Wikipedia talk:WikiProject National Register of Historic Places/Archive 47

This is an archive of past discussions on Wikipedia:WikiProject National Register of Historic Places. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 40

←

Archive 45

tranches and rotation of maintenance for NRHP articles

I'd be interested in rotating out of the maintenance stuff i do on NRHP-related disambiguation pages and list-articles in some parts of the country that i watchlist. I wonder sometimes if Sanfranman59 or Nyttend or others with a lot of NRHP list-articles on their watchlists would also like to rotate out. I've thought that we could split up NRHP material by geographic areas and allow an annual or other periodic changeover, like WikiProject Military History does, but maybe not with its formal elections process. One technical difficulty is that it might be hard to transfer responsibility to a new volunteer, who might not take on a full watchlist that you have built up.

For addressing Biography of Living Persons issues, a new innovation has come up: "tranches" of articles to watch. See User:Tony Sidaway/Living people/tranches. A volunteer can take on a swath of listed articles to watch, easily. This method, with a bot run, could be used to set up NRHP watchlists that could be divvied up differently and rotated. --doncram (talk) 16:04, 9 November 2010 (UTC)

Aside

(this was a subsection within "Please change the standard citation to omit the link" --doncram (talk) 18:32, 17 November 2010 (UTC)
In the above "There are 36,000 articles in WikiProject NRHP now though," answers a question I've had for a long time - What percentage of NRHP sites have articles? The 36,000 may be a bit high for NRHP sites, as there are a few non-site articles in the Project. The number of sites is about 84,000 (?), which would give 3/7ths or about 42%. It seems high - maybe all the county lists make a big difference (2,000 or -2.5% ???).

The other long time question I have is how many of the sites are illustrated. This number could be more or less than the number of articles, since many sites are only illustrated on the county list articles, and not all articles are illustrated. Any idea on how to get this percentage? Smallbones (talk) 19:37, 4 November 2010 (UTC)

The number of articles is 36,420 plus any without wikiproject NRHP banner, minus 3,147 disambiguation pages, minus 2,274 list-articles, minus some other types, in color-coded assessment table at Wikipedia:WikiProject National Register of Historic Places#Collaboration and review. But, there's no easy way to get number of illustrated articles. I manually counted up 2,059 pics for 5,140 New York state articles, in table at Talk:National Register of Historic Places listings in New York#NYS articles status, as of 1/2/2010, and the percentage is perhaps up to 50% by now. NYS is the biggest state but has Pubdog and Daniel Case and Lvklock and DanTD and others adding pics, including Dmadeo who completed out List of NHLs in NYC as one of the first big fully illustrated lists while back, so its percentage is relatively high. --doncram (talk) 00:41, 6 November 2010 (UTC)

That brings a question to my mind too. Is being on the NRHP list alone enough to warrant notability? I have to say I am more of an inclusionist and tend to follow the mantra that most things are interesting to someone but I am curious of the historical determinations. --Kumioko (talk) 19:50, 4 November 2010 (UTC)

This has been discussed before, most recently at the Village Pump here. Every NRHP nomination form I've ever seen has a list of sources that I believe would satisfy the GNG. So if you can find enough to say about a property, any of them could have its own article. Ntsimp (talk) 20:30, 4 November 2010 (UTC)

Thanks and no problem, I figured that was probably the answer but I wasnt sure. --Kumioko (talk) 20:42, 4 November 2010 (UTC)

Fully illustrated and Address Restricted

I know it's rather vain, but I like to finish fully-illustrated lists. See National Register of Historic Places listings in Gloucester County, New Jersey, with 1 missing photo.

We had an extensive discussion recently about Address Restricted sites and photos of them, but I didn't understand the advice on this matter as being practical or specific.

I don't know that this is an archeological site - in fact New Jersey makes learning anything about their sites difficult to find anything about - but it is AR. Should I list this county as fully illustrated? Smallbones (talk) 18:53, 15 November 2010 (UTC)

If I might make a suggestion: whether or not lists with unphotographed Address Restricted sites "count" as fully illustrated, IF the concensus is, in general, not to photograph AR sites, someone should make a little 100X100 "Address Restricted" image to go in place of the photo. I think that would be more visually appealing than leaving it blank or, as Smallbones did, graying the space out, because it would make the image column look continuous. It has the added benefit of filling in all the listings with an appropriate image, so the list "feels" fully illustrated, even if it technically isn't. Andrew Jameson (talk) 21:18, 15 November 2010 (UTC)

I made such an image - see the upper right corner - and will be glad to tweak it as suggested. I also added it to a sample list sp you can see how it looks there. If people like it, I will upload the tweaked version on Commons and delete this version. Ruhrfisch _><>°^° 21:37, 15 November 2010 (UTC)

This is nit-picky, but I'd prefer a non-white background--light gray would be fine, or some reasonable pastel--so that the image visually fills the box it's in. Take a look at the other images in National Register of Historic Places listings in Gloucester County, New Jersey; there's a white one- or two-pixel space between the image and the table border which would be nice to mimic. (edit: Looking really close, the table background isn't quite the pure white of the image, so technically there is a small border, but the colors are so close on my monitor that I can't tell the difference without looking really obliquely.) (And thanks for leaping on my suggestion. Yay cooperation! :) Andrew Jameson (talk) 21:56, 15 November 2010 (UTC)

Thanks - I have added a light grey background. Is this better? Please nitpick away ;-) I originally tried putting the words on a red circle with a diagonal slash, but that looked silly. May have to WP:BYC to see the new version. Ruhrfisch _><>°^° 22:11, 15 November 2010 (UTC)

Nitpicking - I'd prefer a lighter shade of gray, smaller letters, and 100w x 80h. My reasoning is that the "AR picture" should jump out at you a bit less - part of the Great AR Debate was that our blank spaces jump out at editors and make folks want to fill them. A little subtler "AR picture" should do the trick. BTW, I am far from arguing that all AR sites should have this - but if they are subject to looting or just impossible to get info on, this seems the best way. And my original question (vain as it may be): Does it "count" as a fully-illustrated list? Smallbones (talk) 22:44, 15 November 2010 (UTC)

Absolutely, yes, it counts! --doncram (talk) 23:05, 15 November 2010 (UTC)

I agree it counts. I made the AR image lighter grey, with smaller letters, and only 100 by 80 pixels. Is it OK or does it need more tweaks? Ruhrfisch _><>°^° 00:39, 16 November 2010 (UTC)

I like it, and I agree with pretty much everything else said above. Andrew Jameson (talk) 01:01, 16 November 2010 (UTC)

Like it. Ego properly soothed. I've listed Gloucester County on the FI list as:

National Register of Historic Places listings in Gloucester County, New Jersey: 31 listings + 1 AR, November 2010

BTW totaled up the FI list and there are exactly 100 counties (or other geographical areas) and 3120 pix on all these lists. Smallbones (talk) 01:11, 16 November 2010 (UTC)

(out) The image is on Commons now. I had to rename it as File:Address restricted.PNG‎, as there was already a file named File:Address Restricted.png. Ruhrfisch _><>°^° 01:46, 16 November 2010 (UTC)

Yeah, team! :) Have to look thru the Florida lists to see which ones can be considered "fully illustrated" now. Next month, prolly, after my photo roadtripping binge of late will have settled down. And the pics resulting therefrom have all been sorted, uploaded, categorized, and so on. Allons-y! --Ebyabe (talk) 06:42, 16 November 2010 (UTC)

Thank, Ruhrfisch; I'm looking forward to doing my fully-illustrated dance for some Nebraska counties that've hung up on AR sites.

Could I suggest, though, that we not use the AR graphic too freely? I'd propose that it not be used unless and until a serious effort has been made to get a photo for the site. Otherwise, photographers like Smallbones and I, who get an ego-boost from adding a county to the fully-illustrated list, might be tempted to skip some sites that we might otherwise be able to track down and photograph. Ammodramus (talk) 23:06, 17 November 2010 (UTC)

Here i disagree. I'd rather put the AR graphic everywhere it applies, and remove the implicit call for local photographers to search for and reveal the locations of the address restricted places. Only if there is positive documentation that an archeological site is publicly revealed, documentation which meets a pretty good standard (e.g. state SHPO says the place is publicly known), should we remove the AR graphic. Use it freely, remove it rarely. --doncram (talk) 23:17, 17 November 2010 (UTC)

Applying the AR graphic by default would place it on a great many non-archaeological sites. Too-free use of the AR graphic removes much of the incentive for WP photographers (at least for shallow, ego-driven photographers such as I) to find and photograph sites that really shouldn't be address-restricted at all. Ammodramus (talk) 01:45, 18 November 2010 (UTC)

How about - for archeological sites - apply the graphic freely, remove rarely; for obviously non-archeological sites apply rarely, remove freely; where it's unclear apply if there is no good info after a good search for sources, use your judgement in removing. Smallbones (talk) 02:06, 18 November 2010 (UTC)

Is there a list of NRHP articles that still need to be created?

I noticed on the NRHP that there appears to be 85,822 properties listed on the website but only about 36000 have articles in WP. Is there a list or something somewhere that shows which ones still need to be created? --Kumioko (talk) 18:34, 17 November 2010 (UTC)

Look at the red links on the state, county, or other geographical lists, e.g. National Register of Historic Places listings in Nebraska, National Register of Historic Places listings in Delaware County, Pennsylvania, National Register of Historic Places listings in Northwest Philadelphia.

Smallbones (talk) 18:52, 17 November 2010 (UTC)

Thanks, Im building a list of US related articles that need to be created and adding it to the WPUS page. If you take a look in a couple days there should be a bunch there. --Kumioko (talk) 18:58, 17 November 2010 (UTC)

I have a question regarding these missing articles. If a way could be devised to create these articles in an automated or semi automated fashion (they would probably only be stubs though) would that be of interest to the project. To create 50K articles manually over time (assuminng that all NRHP articles are notable that is) will take a long time and after looking at the fairly consistent structure of the content of the NRHP site I think this might be possible. This would allow at least something to be said about them and allow others to expand on them. --Kumioko (talk) 19:23, 17 November 2010 (UTC)

There was an attempt to do that (see User:NrhpBot); I believe it wasn't a great success. I'm personally opposed to the mass creation of stubs (IMO a redlink is a greater to incentive to write an article than a stub is). Niagara ^{Don't give up the ship} 20:04, 17 November 2010 (UTC)

Thanks, the problem with red links is that most users wont create one (or cant) but if a stub for something exists they are much more apt to jump in and edit it or expand it. Based on the conversation developing below though I am less inclined to think its able to be done. --Kumioko (talk) 20:09, 17 November 2010 (UTC)

Template:NRHP-PD and NRHP PD material

I noticed the following dated from this summer Template:NRHP-PD which gives

This article includes text in the public domain from the National Register of Historic Places.

which would be useful if we were copying anything directly from the NRHP to avoid plagiarism problems. I'm just wondering whether there is any useful information that can be copied directly from the NRHP that is public domain?

The obvious question is whether the nomination forms are PD. I'd argue that the are since they are administrative rulings of the government. The principle is essentially the same as for why court cases can't be copyrighted. The nominations themselves ARE the National Register, and the National Register itself is a Federal government document that everybody has the right to access, copy, etc. Nevertheless, I've tried this argument before regarding photos in the NRHP and it has been rejected here. (Are photos somehow different?) The counterargument was that the photos were produced by individuals who had copyright before they submitted them to the NRHP. Does the same reasoning apply to the written government form (the nomination)?

Before anybody goes and tries to copy a nomination directly, I have to say that this would in general make for an awful article. Too much editing would be needed to make for a readable article. But in some cases it could be useful, particularly info from summaries. I'll put a particular case below, which some may think muddies the water for the general issue. Smallbones (talk) 19:16, 17 November 2010 (UTC)

Typically all works legal or otherwise generated by the US government are freely usable. However I think it is very bad form and just plain lazy to copy paste information from any site, even government for use in WP. As far as the images go, IMO if they are displayed on the NRHP site then they should fall under the Govt distribution clause unless stated explicity otherwise or are works done by an artist not working for the gov (such as paintings or statues). Im not a lawyer though. --Kumioko (talk) 19:23, 17 November 2010 (UTC)

(ec)"Generated by" might be controversial here, and I dont know about the "Govt distribution clause." There has been (and probably should be!) fairly direct copying from the General Services Administration site, with proper attribution and good editing to put it in our format. I support this copying in that it gets a very good article up and running right away, and provides a platform for further writing, editing, photos, etc. that Wikipedia can be very good at. The key to avoiding bad form is the attribution - which is what the above template is all about. Smallbones (talk) 19:45, 17 November 2010 (UTC)

Specific example - the Dorrance Mansion is covered fairly well at http://www.nps.gov/history/nr/travel/delaware/dor.htm which at first glance appears to be a NRHP website. No such luck - it is done by a consortium or some such - of which the NRHP is the leader. So no direct copying with attribution - which could easily get rid of dozens of red links if all the material from that website could be used. It turns out that the site's article on the Dorrance Mansion is just a summary of the nomination form - done by an intern AT a local college, not a NRHP intern FROM a local college. My biggest problem with using the material from that website is that I would be plagiarizing the Nomination form (word-for-word in most paragraphs). If the material was PD i wouldn't have that problem (there were other problems, however, such as the accuracy of her summary). Smallbones (talk) 19:45, 17 November 2010 (UTC)

In general consensus has been to avoid using either text or images. Both are in many cases prepared by private individuals, consultants or state officials who are not working for the Federal government either as employees or as contractors. Therefore, regardless of what's posted by the NPS in the fine print at the bottom of the screen, their work is not PD because the NPS has no authority to waive their rights. I've made specific inquiry to the NPS about images, and their answer is that it's the responsibility of the end user to determine copyright/public domain status, and that it is very unlikely that material posted at NPS Focus can be used in a free-content environment; the same condition would certainly apply to text. Acroterion _(talk) 19:34, 17 November 2010 (UTC)

Interesting thanks for clearing that up for me. --Kumioko (talk) 19:43, 17 November 2010 (UTC)

Which would mean that the template is essentially useless - it implies PD material where there is essentially none.!?

I came to a similar conclusions back when this template was created (see Archive 41). Send it to WP:TFD maybe? Niagara ^{Don't give up the ship} 19:58, 17 November 2010 (UTC)

Note that nomination forms completed by federal employees (this is often the case with listings located within national parks) are in the public domain. But I agree, it's poor style to merely copy text from a nomination form. Bms4880 (talk) 19:59, 17 November 2010 (UTC)

Just to be clear - I don't think that articles such as United States Post Office and Courthouse (Pittsburgh, Pennsylvania), which is taken with attribution from the GSA website, are in any way "poor style." We could use many more of these articles and have a link to the source at Wikipedia:WikiProject National Register of Historic Places/GSA federal building links. We should understand that we've got over 50,000 articles to go - see above thread - and can use all PD sources available to us. One similar situation involved copying the PD biographies of all US Congressmen, that was done a long time ago on Wikipedia. I find those copies to be an invaluable resource from time to time, and note that many of these bios have now been much improved. To the extent that we can use PD sources - I think we should, at least as a start to articles. Copying a NRHP nomination verbatim, however, would certainly be bad form - if only because they are nearly unreadable to a general audience. Smallbones (talk) 22:13, 17 November 2010 (UTC)

Fully Illustrated Lists

This has been a banner month for fully illustrated NRHP lists (see project page). I count 7 FI lists completed in November. Dear to my heart is National Register of Historic Places listings in Central Chicago completed by User:Alanscottwalker with 113 listings (6th largest among FI lists). Central Chicago is one of America's architectural jewels and certainly deserves a FI list. User:Bobak topped up large lists in Hennepin and Ramsey Counties, MN (which have a couple of nice grain elevators!). New York State has several - too many to track down all the attributions - probably by User:Pubdog and the usual NY NRHP mafia. User:Ammodramus is keeping up his usual pace in Nebraska. User:Royalbroil completed National_Register_of_Historic_Places_listings_in_Wisconsin#Green_Lake_County. Apologies for anybody I left off. Smallbones (talk) 17:46, 20 November 2010 (UTC)

Monadnock Building at FAC

The NRHP/NHL article Monadnock Building is now a Featured Article candidate. People on this project may be interested in the review. --Nasty Housecat (talk) 21:42, 20 November 2010 (UTC)

NRHP-Gov image links

Where is everybody getting the links of images of NRHP sites from the official NRHP website? Because every time I try, I keep getting the error message "The PDF file for this National Register record has not yet been digitized." I really need this for two sites; Wakefield Upper Depot, and any historic district that Wilson (Amtrak station) might be in, assuming the station is in a historic district. ----DanTD (talk) 00:12, 21 November 2010 (UTC)

As a thought, have you also looked at the HABS/HAER photos here]? Einbierbitte (talk) 17:11, 21 November 2010 (UTC)

Unfortunately, the files containing nominations and photos for NRHP listings in most of the U.S. have not yet been added to the NPS Focus website. --Orlady (talk) 18:08, 21 November 2010 (UTC)

In the past I have had good luck with contacting a user in that area and asked them to take a picture and upload it. --Kumioko (talk) 18:32, 21 November 2010 (UTC)

Einbierbitte, I took your suggestion and none of the HABS/HAER photos had anything to do with either location. I still don't know if Wilson Station is part of any historic district. Rocky Mount (Amtrak station) is apparently part of one, but I forgot what it was. ----DanTD (talk) 04:30, 24 November 2010 (UTC)

Version 0.8 followup

Hi, and thank you for your feedback on the Version 0.8 selection. Unfortunately, I can't see any of the articles being added for now - we look at a lot of stats for judging the articles, but they usually have to be pretty major/mainstream topics to be included. The FAs listed all seem to be fairly specialized. If you know of any specific reasons that the stats may be wrong, or if I've overlooked something, please let me know. Likewise, with the articles you propose removing: These are often crosslisted with other WikiProjects, and may well have been included for those projects' reasons. That means we typically need a strong, specific reason for removing an article, such as gross copyright violations (that occurred on some Wagner operas) or a technical glitch on our part, etc. Please let me know if any of these "to be removed" articles fall into that category. Sorry I couldn't be more positive this time, but I still appreciate the time you've taken to look over our selection. Thanks again, Walkerma (talk) 05:48, 25 November 2010 (UTC)

A consideration for cross project consolidation of talk page templates

I have started a conversation here about the possibility of combining some of the United States related WikiProject Banners into {{WikiProject United States}}. If you have any comments, questions or suggestions please take a moment and let me know. --Kumioko (talk) 04:50, 27 November 2010 (UTC)

Ackerville Baptist Church of Christ

This short NRHP stub, by me, is up for deletion at Wikipedia:Articles for deletion/Ackerville Baptist Church of Christ. Altairisfar 17:48, 27 November 2010 (UTC)

It was speedily kept - with almost all 8 commenters saying that NRHP sites are (almost) automatically notable. Smallbones (talk) 04:00, 29 November 2010 (UTC)

A civil discussion about sources

Hi there. We're having a discussion about a few sources including letters by the National Park Service (regarding eligibility and/or listing on the NRHP) as well as letters and photos by a state historical society and their status in articles about historical buildings. These are sources that were found at the Minnesota State Historic Preservation Office and will impact future NRHP articles in Minnesota. It would be nice to set a consensus policy now, as this could easily be a problem elsewhere. If you find this of interest, please take a look at the discussion at Talk:Salvation Army Headquarters (Saint Paul, Minnesota) and let us know your opinion. Thank you. --Bobak (talk) 07:53, 5 December 2010 (UTC)

Suggested policy change to the tagging of non article items

I have submitted a proposal at the Village pump regarding tagging non article items in Wikipedia. Please take a moment and let me know what you think. --Kumioko (talk) 02:00, 6 December 2010 (UTC)

One percent of Wikipedia

We have 36,640 articles tagged by the WikiProject now, including 2,000 list-articles and 3,000 disambiguation pages. There are some NRHP list-articles and individual articles lacking Talk pages and Wikiproject tagging, and there are many articles tagged by us that are not primarily about NRHP-listed places. But, notice this is 1% of 3,483,756 articles in the English wikipedia (per Template:numberofarticles). I think that's pretty amazing, that we have this many articles in progress, where local editors can arrive and add pics and develop material.

We're being translated into Portuguese now, too: see Anexo:Marco Histórico Nacional na Dakota do Sul for the South Dakota NHL list. --doncram (talk) 00:24, 24 November 2010 (UTC)

1% of Wikipedia articles is pretty amazing. It's probably not 1% of text since our articles are often stubs or otherwise quite short. On the other hand, we do have a lot of nice photos ;-) Yes 1% is pretty amazing. Congrats to (fill your name in here). Smallbones (talk) 04:58, 24 November 2010 (UTC)

Compare to WikiProject SHIPS, with 41,442 now per its stats. It has 1,000 fewer lists, 1,000 more disambiguation pages, but has 7,000 more category pages than we do. So about the same number of basic articles + list-articles. They have many more featured articles and lists than we do, however. --doncram

Thats not even inlcuding the 30,000+ NRHP related articles that havent been created yet. --Kumioko (talk) 16:57, 24 November 2010 (UTC)

glad I could help .... I think I've done abt 3,000--Pubdog (talk) 03:12, 8 December 2010 (UTC)

omissions and incorrect inclusions

I just discovered that 5 NRHP listings during 2008 were omitted from one county list-article, because its March 1st table-izing used the NRIS version then available, and didn't capture items covered in the March 13, 2009 and still-current version of NRIS. I detailed this out at Talk:National Register of Historic Places listings in Douglas County, Nebraska#Park Avenue Apartment District. How many of our list-articles would the same problem apply to? Also recently Nyttend found several list-articles, including National Register of Historic Places listings in Poughkeepsie, New York which erroneously presented as NRHP-listed, places that in fact were not NRHP-listed in the end due to owner objections. I surmise those are cases where table-izing was done prior to our knowing more about the NRIS codes for owner objection. I think overall we've been rather conscientious, but can't stop all errors. I suppose there are other types of errors of omission and incorrect inclusion that we could have made. Sanfranman59 reports that a new version of NRIS is coming available soon. Is there some way we could use the new version to audit the accuracy of some of our county lists? Should we? --doncram (talk) 18:32, 6 December 2010 (UTC)

More NRHP infobox shrinkage

I just found out that the NRHP portion of the infobox in Bayport Aerodrome has shrunk in half. Can anybody fix this? ----DanTD (talk) 02:44, 8 December 2010 (UTC)

Construction equipment on the Register

Having just finished Marion Steam Shovel (Le Roy, New York), and having some pictures of Dipper Dredge No. 3 that I will at some point upload, I'm wondering if we have enough old construction equipment listed on the Register outside of upstate New York to justify a "Construction equipment on the National Register of Historic Places" category (and, by extension, enough created articles). There's got to be a few more out there.

If we do create this, maybe we can also create a higher-level cat along the lines of "Heavy machinery on the National Register of Historic Places" to hold the construction equipment and the locomotives, all to be part of a general "Industry-related listings on the National Register of Historic Places" topic cat I think we could have eventually. Daniel Case (talk) 23:50, 19 November 2010 (UTC)

A general "Industry-related listings on the National Register of Historic Places" topic cat might be too big - as there are a lot of industrial buildings, which I don't think you had in mind. Train freaks might object to mixing steam shovels with locomotives lest their thoughts get derailed.

BTW, I just saw some great grain elevator pix in Ramsey and Hennepin Counties, MN. Smallbones (talk) 17:29, 20 November 2010 (UTC)

Grain elevators are agricultural facilities, not industrial ones. I didn't mean that the "industry-related" cat would contain actual articles, just other cats (save for any oddball listings that probably couldn't fit into other categories, the way San Francisco cable car system is in Category:Listings related to transportation on the National Register of Historic Places). My idea was that since locomotives and construction equipment are both heavy machinery, a cat for both those cats might make sense. I really don't care what foamers think. Daniel Case (talk) 22:02, 20 November 2010 (UTC)

Just FYI, Smallbones might want to check out the page List of grain elevators. --Orlady (talk) 18:13, 21 November 2010 (UTC)

Neat, see Zip Feed Mill for a particularly dramatic photo. BTW Canada kicks US butt on this list, I'll suggest that anybody who knows of an NRHP grain elevator add it to the list. For those of you who don't think that a grain elevator can be beautiful or industrial (or even beautiful and agro-industrial at the same time) I'll suggest taking a look. (Sometimes US east coasters can be so provincial!) Smallbones (talk) 01:22, 23 November 2010 (UTC)

The Franklin Feed Mill makes the list at National Register of Historic Places listings in Sheboygan County, Wisconsin and I took a photo. I didn't find it to be a particularly interesting or unusual structure so I was surprised it was on the register. Royal broil 03:46, 9 December 2010 (UTC)

Having fixed the link above, one can see that the locomotives already have their own category; if you want to make that a subcategory of your proposed heavy machinery category, that's fine, but I don't think it would be just the foamers who would object to losing that category. Mangoe (talk) 17:02, 10 December 2010 (UTC)

Two listings for one property?

A few minutes ago, User:Ebyabe addressed a problem to me; Apparently the Opa-Locka Railroad Station and the Harry Hurt Building have the same address. I have a strong feeling that they're the same place, especially since the picture he took of the Harry Hurt Building looks a lot like a railroad station. ----DanTD (talk) 17:03, 8 December 2010 (UTC)

They aren't, though there are a bunch of people out there who think they are. If you follow the geohack for the railroad station, the pointer is almost due west of the actual station, which is the longer building on the north side of the tracks just eastbound of the present station platforms. The pointer for the Hurt building by contrast is directly south of the actual building. The confusion is that the they have the same street number, but on different streets: the train station is 490 Ali Baba, and the Hurt building, while one side of it is also on Ali Baba, is on 490 Opa Locka Boulevard. It would help if the coordinates were a bit more accurate. Mangoe (talk) 18:02, 8 December 2010 (UTC)

Okay, I just went through Geohack, and I looked at GoogleMaps Street View, and I saw a "490" on the northwest corner of Opa-Locka Boulevard and Ali Baba Avenue(which is not along the tracks). When I zoomed in on it, I found it had the name "Logan Executive Center" over the door. I think we should bring Ebyabe to this discussion. ----DanTD (talk) 18:42, 8 December 2010 (UTC)

I updated the address on his list of addresses and directed him back here. I'm not sure how to go about getting the picture renamed since I'm not active on Wikimedia. Mangoe (talk) 18:57, 8 December 2010 (UTC)

Well, this book says the Hurt Building has a second floor. The building I thought was the Hurt Building has no second floor, but does have 490 on one of the doors. I'll look at this more later, when I've got more time. I can always have the photos deleted and re-upload them with corrected names, if it comes to that. --Ebyabe (talk) 19:02, 8 December 2010 (UTC)

I see what you mean now. Dang, I drove right by that place! I'll do the delete and re-upload the old RR building pictures. I'd hoped not to have to go back to Opa-Locka, it's gone so downhill since I used to live in Miami. Oh well... --Ebyabe (talk) 19:09, 8 December 2010 (UTC)

Well before you go back, can't you just do a mass-renaming of the existing pictures? I recommend "Opa-Locka SAL Railroad Station," or something to that effect. ----DanTD (talk) 21:08, 8 December 2010 (UTC)

Let me know if you need a rename done for the pictures at Commons and I'll take care of it. I'm an admin at Commons. Please leave a message here on my talk page at the English Wikipedia. Royal broil 03:55, 9 December 2010 (UTC)

I happen to have some ready-made renaming tags for each pic. Unless Ebyabe has a reason to object to them, I'll post them on each one. ----DanTD (talk) 16:17, 9 December 2010 (UTC)

Sorry, Dan. Ebyabe had left a message on my talk page and I already did the rename before noticing your comment. Royal broil 01:37, 10 December 2010 (UTC)

That's okay. As long as it was done. Now I'm just going to add them to the article, and even create a commonscat for it. ----DanTD (talk) 13:55, 10 December 2010 (UTC)

All Caps in ship names

Ever now and then I come across a NRHP list with a ship or shipwreck, and frequently the ship's name is given in ALL CAPS. This is because the ancient Federal database lists them that way, presumably because it was not capable of italic text. But is there any reason Wikipedia should copy this practice? If I correct these, will somebody jump down my throat or go around reverting me? Abductive (reasoning) 12:19, 8 December 2010 (UTC)

As long as you go with WP:MOSSHIP, I can't see a problem with that. --SarekOfVulcan (talk) 13:35, 8 December 2010 (UTC)

Discussed recently at Wikipedia talk:WikiProject National Register of Historic Places/Archive 45#ship vs. SHIP names. Please change to lower case italicized, e.g. show "USS Seattle (battleship)". Please post a note here if there is any contention or any reversions of your edits that way. This is a subject where i think there has been nearly 100% consensus out of editors participating in discussion on the topic. Any editor who disagrees should have to explain themselves and participate here, IMO. --doncram (talk) 03:06, 13 December 2010 (UTC)

Sourcing of all infobox information

I was scrolling through my watchlist and saw that someone put a "citation needed" tag on The Kean--specifically, on the "year built" in the infobox. Adding a citation to the date would be trivially easy: not only is the year given in the NRIS, it's also given in *both* of the other references listed in the article. However, the "citation needed" tag got me thinking: what's the policy on providing a source for the infobox info? I've been assuming the NRIS citation on the NRHP Reference# kind of covers the whole box, but should each fact be seperately sourced? Alternatively, should each fact be mentioned in the article text and sourced *there* instead? Andrew Jameson (talk) 12:57, 9 December 2010 (UTC)

Well, ideally, everything in the infobox should be mentioned (and cited) in the prose as well, but it's doubtful that will occur all the time. I only ever use the NRIS citation for the Refnum (as it's a unique identifier that is rarely used outside of the NRHP/NPS and there is no easy way of incorporated that into the article) and date of listing (in the prose). Personally, I'd just expand the article, instead of adding a citation for everything in the infobox. Niagara ^{Don't give up the ship} 15:18, 9 December 2010 (UTC)

In this instance, it appears that the issue may be that the sourced statement in the article text says it was "designed" in 1931, but the infobox says it was "built" in 1931. Since the source says it was "completed" in 1931, it seems to me that the article text should be corrected -- and the "fact" template can then be removed from the infobox. --Orlady (talk) 16:09, 9 December 2010 (UTC)

General thoughts: I agree articles should stand alone without infoboxes. Similiar to lead sections, all information should be mentioned in the body and citations added there, with the infobox as a summary. In fact citations look odd to me in infoboxes. I like to put coordinates in the body too, although some editors have complained and even taken them out. Coordinates are redundant with street addresses, so one or the other in the article also makes sense. I often do put the NRHP id into the body, and cite an acessable source if possible, although agree that might look odd to some. As for the specific issue, as noted above, the single date in the NRHP just needs to be expanded upon in the body. That is, architect designed in X, construction started Y, building opened in Z. The problem is that these are sometimes hard to dig out. I find the LOC newspaper archive good for this kind of thing, although 1931 is too late, alas. W Nowicki (talk) 18:32, 12 December 2010 (UTC)

I think WikiProject Ships tries to add a separate inline reference for each and every separate item within a ship infobox. See for example USS Becuna (SS-319), for a NRHP-listed NHL submarine in PA. That might be a hard and fast rule for their Featured articles, of which they have many. --doncram (talk) 02:57, 13 December 2010 (UTC)

"Address Restricted," Archaeological Sites, and a Special Case, I think.

There are ten shipwrecks listed in the National Register of Historic Places listings in Keweenaw County, Michigan--all of these are from the same MPS, they are all listed as "address restricted," and they're all off Isle Royale (and, I gather, within the boundaries of the Isle Royale National Park). They all have some archaeological importance, which I infer is why they're listed as "address restricted." However, the National Park Service pretty clearly has no objection to letting people know where the wrecks are located. This is the official NPS Isle Royale "scuba diving" page, complete with photographs and desciptions of all ten wrecks, as well as direct links to commercial dive companies one can hire to visit the wrecks. One of these companies is Black Dog Diving, another is Superior Trips; both companies list GPS coordinates of the dive buoys for each wreck. Given that the coords are publically available, and the NPS explicitly allows the public to dive the wrecks, I think it's reasonable to include the coords in the National Register of Historic Places listings in Keweenaw County, Michigan and on the existing individual ship pages. Is there a reason I shouldn't do this? Andrew Jameson (talk) 19:03, 10 December 2010 (UTC)

I can't see any problem with giving the coordinates. Unlike a number of Native American archaeological sites, where they don't disclose addresses in order to prevent looting, I think the main reason they don't list addresses for shipwrecks is because shipwrecks don't have street addresses. (Then again, the Edna G tugboat doesn't have a street address, but it has coordinates. It's still above water and available for tours.) I listed the coordinates for four wrecks I've dove on, the Madeira (shipwreck), Hesper (shipwreck), and the Samuel P. Ely (shipwreck). Thomas Wilson (shipwreck) doesn't have coordinates, so I should add them. In general, if the location of a site is publicly accessible somehow, Wikipedia doesn't need to keep it a secret, even if the National Register marks it restricted.

As an aside, the recreational scuba diving community has actively encouraged divers to leave wrecks as they found them and to avoid taking artifacts, in order that future divers may be able to enjoy diving there. Sometimes they have engaged in preservation projects, like the Great Lakes Shipwreck Preservation Society's projects to keep the Ely from falling apart after being beaten by Lake Superior's weather. --Elkman ^(Elkspeak) 21:08, 10 December 2010 (UTC)

It's always good to check here about adding things to AR sites, thanks. This is pretty clear that we won't be doing any damage by adding the coords, in fact why not add all the NPS pictures? I know A J wants to get all those photos too, but I'm sure that water will be pretty cold for another 6 months. Smallbones (talk) 23:20, 10 December 2010 (UTC)

Yeah; the pix on the NPS page are all individually and explicitly marked as NPS photos, so one of the things on my List of Things to Do was to snag the images to populate the list article (although if someone else wants to upload them, I'm fine with that). That would make the National Register of Historic Places listings in Keweenaw County, Michigan completely illustrated but for one site (which, if we're lucky, there might also be a NPS photo of). And since I'm unlikely to be able to drive to the wreck sites, I'm unlikely to be able to take a photo myself. :) Andrew Jameson (talk) 02:24, 11 December 2010 (UTC)

You could always do what I've done for Florida shipwrecks, and take photos of the water where it's sunk. It's not great, but not being a diver, it was the best I could figure. Some that I've gotten are the Vamar and the City of Hawkinsville. If they're publically used by divers, I don't see a problem with adding the coordinates either. --Ebyabe (talk) 03:42, 11 December 2010 (UTC)

Completely illustrated counties if unphotographed shipwreck(s)?

A related question that I've been meaning to ask for a while is wondering if a county is considered complete if there are some unphotographed shipwrecks. It would take extraordinary means for an average Wikipedian to photograph a shipwreck that is several miles from shore considering that most of us don't own a boat or have deep-water diving equipment and cameras. Royal broil 21:43, 12 December 2010 (UTC)

Have you considered reaching out to local diving clubs (probably via local dive shops) to see if someone who does dive has a picture you could use? I have been trying that method to get pictures of the Land Tortoise on the bottom of Lake George, the only NHL shipwreck in the Northeast and (I think) the only one in fresh water.

Also, there are historical markers on the shore for these things. The Land Tortoise's NHL plaque is onshore as well, and at some point I'll upload a pic of it if I don't hear from the dive group I've contacted, the one whose members discovered it in the first place. Since it's a federal work, the NHL plaque wouldn't create copyright issues with the text. Daniel Case (talk) 05:42, 13 December 2010 (UTC)

In some cases, artifacts from a wreck could be on display in a museum. If so, one could ask for permission to photograph one or two, I suppose. --Ebyabe (talk) 05:52, 13 December 2010 (UTC)

You could also see if you can find who did the exploration of the wreck. Oft-times its a university, and you could try contacting their archaeology department to see if they can help. --Ebyabe (talk) 05:56, 13 December 2010 (UTC)

Or get historical pictures of the ship itself. Smallbones (talk) 15:20, 13 December 2010 (UTC)

Major infobox update

See discussion at Template talk:Infobox NRHP#Meta-template compatible.--Dudemanfellabra (talk) 20:28, 13 December 2010 (UTC)

Infobox maps for AR sites

I am slowly, very slowly, coming along with a major expansion of Pike-Pawnee Village Site, an address-restricted archaeological site in Webster County, Nebraska.

The AR designation is very valid, since shiny things have been recovered from the site, and I'll refer to those in the article. However, I'd like to have some kind of map in the infobox, just because the verbal description ("on the Republican River between Guide Rock and Red Cloud in Webster County, Nebraska") wouldn't convey much to a lot of readers.

My inclination is to use the Commons map showing the location of Webster County within Nebraska, and to use it like a district map in the infobox template, with a caption "Webster County, Nebraska" or the like. Does this seem reasonable, or can someone suggest a better approach?

--Ammodramus (talk) 22:05, 15 December 2010 (UTC)

That's exactly what I would do. I was actually thinking of suggesting the |district_map= parameter as I was reading the beginnings of your comment, and I see that you were thinking the same way I was. If there is any image showing the Republican River, I might go with that, however. Just depends on how specific you want to be without endangering the site.--Dudemanfellabra (talk) 23:51, 15 December 2010 (UTC)

Adding sites to list articles

I've found a couple of sites that don't appear in their county lists, but that appear to be legitimate NRHP listings.

One is the Z.C.B.J. Opera House (Clarkson, Nebraska), which does not appear in National Register of Historic Places listings in Colfax County, Nebraska. It shows up when I look at Colfax County in the NPS Focus site; and it's on the Nebraska State Historical Society's NRHP in Colfax County page. Recent ground-truthing indicates that it's still standing and doesn't appear to have been degraded: see Commons:Category:Z.C.B.J. Opera House (Clarkson, Nebraska). The Elkman tool returns a note: "This property may not actually be listed on the National Register - listing code is DR".

The other site is the East Riley Creek Bridge (not the same as the Riley Creek Bridge), which does not appear in National Register of Historic Places listings in Republic County, Kansas. Both Riley Creek bridges appear under Republic County at the Focus website; both are listed in the Kansas Historical Society's Historic Places in Republic County page. Both bridges are apparently still there, and match the descriptions in their nom forms, available at the KHS page; see Commons:Category:Riley Creek Bridge on 170 (Republic County, Kansas) for Riley Creek Bridge, and Commons:Category:East Riley Creek Bridge on Queen (Republic County, Kansas) for East Riley Creek Bridge. The Elkman tool returns the DR note for East Riley Creek Bridge, but not for Riley Creek Bridge.

Can I add these to their respective list articles, or does this require further investigation and/or discussion and consensus?

--Ammodramus (talk) 00:58, 19 December 2010 (UTC)

About the Z.C.B.J. Opera House one, the Nebraska State Historical Society page for Colfax County that you link states it was listed September 28, 1988. That is not inconsistent with the NRIS entry for it showing "DR" status, which means "DATE RECEIVED/PENDING NOMINATION", as of May 24, 1988. It was pending in May; it was listed in September according to a source. Why is the updated status not in NRIS? I dunno, but there have been a few other cases like this nation-wide, I am recalling one in California and i think some in Oregon. In the California case i was able to further find the NRHP Weekly Announcement listing of the place, also a few months later, which somehow did not get entered into NRIS. I for one would appreciate if u could please just put a note about the case into wp:NRIS info issues NE, to document why we are recording something into Wikipedia that is contradicted by NRIS, and then u should go ahead and add the entry. Note, the Z.C.B.J. Opera House item shows up in a dab page, because i accidentally or deliberately included such items whether they had LI = Listed status or not, while Elkman's county-list-table generator properly excluded them from the main county lists (I do wonder about adding subtables for county lists to show delisted cases, owner objection cases, and/or similar ones). Also, maybe we should look at all of the old "DR" status ones, nation-wide. --doncram (talk) 01:49, 19 December 2010 (UTC)

P.S. Actually the September 28, 1988 listing for that Z.C.B.J. Opera House appears in the weekly announcement at the NPS, on page 156 of the big PDF for all 1988 listings, available by link over at left on NPS's Weekly Listings page. It is definitely NRHP-listed. It's an NRIS omission error. It still would be nice to document this. --doncram (talk) 02:04, 19 December 2010 (UTC)

About the East Riley Creek Bridge, also status "DR" in NRIS as of November 28, 1989, the Kansas site you link to states that it was listed in National Register January 4, 1990. Can u check the corresponding 1990 big PDF for that? --doncram (talk) 02:10, 19 December 2010 (UTC)

Thanks for checking up on the ZCBJ and for pointing me to the Weekly Listings page. I checked 1990 and found both Riley Creek Bridge and East Riley Creek Bridge. They were apparently omitted from the lists in January; they appear on the listings for May 11, 1990, with their dates given as Jan 4, 1990. --Ammodramus (talk) 13:40, 19 December 2010 (UTC)

National Historic District as a phrase

I notice that about 50 usages of "National Historic District" as a proper noun, linked phrase have crept back into Wikipedia. I think these are all bogus and propose deletion of the term, currently a redirect, at Wikipedia:Redirects for discussion/Log/2010 December 5#National Historic District. Please consider commenting there! --doncram (talk) 14:05, 5 December 2010 (UTC)

Whether the term is "correct" or not, people do use it... and that means it is a viable search term. I think having this title redirect somewhere is appropriate... the only question is where to point it to. What about: "Historic District (United States)"? Blueboar (talk) 15:51, 5 December 2010 (UTC)

Rather than split the discussion, please discuss at Wikipedia:Redirects for discussion/Log/2010 December 5#National Historic District. --doncram (talk) 15:54, 5 December 2010 (UTC)

OK... I have done so. Blueboar (talk) 15:56, 5 December 2010 (UTC)

RFD concluded by User:JLaTondre with "The result of the discussion was Disambiguate. If there are remaining issues remaining with the wording (though there does seem to be general consensus), that should be handled at the dab's talk page.". --doncram (talk) 04:01, 22 December 2010 (UTC)

Updated NRIS database files available for download

The NPS has released updated public data files. I'm not sure what date should be used in references for this version. The latest "CERTDATE" in the PROPMAIN table is July 9, for 5 nominations that were pending at the time (four of these were listed August 16 and the other August 23). The CERTDATE for the most recent listed property is June 25 (Main Street Commercial District in Little Rock, AR, announced July 2). I think we should probably use the July 9 date. What do others think?

I'm going to start going through and documenting sites that are listed in multiple jurisdictions, adding them to our list articles as appropriate and accounting for the duplicates in the tally tables. This is something we don't get from the weekly announcements where they only indicate the primary location. I plan to source my additions to this release of the database, so I'd like to come to some consensus on the date.

We may also be able to fill in some missing geocode coordinates using the "spatial" data file. The most recent CERTDATE in this file is May 28. According to the download web page, the Google Earth layers they have available are more accurate than the data in the "spatial" file, but they are only complete through the beginning of 2007. --sanfranman59 (talk) 02:02, 20 December 2010 (UTC)

Thanks for this info. New NRIS references using this can be defined as <ref name=nris>{{NRISref|version=2010a}}</ref>, which currently shows the July 9 date that you suggest.^[1]

^ "National Register Information System". National Register of Historic Places. National Park Service. July 9, 2010.

--doncram (talk) 02:27, 23 December 2010 (UTC)

just do it by bot now

User:Kumioko is suggesting using the next version of NRIS, coming soon, for creating the missing 55,000 NRHP articles. I would like to support that, now, IF certain conditions can be met:

new articles can be formed for redlinks among the now-pretty-well-disambiguated names appearing in state/county/city NRHP list-articles
new articles would be at a "pretty good stub" standard which at least equals the standard set by Elkman's NRHP infobox/article generator system. This standard is pretty high, and includes:
1. Infobox including name, NRIS reference number, date, etc.
2. Infobox includes coordinates (there may be some issue about which coordinates data to draw from; there are different coordinates databases and versions)
3. Infobox includes MPS / MRA documents based on wp:MPS table of document URLs
4. Categories are included that are relevant, including county and architectural categories
5. Talk page with state-specific wikiprojects included (including getting the Georgia (U.S. state) one correctly) an d reqphoto statements
new articles would include relatively easy to implement improvements:
1. A basic introductory text drawing on available NRIS information should be included. This text should be different for historic district articles vs. regular properties. It should draw from NRIS fields on historic use of the property, and state the area of the property, and otherwise put out more info in separate sentences that can be manually adjusted later.
2. new articles in states where the NRHP nomination documents are online, should from the beginning include references composed for those documents (references which would still require some manual fine-tuning to add author name, other)
3. MPS / MRA documents, where available, should have references composed (also requiring further fine-tuning)
the article creation should be rolled out in a few test counties or states at a time, preferably first in states where NRHP nomination documents are online from the National Park Service, allowing for editors to absorb/check/improve those articles and to identify improvements for a next phase
local editors who are working on article creation in a given area, i.e. have reasonable expectation to create all articles in their area in next couple years, can request that their area be excluded from the mass creation
I would very much hope that concerns of Elkman and other v. experienced editors be invited and addressed

I hope this can be considered as a reasonable proposal and that moderation in discussion can prevail. It should be noted:

There has been considered, intelligent opposition to mass stubbing for at least a couple reasons:
- Stubs might be poor quality. I myself opposed one previous stubbing campaign, mainly because i thought it created articles that did not meet already-achievable standards.
- Stubs might take away editors' pleasure in creating brand new articles.
- Probably other good concerns have been raised.
There have been good results from previous stubbing campaigns, e.g. Pubdog's completing all the Maryland articles with good use of a Maryland-specific webpage system; a nearly complete New York campaign using the NRHP documents in New York States' system; all stubs in some Connecticut counties and cities being created using NRHP documents online at NPS; all stubs in Henrico County, Virginia being done with NRHP documents; all Massachusetts and Florida articles having been started also.
A reason to create articles is to finish off the issue of redlinks on disambiguation pages, which has been quiet recently, but is removed as an issue entirely if articles are just all created. After a mass creation is done, any remaining NRHP redlinks on disambiguation pages would represent errors or needs for redirects and then could be productively all resolved.
The main reason to proceed is that it would directly advance Wikipedia greatly, and allow for more rapid follow-on improvements. It would allow for many more new/future editors to contribute photos and material. It would allow experienced editors to get out of tedium of the first steps of article development. It would allow for smart AWB or bot-like tactics of mass improvement of articles. It would put all the NRHP places into reqphoto reports of photos needed by state.

--doncram (talk) 20:25, 4 December 2010 (UTC)

Sounds good in theory - but it needs to be fleshed out a bit. Could you give an example of what this type of stub would look like? What does NOW mean - i.e. what type of time frame would be the goal? Is there some type of bot involved? Off the top of my head, if there were 10 editors involved doing 10 stubs per day for 5 days a week: 55,000 stubs /500 = 110 weeks or about 2 years. Obviously, I'm missing something. Smallbones (talk) 21:14, 4 December 2010 (UTC)

I meant for a bot run to create stubs for all 55,000 NRHP articles needed. Getting a bot to equal/surpass Elkman's quality would take some effort, say it takes a month or two, then do some testing for a few counties, then rollout all 55,000 articles a couple months from now (excluding some counties where editors prefer not to have them done). --doncram (talk) 21:25, 4 December 2010 (UTC)

I think its definately worth the effort. Before I make any promises though I would need to see the state and format of the data in the access database. Once i see that I will be able to make a better determination of the effort involved and the end-state article quality. I think it is absolutely worth the investment of time. Even if we shoot for completing the articles by June it would be a marvelous achievement. I also agree that the article quality should be decent and not just cut and paste a bunch of crap from the site. I also agree that we should do a roll-out rather than a mass dump to make sure that the article quality is maintained and that any problems (there are bound to be a few snags along the way) are worked out. If I can get the bot working it could potentially do several hundred a day...or a set number based on the group/county/category/whatever we pull in at that time. --Kumioko (talk) 00:36, 5 December 2010 (UTC)

There was a bot several years ago, the NrhpBot (see User:NrhpBot and [[User talk:NrhpBot) which created a few hundred articles in Ohio to start, using NRIS info, but apparently it did not have a lot of support. Seeing its work later, i thot it needed more refinement. Since then Elkman's standard has gotten to be pretty good for the infobox and for categories, though Elkman's version does not draft main body text. Kumioko, regarding requirement 1, can you create a bot that works from a given NRHP list-article, which would create articles for just the red-link items in the name column, and using the names supplied there? Those names now reflect good disambiguating work. We don't want a bot mindlessly overwriting articles due to NRIS names being non-unique (e.g. there are lots of NRIS listings for "First Presbyterian Church", but our list-articles have bluelinks for some and redlinks for others, in format "First Presbyterian Church (City, State)"). --doncram (talk) 15:54, 10 December 2010 (UTC)

P.S. This original version of Hughes Manor article is one created by NrhpBot. --doncram (talk) 16:42, 11 December 2010 (UTC)

The premise that a stub will encourage editors to write on the subject is untrue. Many stub creation drives in many topic fields, whether by bots or people, have instead resulted in stubs that never improve. Worse, these stubs need improvement the moment they are created, sinc they always contain errors and lack the important local imput. Abductive (reasoning) 19:39, 11 December 2010 (UTC)

I see your point, but don't completely agree. A number of the NRHP stubs I created long ago have been expanded. I've always believed it's more likely a stub will be expanded than a redlink made into an article.

On the other hand, it does make me think of a potential problem. We still have stubs questioned for notability on occasion. Should we create thousands of new stubs, it's likely that would increase exponentially. I'm not saying don't make them, just that it's something to consider. --Ebyabe (talk) 20:16, 11 December 2010 (UTC)

I would prefer to see them not created by bot but rather do them by hand. A bot will only have minimal information and I keep seeing awesome articles done by hand. When I see a list of redlinks for a county, I'm able to see what needs to be done. Having bot-created articles will mask it. Royal broil 20:58, 11 December 2010 (UTC)

[edit conflict] I have to agree with Abductive, I'd rather be given the oppurtunity to start a whole article than to expand / cleanup someone else's; now, granted, I will expand stubs if I come across them in certain locales, but it's for the same reason a fish will bite a fishing lure, not because they're hungry, but because it's funny/strange looking and they can't stand looking at it. Generally, when I do expand those stubs, it's amazing the interesting details that are uncovered if only a little more time was spent doing some research; something no bot could ever do. Niagara ^{Don't give up the ship} 21:05, 11 December 2010 (UTC)

I expected you'd want Pennsylvania NRHP articles kept out of the drive, which is okay by me. Actually, the bot can do research, or at least facilitate it. It could include the NRHP nomination document as a direct link for states where those are available(part of my proposal), and it could further include a template providing link to search results on the topic, calling for readers to add information and to fix up what the stub can provide. We could use a customized template, adapted from the template:find sources. See, for example, Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL . Call this template:Find NRHP sources; see Template talk:Find NRHP sources for development of that. --doncram (talk) 22:24, 11 December 2010 (UTC)

I concur with USer:Niagara and User:Abductive. While I can appreciate the desire to reduce the number of redlinks, I've found that for the New York articles I've been working with where "skeleton stubs" exist to create my version of an "expanded stub" (i.e. link to the Registration Form and photos with critical features of the property noted and anchoring the infobox items) that I'm basically starting the article all over again. I've found it's just as much work as starting from scratch.--Pubdog (talk) 22:29, 11 December 2010 (UTC)

Pubdog, if you could have a bot run for a New York state county, right when you were starting to do that county, I think you would find it helpful and would choose to have it run. It is harder to write a bot to start articles for Pennsylvania and for New York, though, because those states have idiosyncratic state department systems where the NRHP noms are located. Starter articles for Connecticut or North Dakota or other states where NRHP noms are systematically on-line at NPS, would be better. In these states, it would help you a lot more, to have stubs created by a bot. I am, personally, bored by the repetition in beginning work at the current level of cut-and-paste from Elkman output, and there are about 12 steps after that which i perform which are routinizable. Also, the purpose is not just to reduce redlinks for the sake of reducing redlinks. One other reason is to reduce the considerable work involved in dealing with wave after wave after wave of new editors arriving at disambiguation pages and making problems, work which has been almost all done by me personally. Another reason is to facilitate in organized drives, such as creating all the U.S. Post Office articles to help deal with article naming issues for all of them. --doncram (talk) 22:57, 11 December 2010 (UTC)

So if I am understanding this correctly you would like to skip the bot and create the 55000 articles by hand, is that correct? Although I agree that, of course adding an article by hand is almost always going to be better than by bot I wonder how long it will take to create all those by hand. --Kumioko (talk) 23:12, 11 December 2010 (UTC)

What's the hurry? There is no deadline. Niagara ^{Don't give up the ship} 23:41, 11 December 2010 (UTC)

I'm with User:Pubdog, User:Niagara, and User:Abductive on this. As a WP reader, I find stubs somewhat annoying. To me, a blue link in an article represents the promise of another article of substance. Too high a frequency of stubs could lead users to conclude that WP was a less-than-useful information source.

A problem with a mass stubbing campaign, especially one conducted by bots, is that it'll lead to duplicate articles, which will have to be merged when discovered. For example, the NRHS name Pawnee Indian Village Site was a redlink a week or so ago, before I turned it into a redirect to the existing Pawnee Indian Museum State Historic Site. A bot, or even a human editor trying to turn out lots of stubs quickly, might not have found the existing article. A live editor who's interested in the subeject and trying to produce a useful article is a lot more likely to catch that sort of thing.

--Ammodramus (talk) 23:54, 11 December 2010 (UTC)

I don't see that last point as a problem, but rather it suggests all the more reason to proceed: to facilitate our making good connections to articles. Your example is really pointing out that the NRHP list-article National Register of Historic Places listings in Republic County, Kansas was lacking, by showing a red-link, rather than connecting by a redirect to an article that had been created. I think creating stub articles often leads to discovery of related articles which they could be redirected to, and that is a good thing! There are lots of wikipedia editors and readers out there who do swarm over new articles, and make those connections. And, those kind of connections have been being made since the list-articles were completed out in the drive ending July 4, 2009. So you found one example, but there are many fewer to find now, and maybe creating articles is the best way to force them to be found.

I hear the concern about not starting too many at once, beyond what can be absorbed by ourselves and other editors swarming over the new ones. I opened this thread and might have been perceived as calling for a bot to create ALL the stubs right away (and maybe that is what Kumioko would be inclined to do), but I really mean for it to be a good tool to use in a somewhat more focussed way. For example, I would like to run the bot on one county at a time in each of several states, and call for editors there to swarm over them, then do another county after the one had been somewhat absorbed. My having created stubs for all of New Haven, CT is one example of where other editors have come in pretty well. I think that list-article and its articles' existence really helped spur one competitor in CT's recent photo contest, who i suspect might be a Yale student, to participate in that contest and to make a big contribution of many photos there. Maybe running the bot to help create articles for cities of major universities is a good focus, too.

Also, let's keep in mind that anyone can create stub articles, and I/we are not really asking for a Vote on whether or not to proceed at all. Some kind of more automated process will be run, if only by me running a private version of my own to give me material to cut-and-paste into targeted article drives, like for Odd Fellows Halls or for U.S. Post Offices. I do hope that there could be suggestions on how to meet and improve upon the current standard of using Elkman's output, for when a bot is run. --doncram (talk) 00:43, 12 December 2010 (UTC)

Before we make the decistion to, or not to, create articles in an automated fashion lets see what the data looks like and go from there. Its my guess that the data will be in columns and rows and will probably contain most or all of the fields on the package that is submitted to get a location added to the NRHP. If this is the case then I should be able to create a basic article format that can be used and generate a low start with the infobox. I'm guessing somewhere similar to 330 West 42nd Street. This also depends on how you define a stub. If you define a stub as an article like 16 Altamont Terrace then I am almost certain I can do better than this. Until we get the data though we just won't know. It could very well be that the whole idea is completely impossible based on the data. --Kumioko (talk) 01:04, 12 December 2010 (UTC)

Here is Hughes Manor article in version that a bot should be able to provide, which is Elkman's output plus text that could be automatically created from the available NRIS fields for historic subfunction, for architecture, etc. Not all possible fields were populated for that NRHP place, so there are bells-and-whistles not illustrated for that one case. NRIS is a database, which you can read in Microsoft Access or other database software. The Elkman basic version, with NRHP infobox and some more material, is produced by this report from Elkman's system, based on the March 13, 2009 version of NRIS database. The NRIS database, supposedly soon to be updated, is available for downloading here. --doncram (talk) 01:29, 12 December 2010 (UTC)

Here is St. Luke's Chapel (Stamford, Connecticut) in version a bot should be able to provide], based on NRIS plus the table available at wp:MPS. This is a good example because it includes a NRHP document (available online for CT) and also an MPS document (whose title is identified in NRIS, and whose URL is available in the wp:MPS table). The 330 West 42nd Street example is for a place that is actually complicated because it is further designated as a National Historic Landmark. We have created articles for all National Historic Landmarks already, so the bot doesn't need to address those complications. The current version of 16 Altamont Terrace is short but i think it would be hard to do that well; that is one of the Maryland ones well-started by Pubdog using the Maryland website info, not just NRIS. I would hope for just a lower standard like the St. Luke's Chapel and Hughes Manor examples, for now. You need some examples of historic districts, which have to differ somewhat, too. --doncram (talk) 01:50, 12 December 2010 (UTC)

If, Kumioko, you are expecting to create articles as detailed as the 330 West 42nd Street one, you, my friend, are expecting way more than you should from the NRIS database. As Doncram states above, this article (though still quite short in my opinion) is exponentially more complicated and well-structured than a bot running from the NRIS data could create. In all reality, the two examples Doncram shows above – Hughes Manor and St. Luke's Chapel – are about the quality I would expect to come out of a bot.

I am opposed to the creation of these bot-articles solely because of their low quality. "Articles" like Hughes Manor and St. Luke's Chapel are better off redlinked in my opinion until they can have justice done to them. Now if a bot were to come along and be able by some magic pixie dust to create articles of the 330 West 42nd Street quality, I would emphatically support this effort and would devote many hours to making sure it went off without a hitch. Alas, though, a bot is not a magician, so this will not happen any time soon. The best that this bot will be able to spit out is the crud that Doncram has shown above. If you are fine with articles of this quality cluttering up Wikipedia, then by all means go forward with this bot. I, however, will remain vehemently opposed to the idea in its current form.--Dudemanfellabra (talk) 04:53, 12 December 2010 (UTC)

Okay, derisive language is not called for. You can choose to get angry that i or other editors haven't done something you want done, like develop those two particular stubs further, but that won't get you anywhere. Meanwhile, many other stubs that i started, like Bremo Historic District featured as a DYK on the front page of Wikipedia right now, are constructively involving new and old editors and becoming quite nice. The current version of Hughes Manor and current version of St. Luke's Chapel are fine as they are now, with just a tad of editing following on what a bot could provide. Luckily, we don't need any permission to just proceed, as it is Wikipedia policy that stubs are allowed and in fact wanted. --doncram (talk) 15:05, 12 December 2010 (UTC)

If a bot is capable of doing what doncram did with the Bremo article, then I'm sure everyone would be fine with the bot. I know I'd withdraw my opposition. Mass creating articles with a bot against consensus is a problem however. Is this what you are suggestion doncram, or did I misunderstand your comment? Are you replying to possible complaint against human-created stubs? Royal broil 21:28, 12 December 2010 (UTC)

This is the stub version of Bremo Historic District created during drive that created articles for all National Historic Landmarks in 2008. It included NRHP document and NHL summary webpage links (with a formatting error that was corrected later), and coulda nearly been created by a bot, plus a little human retouching. The current version of Bremo article was nicely much more developed, mostly recently by editor KimChee. My point is that many stub articles do get further developed later. And stubs like the Bremo one once was, and like the St. Luke's Chapel one right now, are a good contribution for providing starter information and especially for linking to the good NRHP document sources available for those ones. At this point i am saying i would very much like to have a bot tool going, to make focussed series of NRHP stub articles where desired, working off of any list like List of Odd Fellows buildings (which i did all manually) or National Register of Historic Places listings in Cass County, North Dakota (which is in a state where NRHP docs available) or the U.S. Post Office and Courthouse dab page (done manually recently to address article naming issues). In the last year i've created a good number of such series manually, and i'd rather have a bot do the drudgery, letting me focus my effort on some touchup and also calling in other editors. Does that clarify? Royalbroil, if u wish for Wisconsin county list-articles to be excluded, that is fine. --doncram (talk) 22:01, 12 December 2010 (UTC)

Good, I hoped that you didn't plan to have a bot create articles against consensus. I'm sure a bot operator wouldn't spend time under those circumstances. So far I count 5 opposed and 2 supporters (I don't know smallbones' position but s/he questioned the plan which may indicate a tendency toward opposition). I don't know others' opinions, but I have a strong preference to full articles appear slowly one by one - all done by hand. As noted above, there is no time limit to finish Wikipedia. I understand the importance of stubs as I've been here for many years - back when stubs were very important to get the ball rolling. However, at that same time, many people were posting unsourced articles based on tribal knowledge off the top of their head. Wikipedia has developed a lot since those years so I don't see a big need for 55k quick stubs now. The stubs that I encounter now which were made about 1 year ago have seen little editing or improvement. And I wasn't speaking on behalf of Wisconsin, I did mean the entire U.S. Royal broil 01:09, 13 December 2010 (UTC)

The individually-crafted article approach has served the project well. I see no reason to change; we're not in a hurry and I see little value to be gained by creating articles that amount to placeholders. Back in the Good Old Days, this kind of automated article creation was seen as necessary to create a critical mass of articles. Nowadays the emphasis is (and should be) on the creation of high-quality articles, which can only be done by human authors. I appreciate the enthusiasm behind the proposal, and understand its attraction, but must oppose the use of a bot. Acroterion _(talk) 22:55, 13 December 2010 (UTC)

I'm not sure that it is good to know where several of you all stand on the hypothetical question of whether you like what a bot would do or not, because I think you're presuming that a bot could only do as well as the Elkman NRHP generator. Negative protestations aside, the fact is editors here do use the Elkman NRHP generator to start articles. I do believe that editor Orlady started one or two NRHP articles with deliberate point not to use it, but editors Acroterion, Ammodramus, Daniel Case, Dudemanfellabra, Pubdog, Royalbroil, Smallbones, and every other NRHP editor do start with Elkman NRHP output when starting articles, I believe. Unfortunately, Elkman is no longer developing the NRHP generator. Elkman has kindly maintained the generator, and provided it at his own website, and it does provide a great service, and I am not at all complaining. We are all volunteers.... But, we can get better starter articles. I think it is crazy to want to start articles at a lower level than can be done, when there are so many to start.

Let's try another question. Please consider National Register of Historic Places listings in Cass County, North Dakota. I was developing and used a semi-automated approach yesterday and today to develop starter articles for about half of its NRHP listings. There were a few articles started already, all i think lacking the NRHP nomination document which is in fact available for almost all of these. I think in the new articles that i started, that there is some good starter information. I used the semi-automated approach to get an article like this first version of Robert Lindemann House and then with another few manual edits i refined the NRHP nom doc reference and developed the article a little bit, resulting in this current version for that one. Don't you think the readers and editors of North Dakota are better served by having some or all of this done? --doncram (talk) 15:16, 14 December 2010 (UTC)

Yes, many of us use the wonderful tool that Elkman has created and maintained; in fact, it is even talked about in the style guide of this project. The problem with your analogy, however, is that the infoboxes that come out of Elkman's generator are not currently cluttering up Wikipedia; they are on a completely separate site where average readers are never bothered by them. If an editor hasn't gotten around to "fixing up" the code given by this generator, no readers can see the raw code until someone has enough will to create a decent article about it.

With this bot that you're trying to get going, there will be thousands of these "raw code" articles in contact with readers every day, and they will most definitely be created faster than we can fix them up. In the interim period, many will have an unsatisfactory experience – as several people mention above – by clicking on the links and finding a bunch of blanks (I mean literal blanks.. how can you see it as helpful at all to have "It served historically as a ___."?!) and "maybes" and "and/ors" around. Like nearly everyone here has said, a bot – given the present state of the NRIS database – can never be as good as a human editor with the will to create the article.

Now all that said, if you could somehow convince Elkman to upgrade his system and implement all these ideas, then I would be all for that because the "raw code" wouldn't be on Wikipedia coming in contact with readers. We could use the tool the same as we do now, except for we'd have less to fix up when we started a new article. The point is, though, that humans would still do the actual editing on Wikipedia.

If you can't convince Elkman to upgrade his tool, then why not ask around for people that can create another one? Maybe Kumioko has a website like Elkman? Or why don't you figure it out and create your own since you're so supportive of the cause? I do sympathize with your view of making it faster and less of a hassle to start articles – and I would expect many here do as well – but our problem is that we don't want this raw code on Wikipedia at any point. As many have pointed out, there is no deadline, so don't worry about it. Maybe if your editing feels a bit monotonous, you should try slowing down on article creation yourself and spend more time with each one. That Robert Lindemann House has a wealth of information in the NRHP document that you've yet to extract that could make for a pretty decent article if you took the time to put it in there. You'd be doing something different, so the monotony of it all would be suppressed, and who knows... you may even like it ;).--Dudemanfellabra (talk) 19:15, 14 December 2010 (UTC)

(ec) Oh, is there still a ___ in one of the articles? I am so sorry. It would be great if you'd help edit them. I don't mean to leave any such blanks in an article. The generator version that i am working on is still being refined; i'll try not to leave any more such blanks.

Actually, i find no such article, not one where a ___ got left in. What are u complaining about? My current version of generator does require a little tidying by a following edit or two, such as to join lines and remove some extra spaces. --doncram (talk) 22:57, 14 December 2010 (UTC)

What I'm complaining about is that if we went through with a bot and added thousands of articles automatically, they may have blanks in them. There will be some indeterminate amount of time between the bot run and when human hands can get to the newly created articles. In that interim time (which could be many months, depending on how many articles are created and how fast we go through them), there will be crappy things like blanks in these articles that I simply am not going to support no matter how much you try to rationalize their existence.--Dudemanfellabra (talk) 01:00, 15 December 2010 (UTC)

About "and/or" in an article, such as in the sentence "It was built in 1879 in Italianate style, and was designed and/or built by Pray Brothers.[1]" included in the James Holes House article, that relates to ambiguity in the NRIS field that Elkman's generator presents as "architect=". I happen to think that it is fine and good to have such a correct statement in an article in mainspace. That invites the reader or local editors to dig into the available source, the NRHP nomination document, and figure it out more properly. This is better than what appears in many hundreds of NRHP articles, where it is blithely stated in the NRHP infobox that a place has "architect = Pray Brothers" while there is no clarification in the text that maybe they designed it or maybe they were builders. In my view, it's not better to state nothing, when the only surely correct statement is an ambiguous one. There are many incorrect NRHP articles where subsequent editors believed the infobox statement and wrote out that the named parties were architects or designed the place, where it happens they should not have made such an assumption. --doncram (talk) 21:12, 14 December 2010 (UTC)

The James Holes House article is the only one where the NRHP doc is not available online, so the ambiguity will remain there, unresolved until someone requests a physical copy. --doncram (talk) 22:57, 14 December 2010 (UTC)

This first generated version of Duane Mansion article shows a correctly ambiguous statement, which led me to provide, a few steps later, the clarification that Solomon Kelly was the builder, as appears in Duane Mansion now. The architect is not known. An article without the ambiguous statement would likely have shown "Architect: Kelly,Solomon" in its infobox, alone. --doncram (talk) 21:21, 14 December 2010 (UTC)

But an article created by someone who has done the necessary research on a place before creating the article would have never had the problem in the first case. That's my beef.--Dudemanfellabra (talk) 01:00, 15 December 2010 (UTC)

I'm still opposed to having a bot generate full articles for two reasons. One is that a bot can't completely and accurately add all the categories to an article. The article Northern Pacific Railway Depot (Fargo, North Dakota) didn't have the category of Category:Railway stations on the National Register of Historic Places, nor did it have the category of Category:Cass Gilbert buildings. There are other structural issues that a bot can't fully complete. But, more importantly, a bot can't do the research that's needed to make an article complete and interesting to the reader. For example, while Northern Pacific Railway Depot (Little Falls, Minnesota) is still kind of a stub, it at least mentions the Little Falls depot in the context of Cass Gilbert's career. Yeah, neither one of these depots is as important as the Woolworth Building, one of Gilbert's later works, but given that Cass Gilbert is a notable architect, there should be more to say about the building. In fact, here's an easily-available source. I'd much rather focus on improving and expanding the existing stub articles than on generating new articles with only the thinnest of content. --Elkman ^(Elkspeak) 20:58, 14 December 2010 (UTC)

Thanks for that Cass Gilbert Society link, which i just added to the Fargo depot article. Why not just create all the missing Cass Gilbert-designed NRHP places articles now, and put the Cass Gilbert buildings category into them directly? I found that just creating all the List of Odd Fellows buildings articles was a great way to get the corresponding thing done for those. And it allowed for category-focused editors to consider whether the category should be further split by state or whatever, once they could see how many there were. Having all the articles created would allow an editor to add such good Cass Gilbert Society links to them all, efficiently, too. --doncram (talk) 22:45, 14 December 2010 (UTC)

:Facepalm:--Dudemanfellabra (talk) 01:00, 15 December 2010 (UTC)

I had to follow the link, to learn that was "an expression of embarrassment, frustration, disbelief, disgust or general woe". Hmm. I don't care for the disrespect. But, well, why should we withhold the information from the Cass Gilbert article that there were a number more notable buildings designed by this person, which were NRHP-listed and mostly are surviving. I'll add the NRHP-listed ones to the article or to Talk:Cass Gilbert shortly. And why not turn the redlinks among those, into bluelinks, to get the development started sooner rather than later? I myself have created a bunch of architect articles basically with lists of NRHP-listed places they designed, and I think all this information really helps out a lot. I get the feeling that some might want to keep this information secret? Maybe i am too impatient in some way, but I just don't get the negativity. --doncram (talk) 03:46, 15 December 2010 (UTC)

I suspect that Dude's reaction is mainly on the frustration side. Almost everyone but yourself is opposed to the idea of having a bot creating the articles but you keep advocating for it. I mean no disrespect toward you in any way. For me, Elkman hit the nail on the head. I'm opposed to mass creation of stub articles with "thinnest of content". Let's spend the time to do these 55k articles right way instead of blasting out stubs. Using Elkman's tool is a great idea for a starting point when someone wants to start an article. Royal broil 04:28, 15 December 2010 (UTC)

I have to weigh in with Doncram. I can't see why anybody would object to a tool that can be used 1 article at a time with editor followup - that really is what the infobox generator is now. If a tool is designed to be used for 10 articles at a time with quick editor followup - you'd get about the same results. Yes there might be some abuse, just as there is some abuse now with the infobox generator. But the "abuse" has a very low downside - there are articles that don't even give as much info as the infobox - but so what? They can be added to, have photos added, have related links added to related articles very easily. A bad stub is really somewhat better than a red-link imho, but I could understand somebody with the opinion that it's slightly worse than a red-link. What I can't understand is somebody thinking a bad stub is much worse than a red-link - why? you can always start again from scratch. I suggest we give our blessing to a small batch stub creator, monitor its use so that at least there is some human review by the stub creator, and see what happens. There's almost no downside, and I believe it is within Wikipedia rules even without the project's concurrence. No reason to fight it, let's help it instead. Smallbones (talk) 04:49, 15 December 2010 (UTC)

Thanks for speaking up. You're wrong about having a bot do anything against consensus. The bot policy at Wikipedia:Bot policy#Bot_requirements says, "performs only tasks for which there is consensus". Royal broil 05:52, 15 December 2010 (UTC)

See Talk:Cass Gilbert#NRHP ones for a list of the NRHP-listed Cass Gilbert buildings. I created articles for those lacking, except for 3 Minnesota ones. I've repeatedly heard of this architect, i guess for the Woolworth building and for his deigning to design the railroad station in New Haven, Connecticut. But i had no idea he was a pioneer in warehouse design, for huge reinforced concrete structures such as the Austin, Nichols and Company Warehouse, Brooklyn, the U.S. Army Military Ocean Terminal the Brooklyn Army Terminal that disembarked 85% of U.S. troops during the world war, and R.C. Williams Warehouse, Manhattan.

I am "embarrassed, frustrated, disbelieving, disgusted and woeful" that WikiProject NRHP had not previously done its job, for architect editors and for readers, to creating those articles and more, such as the crucial Cass Gilbert National Register District article. :) Seriously, what to be most concerned about here depends on your perspective, your goals. I think our goals should be to get info out there and to facilitate many other potential editors' continuing work, not to save up topics for us to work on individually as part of a small private club. I prefer Gilbert's massive public works, over a charming boutique cottage for some rich fancy-pants. --doncram (talk) 15:46, 15 December 2010 (UTC)

Of the stubs you've created, about how many, percentage-wise, have been expanded since their creation? Bms4880 (talk) 16:03, 15 December 2010 (UTC)

Hard to say. I browsed in List of NHLs in TN to check some that i started during the NHL drive ending July 4, 2008. Rattle and Snap and Hiram Masonic Lodge No. 7 and others have been edited since, but weren't adopted by any Tennessee editor, and are not much larger than how i started them. You can show the way, but u can't make the local volunteers do anything. But I think the pages have served okay, supporting the NHL list-article which itself gets a fair amount of viewers, 1000-2000 per month. And i browsed our GA, B, C, Start articles and see some that i created, many that i contributed to early on and/or wanna take indirect credit for (because i specifically provided an editor with an example to work from, or the like). Sorry can't be more quantitative easily. --doncram (talk) 20:49, 15 December 2010 (UTC)

Just a note... the community currently frowns on bot created articles. While they are not banned outright, they are strongly discoraged. If you are going to go this route, make sure you follow Wikipedia:Bot policy#Mass article creation. Blueboar (talk) 16:52, 15 December 2010 (UTC)

break

I don't know where the "are strongly discouraged" for bot created articles comes from. It's not in Wikipedia:Bot policy#Mass article creation. But this does raise the question of how approval for the bot would be obtained. I'll ask doncram to fill us in on the general strategy/procedure.

I'll also ask folks not to say "I'm completely against bot creation" but rather focus on "What the bot needs to do in order to gain my support." It's pretty clear that a bot that would just take the infobox and create a few sentences from that (or similar) doesn't have general support from editors here. OK - I don't think doncram would go ahead and do that without support. I'll suggest that a bot that could do the following in the manner suggested should have support and ask others to put in their lists.

A bot should:

include in proper English sentences the large majority of the information now given in the infobox. Not all the information needs to be put in text form, e.g. the coords which are in the infobox and at the top of the article - I view coords in the text as overkill most of the time.
provide a live link to a reliable source.
it should not create thousands of articles at a time that appear without human oversight, rather
- it might create 20-50 articles at a time, with the subjects or titles requested by users (including doncram who I assume would be the operator)
- with the results put in a sandbox or sub-page of WP:NRHP
- from which human editors could "promote" the stubs to real articles given that certain requirements are met, e.g. no obvious grammatical errors, ___s, with at least 100 words.

I think this type of article creation is important because with 500 articles created per week, we could approach "complete coverage" of NRHP sites within 2 years, or perhaps only 100 articles created per week get "complete coverage" within 10 years. Now I'd guess we are doing 20-50 new articles per week. That's 20-50 years.

"Complete coverage" is important if only because of the rotten job the NRHP does in getting its information out on the internet. For large areas of the country a high school student cannot go to any easily searchable database to find basic information on an NRHP site that is right down the street from where he lives. And if he can find that internet site, much of the info provided is essentially unreadable to inexperienced normal human beings. Having a readily available site like Wikipedia that could give 5 or 6 basic pieces of information (architect, date built, date listed, style, etc.) on all but the newest NRHP sites would be a great service to the entire country and would also attract many new editors to Wikipedia who could improve the stubs. Smallbones (talk) 18:37, 15 December 2010 (UTC)

I could support such a procedure with a holding pen for articles to be reviewed and further edited before being promoted to article space. Acroterion _(talk) 18:46, 15 December 2010 (UTC)

I also like the idea of a holding pen. One of my major concerns stated above is the idea of having these bot-created articles in mainspace for users to see. If, however, we wanted to output the code to (a) subpage(s) of WP:NRHP or to someone's userspace, I would be less hesitant to oppose this. In fact, I would be emphatically supportive of this, so long as we have a strict process for "promotion" of the stubs. Some of you may know that I am already actually doing this with National Register of Historic Places listings in Lauderdale County, Mississippi. I have all the Elkman output at User:Dudemanfellabra/NRHP, which I eventually plan to expand using NRHP documents I obtained via email and other sources I have obtained (evidence on my to-do list). I can't really complain about this proposal since I'm already doing it myself haha..

Now... I think part of that strict process for promoting articles would be that if there is a nomination document available online for the site in question that all of the information (or at least a good chunk of it) be added to the bot-stub before it's put into mainspace. Before some of you go saying that's too high of a bar, think of all the NRHP listings that don't currently have documents online. That's thousands of stubs that with little editing after the bot run that could be promoted. The ones with documents or other big sources present could be left for those that like to expand articles more than they like to create new ones. This would basically weed out all the sites with little information and create those articles (which would probably take a long time to grow to greater-than-stub status anyway), and it would also highlight potential expansion areas for editors focused on that.--Dudemanfellabra (talk) 19:49, 15 December 2010 (UTC)

Thanks for focusing on what we can support and what we can agree on. As far as a "strict process" for promoting the articles off the holding page: we'd have to be careful - I don't think we can impose (with penalties) requirements that are stricter than for an editor creating a stub on his own. But what we can do is jawboning and to put a bit more muscle into it, let editors know that if they don't meet the consensus rules for promoting bot articles that the operator won't produce his requested stubs. "A good chunk of" above could be interpreted in many ways - how about something more specific like "200 words" or "2 full paragraphs"? Smallbones (talk) 20:22, 15 December 2010 (UTC)

A virtual holding pen for articles created by a bot can be created by using a variation on an template:Under Construction tag, like Dudemanfellabra set up for "NRHP infoboxes needing cleanup" or like i set up for template:NRHP dab needing cleanup (current size 0 articles in Category:NRHP dab needing cleanup), to put the new articles in a category whose size can be posted publicly. The size shouldn't be too big, or if it gets big temporarily it should be incumbent upon someone to be emptying it fast: yes, jawboning could/should then be applied. Avoid unnecessary steps of requiring page moves and so on. When u start an article, checking "what links here" only works if u r in mainspace, as a good tool to use to find bits of material already out there to copy in, or otherwise. While creating 10 or so Cass Gilbert related articles just now, i went back and forth to other articles that way. Often when u create an article at the NRHP official name, u soon find that there is a pre-existing article under an official NRHP alternative name or under some similar name; u would not find those as easily and u would waste time editing ur article if it is off in some userspace page. It saves steps if u r already out in mainspace, and u just simply ctrl-alt-a ctrl-c to grab ur content, then replace ur new article by a redirect to the older page, and then follow ur redirect with ur content in hand. The redirect would be needed anyhow. Also, a pen off to the side in someone's userspace is not available for disambiguation editors, category editors, stub-sorters, NRHP documentation format sticklers, and other specialists to get to work. The NRHP-under-construction tag could be set up to show a couple status levels, like that the article has not been checked at all, that it has been fixed for just tidying needed, that it has been fixed for NRHP nom doc tie-in or not.

I'm not sure what level of bar Dudemanfellabra is meaning to set, but I would agree that new articles in states where the NRHP docs should be available on-line, should immediately (within the virtual holding pen) get complete references for NRHP nom docs and for any MPS/MRA docs included into them, up front. A bot or my current semi-automated process or Elkman's generator can easily draft those references, but it takes a manual editor to find the author/preparer name and date and otherwise complete out the reference. While doing that the manual editor could notice an interesting fact and choose to add it to the article, or could try to provide a characterization of why the place is NRHP-listed, but that can only be encouraged and not required. I don't want to judge in any way controlling upon others, and don't want others judging me, on whether my characterization of why some farmhouse was listed on the NRHP is completely fully adequate or not. The point is to model something good, to get it started, so that the proverbial local high school kid who knows that place can relate to the info and see for her-/himself that she/he can add info from the NRHP doc or from elsewhere. --doncram (talk) 20:35, 15 December 2010 (UTC)

(Resonding to Doncram) I will not support in any way, shape, form, or fashion these new bot-articles being put into mainspace. If I'm going to support it, they're going to be put in user/project space. Yes, the what links here only works in mainspace, but to solve that the bot can create a redlink at the top of the user/project space bot-article. The editor can click on that link to find what links to the future article in mainspace. In fact, the redlink would actually make it easier to copy the material into the article when you're finished anyway because clicking a redlink automatically brings you to a page where you can just paste the material in.

(Responding to Smallbones) How about DYK size (1500 bytes of prose)? Of course there can be exceptions made for very small nomination documents (I have a few from Lauderdale County that even if I copied and pasted the material in wouldn't give me 1500 bytes). I think generally, though, nomination documents have at least that much information in them.

As to the jawboning, I think we should come up with a worse punishment than simply not allowing them to create more (although that could certainly be part of it). What if the editor isn't requesting new stubs but just ill-promoting all the current ones? Maybe if after repeated infractions against consensus (3+?) we block that editor from WP:NRHP and all its subpages (if that's possible) for successively longer times (1st offense=1 day, 2nd=3 days, 3rd=1 week, 4th=1 month, 5th=until drive is over).--Dudemanfellabra (talk) 20:57, 15 December 2010 (UTC)

Well, I doubt that that solution is possible. I wanted to be sure that folks know that we'd have very few tools at our command to impose rules on this. That's why I included Samson above. In the Bible he was caught without his weapons by the Philistines, but still managed to come up with an ad hoc solution. He found the jawbone of an ass at hand and took care of the problem that way. I'd be willing to play the role of the jawbone, if other folks played the other roles :-) Smallbones (talk) 21:13, 15 December 2010 (UTC)

You really really hate that R. C. Williams Warehouse existed in mainspace for 17 minutes before i completed out the author and date information in the NRHP nom doc that it referenced, don't you? Hmmmph, well, i think ur articles using off-line NRHP nom docs, where u don't provide author or prep date at all, deserve serious punishment... :( Not serious about punishing, as i'd rather see the articles, tho actually i do hope u will learn to give those credits. I don't think u or i should be enforcing some arbitrary standard. :) But seriously, what is your beef? For a new article, there are no readers other than newpage patrollers seeing them. Something offends you which i don't understand. --doncram (talk) 21:41, 15 December 2010 (UTC)

Obviously an edit conflict, and not directed toward me. I'll stop stupid jokes about jawboning, if you try to focus on what we agree on. Smallbones (talk) 21:49, 15 December 2010 (UTC)

Yes, my reply was towards Dudemanfellabra. I'll try. It occurs to me that Dudemanfellabra is speaking about what's appropriate for Mississippi articles, where the NRHP noms are mostly NOT online, and for Meridian where even the MPS/MRA ones are confusing. I'm speaking mostly from National Historic Landmarks anywhere and from New York and Connecticut and other states where NRHP docs are online. For Mississippi, it makes sense to focus on harvesting most of the NRHP doc info from an offline copy, when creating an article. Where NRHP docs are online, that is not so important. The reader is served by a short summary with some interesting bits of info, and can dip into the linked NRHP nom doc for more. Too complete a summary is not even appropriate, IMHO: the source should be bigger than what you take from it. And I would focus any geo-oriented article generation by bot or otherwise upon the states where NRHP nom docs are available (which is why i did drives on VA, CT, NY, and ND now). Does that help? --doncram (talk) 23:14, 15 December 2010 (UTC)

No, in fact I was talking about all NRHP listings. If you read my comment above, I specifically said a site where "a nomination document available online". The entire point of my comment (which you frequently miss..) was that if a document wasn't online, we could go ahead and create the stub, but when it is available online, we should extract information from it whie the bot-output is still in user/project space before we create the mainspace article. And I'm not talking about just filling out the cut and paste format for showing the document as a reference; I'm talking about actually going in and finding the information to put in the article. (I bold because this is the second or third time I've said it, yet you still fail to understand.) This way at least some decent quality articles would come out of this.

My beef with creating these in mainspace before being worked on by humans is that they will be there much longer than 17 minutes if we create thousands of articles. And besides, to create a quality article, it usually takes more than 17 minutes for the research anyway. Once again, in the RC Williams Warehouse article, there is a document online that you could extract a wealth of information from, yet you choose to continue to leave it in crappy stub format, as I'm sure you would choose to do with many of these new articles given the chance. If you have 10 tons of crap, it's still crap. Quantity is trumped by quality in every instance. If we are going to create a lot of new articles, I am not going to support it unless we focus on quality. You, on the other hand, appear to only be concerned with quantity. That, specifically, is my beef.--Dudemanfellabra (talk) 23:49, 15 December 2010 (UTC)

I find myself mostly in agreement with Dudemanfellabra here. I have yet to create my first NRHP article (I mainly do photos), but the ones I'm considering are redlinks in the Utah and Idaho county lists. When I get around to it, I intend to create them at least at a Start level of quality (and submit them to DYK). Redlinks get my attention; if you stub them, I may not get around to them for many years. Ntsimp (talk) 00:04, 16 December 2010 (UTC)

Maybe creating 1/4 or 1/2 the articles in a given area is a good target, then? Have fun with the DYK level approach, that is fine to do.

(@dudemanfellabra) Hmm, there you go with the insulting language again. But thanks for another unmeant compliment, that i unearthed a good topic with that RC Williams Warehouse article. U and other readers can see that it is interesting. Thanks for admiring it!

Thanks also for reiterating about your views. It didn't and still doesn't make sense to me, grammar aside, that you say "if there is a nomination document available online for the site in question that all of the information (or at least a good chunk of it) be added to the bot-stub before it's put into mainspace". I see u mean if there is info available online, the article must include it all. But why??? The info is available. The interested reader can go and read it. We add real value by making intros available and making all kinds of neat links between articles and topics. We don't have to regurgitate everything that is written nicely in a report about one place, when we can just direct a reader to it.

One other difference in perspective is that you seem to want to craft a few masterpiece articles, when i and some others are working to make interesting list-articles on NRHPs in geo areas or on topics like Elks buildings or List of governors' mansions in the United States. I notice that what i think is your primary list-article, List of RHPs in Meridian, has just 4 descriptions, all written by me i think as an implied suggestion for u, for its 46 listed NRHP properties, and 16 of those are redlinks. I've done far better on lots of list-articles, starting first by creating all the component articles. To get material for descriptions for an interesting list-article, u don't need as much at the individual articles. I'm not going to be angry at you that u haven't developed that one list-article itself, because u certainly don't have to, but likewise it's not justified for u to be angry that i didn't do something u wanted done. We have all not done lots of stuff. I and some others happen to like getting coverage out there, along lines that Smallbones comments about above, which will facilitate getting lots more people involved productively. --doncram (talk) 00:39, 16 December 2010 (UTC)

(outdent) You said: "We don't have to regurgitate everything that is written nicely in a report about one place, when we can just direct a reader to it." –– so why don't we just strip everything out of the Statue of Liberty article and leave a list of links to the references? I'm pretty sure that would agitate a lot of people. The entire point of Wikipedia is to bring all the information from various sources together into a ready-to-read-once format. If a reader gets to an article and just sees a list of links with information, Wikipedia may as well be renamed Google.

We don't add value by putting a link or two on a page; we add value by turning those sources into readily-available information for the average reader to digest. Much the same, a list of sites on what is to most some obscure register commissioned by the government is pointless if all the sites on the list have 2 sentences written about them. The reason I have focused so little on the Lauderdale County list article is that personally I don't really care about it. I've focused on bringing the Meridian, Mississippi article (which mentions the places listed in the context of the history of the city... far more valuable in my opinion than a list) from hardly start quality to now being a GA. When I get the time, I hope to bring it all the way to FA, but that is beside the point. Anyone (including a robot) can make all the redlinks on a list turn blue, and anyone (including a bot) can write two sentences and link to a pdf about a place, but it takes someone with the will to actually discover the history of a place to write a decent article.

The bottom line is that the entire purpose of Wikipedia is to be a self-contained encyclopedia of basically all knowledge – NOT a link directory or a stub farm or a disambiguation free-for-all. Articles are the backbone of Wikipedia. All the other stuff falls into place because of those articles. If we are going to have a bot create anything, it should be articles. And that.. is all I have to say about that.--Dudemanfellabra (talk) 02:36, 16 December 2010 (UTC)

My take on the bot issue: Bots are a great tool for collecting data and compiling facts... ie doing some of the background research for an article. But we should not rely on them to actually write the article. A Human should do that. Yes, it takes a bit longer to do it that way... but we are not in a rush. Blueboar (talk) 15:07, 19 December 2010 (UTC)

2nd break

Again I'll ask everybody to state what kind of bot they can support. Given the no-bot alternative of having on-average 27,000 sites a year (or more) without an article for 20 years (or more), I'm sure there's something we can agree on. So please just state your "acceptable" bot - alternative, and we'll see if we can do it technically. There's no reason that we can't pull the plug if the bot causes problems, and there's no reason that the bot can't be improved or speeded up if a consensus agrees to it later. A single step would be an improvement. For now it looks like the "holding pen" vs the "virtual holding pen" is the main hangup (and they are perhaps virtually identical?). I also possibly detect a bit of a personality conflict. One thing that I have always liked about this project is that everybody seems to do their own favorite tasks and generally add to the whole project, and very seldom do editors insist that things have to be done one way. And compared to other places on Wikipedia there is very little bickering. I'm sure there is a bot-assisted solution here that will not mortally offend anybody. I'm afraid building in penalties for over-promoting bot generated articles would mortally offend many people's Wiki-sensibilities - even if there was a way to do it under the rules, but WP:AGF and trust in your fellow editors will go a long way. Smallbones (talk) 01:15, 16 December 2010 (UTC)

Thanks. Basically i want a system that provides a lot more convenience for a focused article drive, which would generate a conveniently accessible, drafted starter articles, say for all NRHP places designed by one architect, or for all redlinked NRHP places in one county-list-article. If the focused article drives work out well, then the scope could be expanded. I was thinking that a bot might be able to both generate drafts and place them into articles in mainspace (and give an index to those articles for editors to go visit), but now I don't see how a real bot could be instructed to do all that. I'm stuck on what would be technically feasible, given some experience now from my trying to generate articles.

The NRIS database is a little bit complicated, consisting of multiple files that need to be drawn from and merged mostly by use of the key field "refnum". But after some careful merging of these files, a programmer should be able to generate a long report, sorted by state then county then NRHP name ("RESNAME") which is a big sequence of drafted articles (infobox, then main text body, then categories and stub tags). I skip some complications in that. Then at some point it would then not be hard to customize the output so that HD articles in CT, say, get category [[Historic districts in COUNTY County, Connecticut]], which is right for that state. And there are various other customizations to do that way. The main improvement is to have article text drafted, with useful simple sentences like "The district includes 64 [[contributing buildings]] and 12 non-contributing buildings" and "It was designed and/or built by ARCHITECT" and "It includes [[Greek Revival architecture|Greek Revival]] and [[Late Victorian architecture|Late Victorian]] architecture" and "<ref name="nrhpinv3">{{cite web|url=http://pdfhost.focus.nps.gov/docs/NRHP/Text/ REFNUM .pdf |title=NRHP Inventory-Nomination: RESNAME |author=___ |date=, 19 |publisher=[[National Park Service]]}} and [http://pdfhost.focus.nps.gov/docs/NRHP/Photos/ RESNAME .pdf ''Accompanying ____ photos, exterior and interior, from 19___ (see photo captions page __ of text document)'']</ref> and so on, that an editor can keep or delete easily, rather than having to type it all in.

That output, one big file, or just selected parts of it, could be posted to a Wikiproject workpage that editors could cut-and-paste individual articles from, instead of cutting-and-pasting from one-article-at-a-time reports on Elkman's website. Or the just the output for one architect name could be generated and posted. I can see how to do all of that. Editors' manually cutting-and-pasting to fill the redlinks in a given county-list-article which has disambiguated article names prepared (as all our county list articles do now) avoids the need to get a bot to interpret the list-article and do the matchups to the correct non-disambiguated RESNAMES and REFNUMS in NRIS. I don't see how a bot can be told to do that. If a bot can't do that, maybe technically we wouldn't need a bot at all.

Or, should we think that we can get a bot to create articles for the redlinks in a NRHP list-article, and put them out there (within a virtual holding pen limit), working from one list-article at a time? Actually Kumioko indicated further above that a bot could keep its output limited by using a category's count. The trick would be for the bot to grab the right drafted article from a big output list of articles. Should we assume that can be done? --doncram (talk) 03:17, 16 December 2010 (UTC)

I would be willing to support a holding pen if it's in userspace or WikiProject space. Someone would have to request an article be created by the bot, which means it would need to be run manually each time. The person to publish the article to mainspace would be required to provide several other reliable sources so that it would clearly pass the general notability standards on its own, not the "free pass" that's usually given to NRHP articles. Ideally at the level of 1 and 2 from this month. NOT at stub level, minimum of start level or C level. Royal broil 04:15, 16 December 2010 (UTC)

I would support the bot if the holding pen was in user or project space. I'm fine with creating many bot-articles at once (even all 27K.. but not in mainspace of course). I think in cases where nomination forms are available online, the minimum required length (with an exception for short forms) should be DYK-level, and for cases where there are no nomination forms, there should at least be a token effort (i.e. a thorough series of Google searches) to find other reliable sources to add to the stub. This token effort clause should also apply to short nom form NRHP listings. If after checking for nom forms and a thorough search of Google, there is still little to no more information than the bot output, I'd be fine with starting the stub. If, however, there is some information out there and the editor in question refuses to incorporate it, the article shouldn't be moved to mainspace until enough information is incorporated to bring it to DYK level.--Dudemanfellabra (talk) 04:37, 16 December 2010 (UTC)

Responding to both: Wow, that's harsh. Would u apply this ban to all topic areas of wikipedia, that everything must be written off-line, somewhere else, and should not be allowed into mainspace until DYK length? This new Wisconsin community article would not be allowed. I am dismayed to note that you would have disallowed the National Historic Landmark article drive which created 2,400 articles. I am dismayed to note that it would disallow the nice Maryland and New York articles that Pubdog has created over the last year. I am dismayed to note that your rules would not allow for most NRHP list-articles, which are articles themselves, to exist, because they do not meet DYK criteria. Would you seek to ban new users who started an article? Are u allowed to edit in existing short articles, or not unless it brings the article over DYK length? I wonder, would you allow links from mainspace to the offline system, so that editors/readers could be aware of the articles in progress offline? I don't think this would fly in a wider discussion.

I will grant that you two seem to walk that talk mostly. I browsed both of your early contributions in Wikipedia, and notice a lot of additions to existing articles on NASCAR drivers or whatever, and very few new articles. Besides recent short Wisconsin community articles, i don't happen to see any new articles by you that are not DYK. It seems troubling to me, though, that a ban on new short articles would allow editors to add to stub articles about fantasy game characters and sports figures and actors and minor legislators and so on, but not allow for people to add info to a local historical site article, because those weren't started before the window on new articles would be slammed shut. It would seem sad that only a few could create articles; this would not be the Wikipedia that everyone can edit. --doncram (talk) 14:22, 16 December 2010 (UTC)

If a tool is created to make it easier to crank out low grade article (it's doubtful I will ever ethusiastically support such a tool, but I probably be able to live with it), there should be more pressure to expand them to somewhat higher quality. I'm still reeling from the mass creation of 2-sentence stubs in Washington County, PA. What I'd support is an off-wiki tool (on Toolserver maybe?), similar to the infobox generator, but, preferably, with a password such that the metaphorical keys could be given to only editors who have earned the trust so that what happened in Washington County is less likely to happen elsewhere (and so the regulars here don't have look over the shoulder of the people wanting to create stubs and monitor this "holding pen"). You generally don't give admin tools to newbies right away, nor access to AWB, no reason to do so here. Niagara ^{Don't give up the ship} 15:01, 16 December 2010 (UTC)

Still reeling? That's an allusion to User:GrapedApe, a good editor with DYKs and good writing, finding his way to start NRHP articles in Washington County, and encountering really rude behavior. He was successfully driven away: "However, I'll probably steer clear of other NRHP articles, because I'd rather not be harassed by some self-appointed arbiters of protocol at the NRHP project." I am ashamed on behalf of WikiProject NRHP. --doncram (talk) 15:40, 16 December 2010 (UTC)

That's no reason to deride me because I think we should be encouraging to spend some time on an article (I believe I also offered to provide the nom forms, but said editor doesn't appear to have wanted to make the effort in expansion). Take a lot at a similar situation with North Hill Historic District (New Castle, Pennsylvania); a new editor spent a little more time preparing the article, I popped in, did some cleanup, provided the link the nom form and other assitance/encouragement. (see here and here). But this is not about me (argumentum ad hominem), it's about making easier to create sub-standard articles. Niagara ^{Don't give up the ship} 16:06, 16 December 2010 (UTC)

Sorry for a misunderstanding. I was avoiding stating that it was Dudemanfellabra's rudeness back then which i specifically found abhorrent, while i thot Niagara was being challenging in a good way. I was friendly to D early on and am sad to see that kind of rudeness applied to me here and to that new editor in the NRHP area. (Some more of what GrapedApe said: "Frankly, I'm surprised that creating stub articles on obviously note-worthy subjects can arouse such negativity, like this rude comment. I fail to see how creating those stubs--with a reference, a full infobox, and relevant categories--can possibly be a bad thing. Featured articles don't spring fully-formed from the mind of contributors: they must start somewhere. Thanks for trying to explain what the dispute is all about, rather than just being nasty." Full discussion displays still at User talk:GrapedApe#Washington County NRHP stubs.) I did then try to make nice, and to support Niagara's request that more be added to the articles. In part i did so by myself trying to use the difficult Pennsylvania document system. The situation there and everywhere is like a gigantic entrapment, designed to turn off new editors. We have 55,000 redlinks, and then nasty commentators if a person steps in and does merely what a bot could do, rather than creating a full-fledged FA article. Why not just run the bot, and post friendly hints on how to navigate the difficult systems and to add to articles, at Talk pages. Create a situation that is welcoming and discuss what we positively want. Remove the entrapment by just creating the short articles. --doncram (talk) 16:23, 16 December 2010 (UTC)

Despite some obviously long standing disagreements, I'd think we're almost agreed here on some minimal use of the bot. I'll note that nothing we say about bot-assisted editing can apply to editors who don't use the bot generated material. If they want to create stubs the old-fashioned way, they may according to Wikipedia rules that have been in force forever. I think the point of agreement can be along Dudemanfellabra's lines.

for sites without on-line NRHP forms, or with very short forms, the bot can be used, after first dumping the results onto a user page or project sub-page, and requiring e.g. a half hour's work, searching for other sources, re-writing material, etc. the stubs can be posted.
for sites with on-line NRHP forms, the use of the bot is discouraged, but if the editor can get the material up to Start level it (should be) ok with the project.
"enforcement" of the above has to be by community monitoring and jawboning (I'll certainly do this and ask others to as well), and refusing to provide new bot-generated stubs to the "offender."

This is not as strict as RoyalBroil would like, and not as much as Doncram wants, but (call me optimistic) I don't think it should "mortally offend" anybody. It also opens up the use of the bot for 1,000s of articles (10,000s?) where we don't have much chance to come up with a solid article, but can at least get the basic facts out. We should schedule a review, say after 6 months or 1,000 bot assisted articles, and check out the technology with some trial runs and complete the bot approval process.

Is everybody at least ok with that? Smallbones (talk) 16:35, 16 December 2010 (UTC)

Thanks for mediating. But, I want to run the bot to finish Connecticut and for North Dakota and other places where the NRHP documents are available on-line, first, exactly because the bot-generated articles are immediately pretty good quality, with a good link to more information. And, there's no reason to do the development in a user or project page, when it is faster/better to do in mainspace (within some virtual holding pen), so while logistically i think the output is best posted to a work page, the best first manual editor step is to cut-and-paste into a mainspace article. I think that RoyalBroil and others don't understand the complications (and i am pretty sure i don't fully see them all either). Maybe best to do a proper demo / pilot for one batch, slowly, for how this can work, before hardening upon specific restrictions. --doncram (talk) 16:48, 16 December 2010 (UTC)

While not ideal, I am okay with these terms. Perhaps my word choice in my comments were more strict than what I was thinking. The main thing that I want to see is some work, not just a posting of the bot results for 55k articles. Start level is fine with me as I have indicated. I couldn't find any recent article at that level to cite in my comment. I meant a cursory "history" section with a small paragraph or something like that. My hope is that this bot result would be used as an excellent starting point. If I were writing the terms, I wouldn't make distinction between forms available online or not. I suspect that I'm missing something on that point and I would appreciate comments on why there's a difference. Royal broil 13:26, 17 December 2010 (UTC)

The distinction between sites where online NRHP documents are available and where they aren't comes down to - without online documents they're not going to be start class with or likely without a bot. Doncram wants to do a trial first without this distinction and without going for start class so we can see what the output looks like. How about for North Dakota, where there are about 300 red links. I think it is fair to at least look at what the result of a trial run is. At that point we'd at least know what we're talking about before setting restrictions. I'd even personally clean up the articles to a "good stub" class, within a few months, if folks really objected to the overall quality. Any objections? Smallbones (talk) 15:56, 17 December 2010 (UTC)

For a small pilot test, see Talk:National Register of Historic Places listings in Tolland County, Connecticut/drafts. The current version provides draft text for two Tolland County, Connecticut articles which are currently redlinks. Feel free, anyone, to start either of those two articles and their Talk pages, and to develop using the NRHP nom docs that you should be able to access within the articles. To access the NRHP nom docs, after you copy-paste the draft you will need to do some tidying-type cleanup within the NRHP nom doc reference (remove spaces and line-breaks and change ** to '' as in this edit for another Tolland county article). --doncram (talk) 19:38, 17 December 2010 (UTC)

I could support the idea of the holding pen(s) being in sub-talk space like you've done above, Doncram. I don't like the fact that you've gone ahead and create many of these articles in that list before consulting with the rest of the project, but meh.. what was I to expect?

I like the output of the bot, but I certainly wouldn't want an article started in the condition of the current output. Why don't we pick a county in Connecticut or North Dakota where all the sites have nomination forms online and most are redlinks... and also choose a mostly redlinked county in a state where none of the documents are online... and use both of those counties for our test subjects? The CT or ND articles should be expanded to DYK/start/reasonable level before being moved into mainspace, and the no form articles should just be tidied up a little and searched on Google. If these two counties go off relatively smoothly, I'd support expanding the measure to creating a /drafts page on all of the county list articles and going from there. Maybe we could set up a project page also, on which we transclude all of the /drafts pages, organized by state, so we can see all of them at the same time. Checking the "Related Changes" to that project page, then, would alert us to any article creation and allow us to monitor it.

Alternatively (and I believe this would be better), all of the output could be sent to the project page instead of creating all the /drafts pages, so just putting that project page on your watchlist could allow you to patrol it. I could support either option, with preference for the latter, but that isn't a green light to go creating hundreds of /drafts pages now, Doncram. Let's wait to see what others think.--Dudemanfellabra (talk) 21:07, 17 December 2010 (UTC)

Glad u like that output. It got better as i tested out editing in some generated articles from earlier drafts and then tried different improvements, responding to the quirky problems for different articles. I don't understand a lot of the rest of what you say, sorry. If u r suggesting putting generated articles for all of one state in a single project page, that wouldn't work, as the page would be too large. A page for one county list-article seems to be about right, and then putting it with that seems to make sense. I'm just gonna proceed with pilot testing. --doncram (talk) 21:50, 18 December 2010 (UTC)

@Dudemanfellabra, I was not previously aware of this "Related Changes" feature in wikipedia. Given that it exists, e.g. this Recent Changes report for Grand Forks ND county list-article, then what? I am missing what then to do with it. I guess if someone was concerned about potential abuse of a specific county list-article's generated /drafts page, they could every few days monitor whether anything had been done with the drafts, by regularly checking there. In some kind of enforcement way, to check if someone/anyone had actually started more than the allowable number of drafts, without bringing them up to "pretty good stub" quality. Is that where u were going? Honestly at this point i am just trying to understand what the Related Changes feature can be used for. --doncram (talk) 04:27, 20 December 2010 (UTC)

P.S. If u want to use that Related Changes feature to monitor the Grand Forks articles, it would be better to apply it to the corresponding /draftnames page, which i haven't mentioned. That's a stripped-down list of the Grand Forks article names only, or of valid redirects to the articles. I needed to create such a corresponding list, which just shows pairings of refnums to wikipedia article names (or valid redirects), as part of generating the /drafts page. For use with Related Changes, that has advantage of stripping out lots of other linked terms that appear in the NRHP list-article. Here is Recent Changes applied to /draftnames for Grand Forks County. For my /drafts programming purposes, I can accept redirects on this /draftnames page, but if the "Recent Changes" feature is seriously useful for something, and if redirects undermine that, I could revise the /draftnames page to just show actual current article names. But again i don't yet grok to whom / how / why / where when Recent Changes can be useful. --doncram (talk) 04:37, 20 December 2010 (UTC)

When I suggested the Related Changes feature, I actually had it backward. I was under the impression that Related Changes shows edits to pages that link to a given page, but it turns out that it shows edits to pages that are linked from that page. Yes, I meant it as a place to monitor activity for article creation, so we could have a good idea of what's going on and if people are following the "rules"... I like the /draftnames Related Changes, but I'd like to see every single site in a state on one page (e.g. one for Mississippi particularly) that we could track. If that proves to be troublesome, ideally we could transclude all 50 states' /draftnames pages into one projectspace article.

Also, is there any way to only run the bot on redlinked sites? This way, the /draftnames Related Changes would only pick up new article creation and the like, instead of showing the edits for the already created pages as well.--Dudemanfellabra (talk) 17:17, 20 December 2010 (UTC)

A more positive use for the Related Changes feature, is to let editors working on an article drive together, to easily see what the others are doing. Like here it shows me that Elkman has actually stepped in to improve a few articles, both those on my own Watchlist and not on my Watchlist, and it conveniently shows ALL the edits, not just the last one. (While other commentators have done...nada!) The feature seems helpful in supporting any collaborative article drive. Also i realize that refining the /draftnames list to show the actual article names, and not to allow redirects, has other benefit (to improve the "What links here" links put above each draft article in /drafts ). So, i like this feature, and will manually include it into the introductory top stuff at the top of the /drafts article for Grand Forks county, and automatically into future batches.

Also, a template for new articles created as part of a specific article drive using drafts, or generally by any batch-generated drafts, can be included into the /drafts themselves (or manually added). It can put new articles into a hidden category that can also be monitored. Note, though, that another good use of the /drafts is to improve already-started articles. The article drive should be about improving them all to include at least the info in the generated drafts. I'm gonna run the batch-generator soon to generate /drafts for Scott County, Iowa, where most of the articles have already been created, but very minimally, to support article improvement there.

About refining what the Related Changes reports, I don't have means when running my batch-generator outside wiki, to know which are bluelinks vs. redlinks. That's an open question for me, whether an in-wiki bot could know which are which. (Kumioko, can you comment?) But, the /draftnames list could be split or edited down, to drop already-created articles. When starting a new batch, splitting the /draftnames into those which initially are bluelinks vs. initially are redlinks could be done manually, easily enough, if it's worth supporting two separate Related Changes reports. The single Related changes report for Grand Forks county /draftnames, set to show the last 500 edits or last 30 days of edits, currently clearly shows which are new articles. Seems like just one Related Changes report would be more than what any article drive has had for support, and would serve pretty well.

Or maybe there is some useful way to use the "Tags" qualifier part of the Related Changes query? --doncram (talk) 19:27, 20 December 2010 (UTC)

Grand Forks County in North Dakota is currently mostly redlinks, and it includes ND's biggest university. Here are draft articles: Talk:National Register of Historic Places listings in Grand Forks County, North Dakota/drafts, in pretty good shape, if i do say so myself, for what a program can generate. :)

In my batch-generator program, one spin-off is a report of MPS's. For Grand Forks County, it reports there are 35 listings in the "Downtown Grand Forks MRA", 4 listings in the "Historic Roadway Bridges of North Dakota MPS", and one each in "Buechner and Orth Courthouses in North Dakota TR" and "North Dakota Round Barns TR". I tinkered with the generator so that it completes out MPS doc references for the first two. --doncram (talk) 21:50, 18 December 2010 (UTC)

I reviewed the North Dakota drafts. I do not support placing hundreds or thousands of sub-stub articles like this into mainspace. I realize this is probably the best that a bot can do. I have a strong background in computer programming and database manipulation (I minored in computer science). This output is at a low stub level and the word choice is too generic and clunky (by design and necessary) to be helpful to many readers. I anticipated output at this level which is why I opposed since the beginning. I'm fine with everyone using something like this as a basis for starting an article since they would almost certainly customize the wording to topic. But the work needs to be up front because too many people have good intentions but end up quitting Wikipedia before they finish because they get overwhelmed or life takes them in another direction. There's no way to keep track by locale of what article are minor stubs like this versus solid Good Articles. That's why I'd prefer to see the redlinks kept until content at least at start class is posted. Like I said earlier, I think contributors need to request each article's output when they want to start on the article. Royal broil 05:57, 19 December 2010 (UTC)

I'm in agreement with Royalbroil here. Again, I'm going to have to object to using a bot that attempts to take the database classification information from the National Register Information System and attempts to turn it into an article. Sure, you can take database tables and machine-generate prose from them, but there simply isn't enough information in the database to give the reader context as far as why a site is historic, or even why the article is worth reading. Here's what your program generated for R.S. Blome Granitoid Pavement in Grand Forks:

Blome, R. S., Granitoid Pavement in Grand Forks is a property in [[Grand Forks, 
North Dakota]] that was listed on the [[National Register of Historic Places]] in 1991.

It was built or has other significance in 1911.<ref name=nris/>

It was designed and/or built by Blome,R.S.,Co. of Chicago.<ref name=nris/>

It includes [[  architecture| ]] and [[  architecture| ]] architecture style.<ref name=nris/>

When listed the property included one [[contributing structure]].<ref name=nris/>

The listing is for an area of {{convert|55 |acre}}.<ref name=nris/>

The listing is described in its NRHP nomination document.<ref name="nrhpinv3">{{cite web 
|url=http://pdfhost.focus.nps.gov/docs/NRHP/Text/91001583.pdf
|title=NRHP Inventory-Nomination: Blome, R. S., Granitoid Pavement in Grand Forks
|author=___ |date=, 19 |publisher=[[National Park Service]]}} and 
[http://pdfhost.focus.nps.gov/docs/NRHP/Photos/91001583.pdf
‘‘Accompanying ____ photos, exterior and interior, from 19___ (see photo captions 
page __ of text document)’‘]</ref>

Doesn't tell you much, does it? Now, upon reading the actual article, I found that it was an innovative pavement type from the early 20th century. It's interesting to read about how the R.S. Blome company developed a pavement that both provided good traction for horses and a hard, resilient surface for cars. It sounds like it was a really involved process, though, which is probably why most city road departments just dump a mix of asphalt and aggregate out of a truck, roll it a few times to make it flat, and then wait for potholes to form the next spring. But I digress. The point is that machine-generated text from database queries can never, EVER provide any amount of context that a human editor needs to include. I'm going to keep objecting to the use of a bot and to the thin stubs that show no use of actual research material. --Elkman ^(Elkspeak) 07:22, 19 December 2010 (UTC)

I agree with Elkman and others. Using bots to create all missing NRHP articles in mainspace based on the NRIS database is a hubristic mistake based on overly optimistic assumptions about the quality of that data and the current level of bot coding ability. Using that data and either Elkman's tool or a newly coded solution to create a starting point for responsible editors is fine. There's a reason that machine written books are not something you see every day. The holding pen, requesting the articles one by one, etc are all reasonable methods to ensure every article is edited and improved by a human before going to mainspace. I don't believe all NRHP articles need to be brought to DYK quality before they are created in mainspace, but it's a reasonable goal to aim for. As for how long it will take to create all NRHP articles, it's irrelevant and its a moving target. dm (talk) 08:45, 19 December 2010 (UTC)

I concur. No way will a skinjob know enough to be permitted to create stubs in mainspace at this point in Wikipedia's evolution. Abductive (reasoning) 12:39, 19 December 2010 (UTC)

Now several of you are criticizing the wrong thing. The set of generated drafts at the Grand Forks County /drafts page are not written to be put by a bot into individual mainspace articles. They are drafted to be better starting points for creating new articles than what the Elkman generator provides, and to require manual editing by the person posting to mainspace. If I wrote a bot request (which this is not), I would design them to be shorter and avoid obvious difficulties.

Note, these drafts are better than Elkman's in several ways:

(e1) They avoid a small factual error in Elkman's that Elkman has so far chosen not to correct, that the NRIS database uses .9 acre size to indicate "less than 1 acre" not .9 acre literally.

Elkman has now revised his NRHP infobox/generator to fix what displays for .9 acre cases, thanks! --doncram (talk) 03:57, 22 December 2010 (UTC)

(e2) They include a templated version of the NRIS reference which fully responds to a now-long-running complaint in discussion thread also here at wt:NRHP, which Elkman has so far chosen not to include.

Elkman has now revised his NRHP infobox/generator to provide the templated version of NRIS reference, thanks! --doncram (talk) 03:57, 22 December 2010 (UTC)

(e3) They ensure that good, available NRHP document references will be included in articles.

(e4) They include some categories that Elkman's does not, such as handling Category:Greek Revival architecture and Category:Early Commercial architecture which are now well-established categories in Wikipedia.

(e5) As RoyalBroil notes, they provide, in choppy draft writing, some useful pieces of text that a writer would edit down in writing a coherent starter stub.

(e6) They provide a useful caution to writers, that the "built=" field in the NRHP infobox does not necessarily provide a built date and the "architect=" field does not necessarily provide an architect. They provide accurately ambiguous statements such as "It was built or has other significance in 1911." and "It was designed and/or built by Blome,R.S.,Co. of Chicago."

(e7) The generator provides qualification of significant dates, by "c. 1911" when a CIRCA field in the NRIS database indicates it should (which i think but am not sure the Elkman generator does not provide).

(e8) It reports multiple significant dates where relevant, e.g. "built = 1879 ; C 1881 ; C 1900 ; : " for the Thomas D. Campbell House, while Elkman's output just reports the first date. This will prompt the editor to write text explaining the significance dates.

Also, they are inferior in other ways: They do not yet include coordinates. They omit some categories and the TrainsWiki wikiproject banner that Elkman's output would provide where relevant. They do not include function and subfunction information that Elkman's left-hand-column displays as reference for editors. Probably they are inferior in ways i haven't yet noticed. In revisions to the generator program, i hope to improve its output for new batches and then also to use it to update old batches.

Also, note the generated texts are better than existing mainspace articles in some ways. For example, they include good NRHP document references which most North Dakota mainspace NRHP articles do not include.

The Grand Forks output provide for a pilot test so far. Is anyone willing to try starting a Grand Forks County article, and see how this works, and provide feedback towards improving the generated drafts? --doncram (talk) 14:54, 19 December 2010 (UTC)

P.S. I've drafted a text to include in /drafts pages, at template:NRHP generator pretty-good request. The first drafted text there was:

Please do not start a new article using one of these semi-automatically generated drafts unless you will revise the article to meet a "pretty good quality" standard within a short time. What that standard needs to be is under discussion (as of 12/2010, the pretty-good-quality-standard requested is being discussed at here), but any article brought up to "Start" article quality would certainly be okay. Please do not start more than, say, five, if you have not brought all previous ones up to that quality level.

The current text box displayed is as follows (which will update):

Please do not start a new article using one of these semi-automatically generated drafts unless you will revise the article to meet a "pretty good quality" standard within a short time. What that standard needs to be is up for discussion (as of 12/2010, it's being discussed here), but any article brought up to "Start" article quality would certainly be okay. Please do not start more than, say, five, if you have not brought all previous ones up to that quality level.

Pls. discuss wording at Template talk:NRHP generator pretty-good request. --doncram (talk) 16:47, 23 December 2010 (UTC)

Also, the /drafts pages so far generated and posted in Talk subspace can be found by What Links To the "NRHP generator pretty-good request" template. --doncram (talk) 16:06, 19 December 2010 (UTC)

Oh, wow, I didn't think we were in agreement! Royal broil 20:17, 19 December 2010 (UTC)

Are we? I dunno. If u want to argue, though, i can argue hypothetically for a real bot to create stubs for all NRHP articles. But, i don't see how that could be done, logistically, like in terms of explaining to an on-wiki bot programmer how to do all the merges and composing from NRIS that Elkman and i have been doing completely off-wiki (so what we are doing is generating draft articles, not running a bot). And, I am pretty sure that going through some batches of off-wiki generated articles will lead to improvements in what off-wiki generations of articles can achieve. So i want to do a few focused article drives, first, but eventually that could possibly lead to a bot using these to place into mainspace. But if u really want to argue, now, about what a future combo with a bot could possibly ever do or not, i could oblige, for sake of argument. :) Also, there's lots of room to argue about what is or is not a "pretty good stub". :) --doncram (talk) 04:11, 20 December 2010 (UTC)

3rd break

As Blueboar noted above, technically my current process of generating batches of draft articles may require a Bot approval, although it is not using a bot. So i opened Wikipedia:Bots/Requests for approval/NRHP batch generator 3.0. One question opened there is whether batches of drafts should be placed at Talk subpage of a NRHP county-list-article, or placed in userspace or WikiProject NRHP project space. Please comment there if interested. --doncram (talk) 16:36, 23 December 2010 (UTC)

Also, FYI, the Tolland County, Connecticut /drafts assisted in completing out stub or starter articles for all its former redlinks. Recent changes in Tolland County NRHP articles shows last 30 days activity. Also, editor Pubdog has elsewhere provided some feedback that I'll address in any new batches. And, Scott County, Iowa editors are contemplating, if not actually using, the /drafts i provided there towards helping them improve their mostly-already-existing articles. --doncram (talk) 16:47, 23 December 2010 (UTC)

NRIS coords conversion from UTM

I am trying to add coordinates information to the output of the NRHP batches-of-articles-generator (BOAG?) program that i am developing. In the new NRIS database the UTM file, and in the old NRIS the UTMZONEM.DBF file, includes coordinates data in UTM format, i.e. with zone and easting and northing fields. These can be converted by complicated formulas to latitude and longitude. Websites will do it for one location at a time. Or there is Steven Dutch, Natural and Applied Sciences, University of Wisconsin - Green Bay's website which provides a spreadsheet to do conversion in a batch. In my software i could implement those formulas.

(A drawback is that it is assuming the UTM data is recorded under the 1983 coordinates system (NAD83 datum), while in fact most of it is under 1927 coordinates system (NAD27). I suspect this might implement what Elkman's system does. BTW, when we write a new infobox with coordinates, i think there oughta be included a coded comment that would be bot-readable, to allow for future improvements of coordinates by bots. Something like  to indicate coordinates being sourced from the 2010a version of NRIS database's UTM.DBF file, interpreted as if they are NAD83 data (even though many/most are NAD27). That might allow a later bot to come by and replace the coordinates, if a NAD27->NAD83 converter is available, for example. Or it might help an editor interested in fixing coords, to find their way to wp:NRHPcoords and decipher what has been done. I dunno.)

Is that the way to go? Or is there some other way to get lat & long coords for NRHP-listed places? --doncram (talk) 20:48, 23 December 2010 (UTC)

Merry Christmas to all!

I hope everybody gets to enjoy the holidays with family and friends. Yes, I know I have a backlog of North Dakota articles to write, but they may not get done until after a Happy New Year. The best to all! Smallbones (talk) 21:05, 23 December 2010 (UTC)

I also wish you all a Merry Christmas and Happy New Year! Einbierbitte (talk) 22:18, 23 December 2010 (UTC)

listed vs. pending vs. removed vs. other

In the March 13, 2009 version of NRIS, counts of NRHP listings by certification status code are:

LI which means LISTED in the National Register : 84326
DR which means DATE RECEIVED/PENDING NOMINATION : 2466
RN which means REMOVED from National Register : 1560
DO which means DETERMINED ELIGIBLE/OWNER OBJECTION : 1051
DP which means DATE RECEIVED/PENDING OWNER OBJECTION : 124
RE which means REMOVED FROM ELIGIBLE LIST : 11
BD which means BOUNDARY DECREASE : 5
DD which means DETERMINED ELIGIBLE/DOE PROCESS : 2
DC which means DETERMINED ELIGIBLE/CERTIFIED DISTRICT : 1
NX which means NATIONAL LANDMARK STATUS REMOVED : 1
PM which means PROPERTY MOVED : 1

and there's one empty entry.

It's come up several times that items where code is DR were actually later fully NRHP-listed.

All the ones identified as RN are well-documented on the internet, erroneously, as being listed on the very date that they were in fact delisted. I think creating really short stub articles on these ones, by a real bot or otherwise, to get the accurate information out, would be helpful. --doncram (talk) 20:41, 27 December 2010 (UTC)

I'd like to know which one was listed as NX. That'd be a nice example to link to in the documentation of the NRHP infobox when it talks about sites that have been delisted from a NPS designation but not entirely delisted from the NRHP.

As to the delisted (RN) ones, many of these articles have been created and possibly will be eventually created. There's no reason to bot-start these – especially because they're no longer actually on the register. That takes the NRHP inherent notability argument out of the question if you ask me. This is good information to have, but I think we should just keep manually creating articles.--Dudemanfellabra (talk) 05:39, 28 December 2010 (UTC)

Btw, if the above numbers are true, then according to this template:

We have created 30,827 articles (not including redirects, dabs, lists, etc). Subtracting 21 delisted properties, it turns out that we have articles for 36.58% of current NRHP listings.
Of the 30,827 articles, 24,593 (79.78%) are stubs.
Only 6,234 (20.22%) of our created articles are Start-class or better.
That means that of the total 84,326 NRHP listings, only 7.39% have a pretty decent (better than stub) amount of information about them on Wikipedia.

Kind of disheartening, huh? haha.--Dudemanfellabra (talk) 06:21, 28 December 2010 (UTC)

I recall some de-NHL'd house in Georgia or some other southern state, in its NHL list-article, where new owners had significantly renovated, so it lost its NHL but remained NRHP. But in fact the NX one is "Lincoln Motor Company Plant", for which we currently do not have an article(?), at 6200 W. Warren Ave., Detroit, Michigan, refnum = 78001521. Shows in Elkman output as status = NX, with date April 4, 2005 to be interpreted as the de-NHL-listing date. It is described properly in List of National Historic Landmarks in Michigan#Former NHLs in Michigan, but the article title Lincoln Motor Company Plant is a redirect.

If anyone's interested in the Lincoln Plant, I can pop over this weekend and take a couple photos. The most historically important portions of the plant have been demolished, but some of the newer portions remain. Note the right-hand photo on the state historic site v. a current Google streetview. Andrew Jameson (talk) 12:59, 28 December 2010 (UTC)

Please do! --Doncram (talk) 18:23, 28 December 2010 (UTC)

About the delisted ones, i wonder if those are relatively non-controversial to do by bot, as they are a) not important articles and b) we have something useful to say by a mere stub, in the newsflash that these are in fact delisted. There may be articles for a few of them, in which case the bot wouldn't post an article, but those are few. Note we don't even systematically tabulate these in our list-articles, though perhaps we should. Six example delisted ones in National Register of Historic Places listings in Grand Forks County, North Dakota#Formerly listed are stubs produced by batch-output then some editing. Even less text would be fine for these, IMHO. These are never going to grow much further. The NPS does not publish NRHP nom docs for these ones online. They're probably all demolished, gone. It can't be important to "save" them for some editor to do by hand.

What i think is more disheartening is that NRHP.COM and ARchiplanet and other sites have 100% coverage of NRIS-listed places in the 2009a version, while we have only about quarter covered in articles. Our list-articles contain pics in some areas of the country, which NRHP.COM does not, but have far less text info. --Doncram (talk) 06:42, 28 December 2010 (UTC)

Dudemanfellabra, your estimate of our articles doesn't exclude many that are merely contributing properties, and many that are architects, and others. I'd be curious to know how many NRHP infoboxes are out there (besides merely CP ones) as probably a better estimate of how many articles we have.

Thanks for pointing to Category:Former National Register of Historic Places, which i'll add to the Grand Forks ones now and put into any future batch-generated draft articles for NRHP places with this listing status. (Elkman could perhaps do the same?). BTW, there's no way that the batch-generator is gonna replace the Elkman service. It would be nice if they keep developing in a complementary way. --Doncram (talk) 18:23, 28 December 2010 (UTC)

Ah, I totally forgot about contributing properties.. well that makes our numbers even worse haha: 30,843 (number has increased since yesterday) articles - 21 delisted properties - 651 contributing properties + 84 individually listed contributing properties (that are also in the CP category by virtue of autocategorization by the infobox) = 30,255 total, which means we have articles for 35.88% of current NRHP listings.... about 0.7% less than before. Assuming that the CP articles are more than likely not started in huge stub drives (confirmed by a random spot check of the first page of the category), most of them are of decent start+ quality. This means that that 7.39% number I mentioned above should actually be lower, although I'm not sure by how much. If I had to guess, I'd say lower than 7.00%. If someone were willing to tabulate the class stats for all 651 CPs, we could find out exactly.. but the point is we just don't have a very good coverage of sites on the register... even after more than 4 years of work. Wow.

As a side note, I'm not even sure these refined numbers are accurate. During my spot check, I kept coming across unassessed articles or articles without NRHP project tags on their talk pages. Maybe we should start a drive to go through all the county lists and make sure they all have project tags? I could do Mississippi if you'd like. After the drive, we could reassess our progress and see if we can't make those numbers jump a little higher?--Dudemanfellabra (talk) 19:05, 28 December 2010 (UTC)

About how many NRHPs are listed, the 84,326 number is the 2009a version of NRIS's number. The recent new 2010a version of NRIS includes 85,923 with Listed status. Sanfranman59 pointed out that 2010a seems to include new listings up to about July, 2010. However our own current wikipedia number, given at List of RHPs, is 85,308. I believe that more than the NRIS numbers because of the many corrections relative to NRIS that our number reflects. I woulda thot our number would be higher than the NRIS 2010a number, because ours includes new listings since July 2010 probably up thru last Friday. I expect there are good reasons. In fact, one reason is that the NRIS listings total i report probably includes double counts for boundary increase changes. Our own number is what all of our work documents and it is what i would hang a hat on. :)

About a cleanup run to add WikiProject NRHP banner, that was done once by a bot request (June 2009?) and could be done by bot again. I happen to think that adding all the suggested banners and the photo request included in the Elkman generator Talk page suggestion is better for new articles (and I am now further adding batch-generated customized "Find Sources" tips, too). But a bot can easily add the NRHP banner for articles having an NRHP infobox or otherwise categorized in one of the NRHP categories. It should also be able to add NRHP banner for list-articles. Also fine if u want to do MS manually or by AWB or other semi-automated tool. --Doncram (talk) 19:25, 28 December 2010 (UTC)

Request for comment on NRHP article now listed as Everhope Plantation

I've been accused of making disruptive changes (User talk:Altairisfar#Twin Oaks Plantation) to an article that I created in January of 2009 with the name of Twin Oaks Plantation. I recently became aware of new sources (including a new book on Alabama plantations that I recently purchased) as to the correct current name, Everhope Plantation, and made some quick changes, including a page move (without adding the sources, though I had planned to do so on my next days off from work). These were reverted with the edit summary "NRHP and sources do not support changes." So the next day, I moved it back and did an expansion, using the new sources. I've been asked to move it back to the Twin Oaks Plantation name and undo my expansion, does anyone else feel that this would be appropriate? Thanks. Altairisfar 22:12, 28 December 2010 (UTC)

I don't think the changes are disruptive, per se. It's a fact of life that NRHP properties are sold, renamed, repurposed, and so on. One of my favorites, the Washburn "A" Mill, is now known as the Mill City Museum and is run by the Minnesota Historical Society. The National Register database doesn't list "Mill City Museum" as an alternate name, even though if you go to http://www.millcitymuseum.org/ they'll tell you all about the museum. Does the Minnesota Historical Society has a little more power to rename something and proclaim it publicly than the private owners of a plantation? I'm not sure, although it's easier to find sources for the renaming of the Mill City Museum than it is to find reliable sources about the renaming of Twin Oaks Plantation to Everhope Plantation. In any case, it's a matter of finding reliable sources, and assuming good faith. --Elkman ^(Elkspeak) 23:06, 28 December 2010 (UTC)

IMO the sources are very reliable, two are recently published books and the newspaper article is online at the Meridian Star website. Using Google search I quickly found photographic proof at this photographer's website. The sign in front of the house says it all. Altairisfar 23:46, 28 December 2010 (UTC)

Certainly seems it can be called Everhope Plantation, as one leading option, given that sign. For disagreements about article names in wikipedia, we have the nice wp:RM process which usually works very well. --Doncram (talk) 00:04, 29 December 2010 (UTC)

architects help wanted

FYI, at Wikipedia talk:WikiProject National Register of Historic Places/Architects2009a are counts from the 2009a version of NRIS for the most commonly named architects. Elkman once reported that Louis A. Simon has the most, but here's more. All of the NRHP-listed works for some of these architects have been identified in their articles (or in their Talk pages), including (from memory) for:

Lescher & Mahoney, a Phoenix, Arizona firm
Berlin Iron Bridge Co.
Ward Wellington Ward, of Syracuse, New York
Cass Gilbert (at Talk:Cass Gilbert)
Ralph Adams Cram (at Talk:Ralph Adams Cram), for Gothic churches

If anyone is interested in doing an article drive on any architect, lemme know and i can provide what i hope would be a helpful batch of drafted starter articles, like those at Talk:Ralph Adams Cram/drafts or better quality.

Also I could maybe run a report of counts by state, of top architects in each state. Any other architect reports of interest? --Doncram (talk) 22:50, 29 December 2010 (UTC)

I could do a better job generating batches-of-draft articles, for what appears in the infobox and what appears in draft text for an article, if some of you could help make a useful correspondence. What i need is to revise that list to provide a good correspondence of NRIS names to good wikilinks, i.e. direct names of architect articles or valid, good looking redirects. E.g., figure out that "Kahn,Albert" should link to Albert Kahn (which uses a pipelink as in [[Albert Kahn (architect)|Albert Kahn]]), and figure out that "Wenderoth,Oscar" should link to Oscar Wenderoth (strangely that last is a redlink right now). Only edit the link-text, please, i.e. what appears between square brackets, except replace "try" by "istobe". I bet that Elkman could use the same correspondence to good effect, too, in his NRHP infobox generator which is our mainstay. Anyhow, I'd like help! Help! :) --Doncram (talk) 02:05, 30 December 2010 (UTC)

Earlier I created stub article for Louis A. Simon and today created Penrose Stout stub. In between, have created architect bios as needed. I encourage everyone to keep bio stubs in mind, especially as architects historically worked locally.--Pubdog (talk) 02:32, 30 December 2010 (UTC)

Thanks! Acroterion is inserting section breaks into the list, so others can edit without hitting edit conflicts. --Doncram (talk) 02:39, 30 December 2010 (UTC)

New version, now in alpha order, of architect names ready for cleaning. Currently just the first 5,000. Some corrections to make good wikilinks are clear, like unwrapping "Lastname,Firstname" names, e.g. from Anshutz,Joseph W. to Joseph W. Anshutz. Some problems are clear, like there must be one correct version for all of these five (all currently showing as redlinks):

806 1 Andrews, Jacques & Rantoul *try* Andrews, Jacques & Rantoul
807 1 Andrews, Jaques & Rantoul *try* Andrews, Jaques & Rantoul
819 4 Andrews,Jacques & Rantoul *try* Andrews,Jacques & Rantoul
821 5 Andrews,Jaques & Rantoul *try* Andrews,Jaques & Rantoul
822 1 Andrews,Jaques and Rantoul *try* Andrews,Jaques and Rantoul

But which name should all five of those be changed to, in editing the wikilink? Assuming one name is chosen, then there will be just one wikilink used for the corresponding 12 (1+1+4+5+1) NRIS entries. Note, clearly the following one is different and should not be changed to include "and Rantoul".

820 1 Andrews,Jaques *try* Jaques Andrews

There are lots of architect-merger-type decisions which are unclear. I suppose i could revise the cleaning report to state in a following row, for architect names having just one NRIS entry, what is the refnum and name and city and state for the building or other property. Maybe that would allow a cleanup editor to do research in some of these cases. Hmm. --Doncram (talk) 08:13, 31 December 2010 (UTC)

Using Elkman's architect query tool, i find out they worked mostly in Massachusetts and Colorado, but they have one NRHP in Connecticut, where i can get a NRHP nomination document. I started Andrews, Jacques and Ranford article based on name used in Hartford Club's NRHP nomination document (just added to its article). Redirected the first two above to that. Will redirect Andrews, Jaques also (an example that shows unwrapping names like to "Jaques Andrews" can sometimes be wrong). In the architect list will add space to 3rd, 4th, 5th wikilinks above. So Jacques vs. Jaques misspellings are accepted via redirects. If it turns out true spelling is the other way, then architect article can be moved, but all NRHP articles created using batch-generator (and Elkman's if he chooses to use the cleaned up set of architect names) will reach the new architect article.

Other cleanups like this, where an editor creates a stub architect article and sets up just the sensible redirects, and cleans up the wikilinks in the Wikipedia talk:WikiProject National Register of Historic Places/Architects2009a list, would be appreciated. Oh, it's a list of architects from the 2010a version of NRIS now, not just those in 2009a version (it changed only slightly to add more; previous typos etc were not changed). --Doncram (talk) 16:06, 31 December 2010 (UTC)

(revised) Only the architects having multiple NRHPs are worth addressing this way currently. But, note many individual architect names have multiple NRIS spellings, e.g. "Name1,Name2" and "Name1, Name2" which can be combined (by putting in the same wikilink "Name1, Name2" for each of them). I wouldn't unwrap that, assuming it is unwrapping "Last, First" to "First Last", unless i am willing to really bet it is one person's two names, rather than two architects. Just combining by using the same "Name1, Name2" is a help requiring no additional knowledge. Also I wouldn't combine what i understood to be "F. Middle Last" and "First Middle Last" though. Maybe in wikipedia sometime later i would set up a redirect from one to the other, but you can do that only if u really know about the properties and persons involved. Like there is a William Aiken who is different from William M. Aiken his architect son, grandson or other relation, and many other architect father-son situations. F. Middle Last could be different person. Unless you know. --Doncram (talk) 16:49, 31 December 2010 (UTC)

How is unwrapping the names wrong? Would this just be "Jacques and Ranford Andrews" or "Jacques and Rantoul Andrews"? Like brothers? That's how I always unwrap names like that.--Dudemanfellabra (talk) 16:44, 31 December 2010 (UTC)

I would unwrap "Smith,Andrew M." because that could only be "Andrew M. Smith", but "Smith,Andrew" could be "Smith & Andrew" partnership. In this case, they're a partnership, 3 persons: one named Jacques, one named Andrews, one named Rantoul. --Doncram (talk) 16:49, 31 December 2010 (UTC)

You are correct (bottom left of the page under Epilogue No. 7):

Robert D. Andrews. Boston Architectural Club Exhibition Catalog (Boston: Boston Architectural Club, 1904). Andrews was a member of [ Henry Hobson Richardson ]'s office staff whose career flourished in Boston in the years following Richardson's death. Longfellow described Andrews's role in Richardson's office as "making pretty drawings of facades and perspectives" for the clients (AWL to ECPL, Nahant, 5 March 1882, LA). Andrews would joing Herbert Jacques and Augustus Neal Rancoul in practice.

This book (bottom of the page under C-5) says that a building in Dublin, New Hampshire was designed by an architectural firm including "Robert D. Andrews (1857–1928) and Herbert Jacques (1857–1916) of Boston." No mention of Rancoul/Rantoul/Ranford.

Also, a Google search on "Herbert Jacques architect" seems to imply he was once president of the United States Golf Association, as was his father (or possibly son?).

I can find no other sources except the Longfellow one above supporting the spelling "Rancoul". "Randford" only returns a few results, mostly in CT, where the most important person to them is Andrews, since he is a native of Hartford. "Rantoul", however, shows up everywhere. Dates on him are 1864–July 1, 1934 (died in Santa Barbara, California) [Source:[1]]. Also a history of Salem, Massachusetts says:

Augustus Neal Rantoul died in Santa Barbara, California. He was one of the sons of Robert S. Rantoul, who was an early Salem mayor and longtime President of the Essex Institute. Augustus was an architect whose firm in Boston was Andrews, Jaques & Rantoul. He was born and lived for many years at his father’s home at 17 Winter Street in Salem.

So pretty much, Rantoul is the correct spelling, while Jaques and Jacques are both used about equally everywhere. The page, then, should be moved to Andrews, Jacques, & Rantoul or Andrews, Jaques, & Rantoul.. not sure which one. If I had to choose, I'd go with the first just because throughout my life I've seen "Jacques" many times but never "Jaques". Who knows, though?

There is a lot of information about there about these guys, including some already on Wikipedia. They apparently designed the Boston Building, the Equitable Building, the Connecticut Governor's Residence, the Paine Estate, and Roughwood. Shouldn't be that hard to get a decent article going.--Dudemanfellabra (talk) 17:35, 31 December 2010 (UTC)

architectural styles

Not surprisingly, "Other" tops the list in frequency of architectural style types noted in the new 2010a version of NRIS. Greek Revival architecture, a favorite of mine, is the top real one with 7,839 NRHPs being recorded as having it. What is Early Republic style, anyhow? Not very common in NRHP, there are just 317 of this type. Early Republic architecture is a link to where something could be written about it; Category:Early Republic architecture is a brand new category. All 88 Exotic Revival architecture NRHPs could be covered in one list-article.

The 40 style types are unchanged in the new 2010a NRIS vs. the 2009a and perhaps all previous NRIS versions. Only 3 of these can be recorded for each NRHP listing, so historic districts whose NRHP nom docs clearly describe many more styles, only get 3. There's one free field for NRIS data entry of another architectural phrase, though, which allows for some unusual architectural descriptions. The 40 standard styles are, by frequency in the new NRIS:

 RANK ARCHITECTURAL_STYLE                                   COUNT
 ---- -------------------                                   -----
   1  OTHER                                                 20453
   2  NO STYLE LISTED                                       17771
   3  GREEK REVIVAL                                          7839
   4  COLONIAL REVIVAL                                       5473
   5  ITALIANATE                                             5360
   6  QUEEN ANNE                                             5355
   7  CLASSICAL REVIVAL                                      4830
   8  FEDERAL                                                4812
   9  LATE VICTORIAN                                         4378
  10  BUNGALOW/CRAFTSMAN                                     3639
  11  LATE 19TH AND 20TH CENTURY REVIVALS                    3184
  12  ROMANESQUE                                             2894
  13  GOTHIC REVIVAL                                         1962
  14  MISSION/SPANISH REVIVAL                                1525
  15  GOTHIC                                                 1501
  16  GEORGIAN                                               1490
  17  TUDOR REVIVAL                                          1464
  18  RENAISSANCE                                            1426
  19  ART DECO                                               1260
  20  LATE GOTHIC REVIVAL                                    1200
  21  COLONIAL                                               1189
  22  LATE 19TH AND EARLY 20TH CENTURY AMERICAN MOVEMENTS    1185
  23  EARLY COMMERCIAL                                       1114
  24  SECOND EMPIRE                                          1098
  25  BEAUX ARTS                                             1005
  26  STICK/EASTLAKE                                          912
  27  MODERNE                                                 773
  28  PRAIRIE SCHOOL                                          760
  29  MODERN MOVEMENT                                         693
  30  MID 19TH CENTURY REVIVAL                                670
  31  SHINGLE STYLE                                           599
  32  CHICAGO                                                 414
  33  MIXED (MORE THAN 2 STYLES FROM DIFFERENT PERIODS)       325
  34  EARLY REPUBLIC                                          317
  35  INTERNATIONAL STYLE                                     272
  36  ITALIAN VILLA                                           258
  37  PUEBLO                                                  135
  38  OCTAGON MODE                                            115
  39  EXOTIC REVIVAL                                           88
  40  SKYSCRAPER                                               87

Different coding of data within the 2010a NRIS might possibly work better for brand-new programmers/users of the data, but is a pain for those with programs expecting the 2009a format of data. I for one have decided to reverse the change in architectural style coding reflected in the 2010a version, by internally reconstructing the ARSTYLCD field that was in the 2009a, so that my programs following that conversion can run the same.

Your correspondent. --Doncram (talk) 06:12, 31 December 2010 (UTC)

Well, if its Early Republic in the U.S. I would say that its either Federal or a subtype of Federal architecture. Altairisfar 08:54, 31 December 2010 (UTC)

Lincoln Motor Company Plant

I've just created an article on the Lincoln Motor Company Plant, it of the singular NX designation. I suspect I didn't get the infobox designation correct, so please take a look. Andrew Jameson (talk) 18:04, 31 December 2010 (UTC)

Fixed the parameters. |delisted= only applies to delisting from the register entirely. If it has only been removed from a single designation, use the |delisted_nrhp_type= parameter.--Dudemanfellabra (talk) 18:09, 31 December 2010 (UTC)

Wait. I thought this was delisted from the NRHP entirely. Is it still technically listed? Andrew Jameson (talk) 19:22, 31 December 2010 (UTC)

I assumed that's what NX meant. If it were completely removed, it would be listed RN. In cases where a site is delisted from both NHL and NRHP status (e.g. Edwin H. Armstrong House), the NPS withdrawal page reads: "The Landmark designation was withdrawn on March 5, 1986 and the property was also removed from the National Register of Historic Places." In the case of the Lincoln Plant, however, the withdrawal page reads: "National Historic Landmark designation for the Lincoln Motor Company Plant was withdrawn on April 4, 2005." I would assume that since the Armstrong House page specifically states that it is no longer on the NRHP and the Lincoln Plant doesn't, that the Lincoln Plant is still on the NRHP but just not an NHL anymore.--Dudemanfellabra (talk) 19:37, 31 December 2010 (UTC)

Hum. If that's true, it should be listed on the National Register of Historic Places listings in Detroit, Michigan‎ (it's not, now). I assumed the Lincoln Plant was somehow NHL-listed, but never "ordinary" NRHP-listed, so that there was no NRHP listing to delist. Andrew Jameson (talk) 19:58, 31 December 2010 (UTC)

Hmm.. you have a point. Looking through the list of NHL withdrawals, The Musical Fund Society example says: "...the Landmark designation for the Musical Fund Society Hall was withdrawn on January 13, 1989. The building continues to be listed in the National Register of Historic Places." In the NRIS database, however, it is listed as LI – not NX or RN – and the article doesn't say anything about NHL status.

Also, the U.S.S. Cabot example never says anything about also being delisted from the NRHP, though the article claims it was. The First Pacific Coast Salmon Cannery Site (redlink) was formerly an NHL, and the withdrawal page doesn't say anything about it being removed from the NRHP as well, yet it's listing in the NRIS database is RN instead of NX, so I assume it was delisted from both.

The Isaac Franklin Plantation's withdrawal page doesn't mention being delisted from the NRHP as well, and it is listed in the NRIS database as RN. The article shows it being delisted from the NRHP (and never mentions NHL status), yet the Fairvue dab page says the site is still on the NRHP but no longer an NHL.

The redlinked Robert Frost Farm's withdrawal page says "The Landmark designation of the Robert Frost Farm was withdrawn on March 5, 1986; the property remains listed on the National Register of Historic Places," yet it shows up as LI in the NRIS database – not NX.

The Gambell Sites have been delisted from NHL status but remain on the NRHP, as is reflected in the article and the withdrawal page. It shows up, however, as LI and not NX in the NRIS database.

There are many more examples, but I think you get the picture. It looks like former NHLs that are still on the NRHP show up as just regular LI in the NRIS database. If they are completely removed from everything, they show up as RN. The withdrawal pages sometime state explicitly that the site was delisted from the NRHP as well, but sometimes it leaves that out. What makes the Lincoln Plant different is anyone's guess. I'd suggest conferring with the MI SHPO to see if they have any further information. A quick email to the NPS to get them to explain the "NX" listing code would be useful as well. I'm puzzled.--Dudemanfellabra (talk) 20:30, 31 December 2010 (UTC)

FI lists

In December there were 3 NRHP lists added to the fully illustrated category, one each in Wisconsin, Michigan, and Florida, with the last pix added by User:Royalbroil, User:Andrew Jameson and User:Ebyabe. Congrats! For 2010 the NRHP FI lists increased from 47 lists with 1301 pictures, to 126 lists with 4025 pix. That's a lot of progress. Cheers to all who contributed. Smallbones (talk) 16:41, 1 January 2011 (UTC)

Thanks! I'm very happy to say that I have fully illustrated lists in three states ... New York, Pennsylvania, and Virginia. Enjoying very much the fun of finding and photographing these historic places!--Pubdog (talk) 23:04, 2 January 2011 (UTC)

re: WP:WikiProject_National_Register_of_Historic_Places/Assessment#Requests_for_assessment

Doncram above suggested some change to this section since folks haven't been using it. I've cleaned out the 3 or 4 requests from 2010, but the 2 or 3 dozen requests from Sept 2009 and before remain. The directions or mechanics of the page don't seem to work well - articles that were reassessed remain in the to-do list at the top of this page. I'll suggest getting rid of the whole section and in the instructions put "If you wish an article to be reassessed simply remove the current assessment. The article will then be automatically be placed in the category of unassessed articles, and be assessed anew in due course." (or similar). If we can't reassess articles within a year under the current system, then we need to do something else. Smallbones (talk) 02:53, 2 January 2011 (UTC)

Images for Address Restricted Sites

We have a lot of "address restricted" sites in Missouri. These are mostly Native American burial sites and such. Obviously we don't want pictures of the actual sites as this kinda undermines the whole purpose of restricting the site addresses. That said, I would still like to fully illustrate our county lists. Are there any generic "archeological site" images that could be used for all these address restricted sites? If not, does anyone have any thoughts on what such an image would look like or the means to create one?HornColumbia talk 18:01, 2 January 2011 (UTC)

File:Address restricted.PNG--Dudemanfellabra (talk) 18:15, 2 January 2011 (UTC)

That works. Thanks!HornColumbia talk 18:19, 2 January 2011 (UTC)

Beat me! By the way, did anyone notice there's a :File:Address Restricted.png? Odd, that. --Ebyabe (talk) 18:23, 2 January 2011 (UTC)

Categories for discussion nomination of Category:Railway-related National Historic Landmarks

Category:Railway-related National Historic Landmarks, is a recently created category that has been nominated by another editor for deletion. It looks like NHLs are currently categorized by location except for this cat and lighthouses, both of which are fairly recent groupings. I don't have a strong opinion one way or the other but, if you would like to participate in the discussion, you are invited to add your comments at the category's entry on the Categories for discussion page. Thank you. RevelationDirect (talk) 04:25, 4 January 2011 (UTC)

Popular pages

I dunno why i was not aware of it before, but came across Wikipedia:WikiProject National Register of Historic Places/Popular pages, which is useful to Watchlist for monthly updates. Adding to my Watchlist now. Oddly, there is a AT&T Building (Kingman, Arizona) which stands out as a stub article but receiving a ton of hits. Perhaps from being target of redirect AT&T Building. Dunno where readers are really looking for, probably not the telephone exchange building in Obscuretown, Arizona (which i hev bin 2). And Keeper of the Register, which should be a Top priority for us, is another stub getting tons of hits. Many higher-quality-rated articles seem okay, not so much from our work but ok nonetheless. -:) -Doncram (talk) 05:50, 5 January 2011 (UTC)

This kind of relates to what I was suggesting in the "importance" discussion above... the articles that get the most hits should probably be marked as "high" importance, and should be a priority for the project to improve and maintain... as those are the articles that the readers are interested in. Blueboar (talk) 14:18, 5 January 2011 (UTC)

Yes, you're right. I should have gone back and said that high readership should certainly be one input to consider, in assessing importance. Your comment did change my thinking, although it is not reflected in the current draft priorities table. Thanks!

I dropped our banner from Gumbo and Battle of Peleliu, which were both showing in our Popular list. The Peleliu Battlefield article, to which i added our banner, needs development as a High priority, being a NHL and simply important, i note. Maybe there are other Popular articles we should exit from, too. --Doncram (talk) 15:10, 5 January 2011 (UTC)

importance ratings

I don't know why, originally, this Wikiproject did not assign Importance ratings to articles. Maybe so little was known, at first, that it seemed unhelpful? But can we agree now to begin to use importance ratings, and use that to guide our article improvement efforts?

FYI, i am finding myself in difficulty at Wikipedia talk:WikiProject Connecticut#New contest, where i am arguing that CT NHLs should be considered High importance for WikiProject Connecticut and hence be eligible for its new article improvement contest. Unfortunately, the National Historic Landmarks article, the List of NHLs in CT article, and most of the 60 separate CT NHL articles don't do a great job, yet, of explaining why they are important to the state and/or to the nation. (Comments there welcome! I'd be especially happy if anyone else would commit to developing a CT NHL article within the contest, if those would be deemed contest-eligible.)

It has been asserted from time to time that having short stub NRHP articles is somehow bad. I happen to think having a short stub is good and is all that is needed for many NRHPs, like for remote farmhouse ones which are NRHP-listed merely because of being the best local representative of a given architectural style. It's not important to further develop such articles, IMHO, although developing them is fine if someone local happens to get interested. However it is important to develop the National Historic Landmark ones; I would give all of those High importance now. I'd be interested in seeing our articles assessment table, at Wikipedia:WikiProject National Register of Historic Places#Collaboration and assessment, show how many of these high importance ones are stubs vs. start vs. better. --Doncram (talk) 18:11, 1 January 2011 (UTC)

For starters, we'd have to amend Template:WikiProject National Register of Historic Places to allow for an |importance= parameter. Then we'd have to create a guideline like this one. Then we'd have to go through all 30,941 (as of 2011-01-01, 18:26 (UTC)) articles and assign them importances. A daunting task to say the least.--Dudemanfellabra (talk) 18:42, 1 January 2011 (UTC)

How to determine "Importance"? Probably the main reason that we haven't had an importance rating is that it would be so subjective, and everybody would have his or her own favorite. It's likely worthwhile to give these ratings, but how would we do it? Would there be disagreements and how would these be resolved? Some factors that might be included, and maybe one that shouldn't be: NHL status - almost everybody here would agree on that - plus maybe number of views - there's probably a few quirks here but if folks are interested enough to read it, then there must be some importance, ditto FA or GA status - if an editor has taken the effort to take it this far, it's probably pretty important at least to that editor (but somehow I'd guess that most editors would disagree with this criterion!). Perhaps date of listing? In many cases it seems that very important sites were listed early, but those with any questions about importance were only listed later. Are there any other "objective" criteria? Doing 30,000 ratings by purely subjective criteria would likely be impossible. Smallbones (talk) 18:54, 1 January 2011 (UTC)

I see in that Wikiproject Cities guideline Dudemanfellabra points to, that at its subsection on "Priorities", there is specific guidance based in part on city populations. That's good and helpful. I also like how editor Markvs88 composed detailed guidance for CT's ones at Wikipedia:WikiProject Connecticut/Assessment#Rating criteria. For NRHP, how about putting all NHLs at High, and start by assuming all others are Low? NHLs are about 2% (approx 2,500 / 85,000) of all NRHPs i think. Say Mid importance should get 5% more. Then ask editors like Pubdog and Ebyabe who have done a ton of articles in NY and MD and FL, say, to go through their states' articles and use their judgment to identify Mid vs. Low importance ones. I think most of us could do a fair job of assessing which are the most important NRHPs, in counties where we're already familiar with all the articles. Or we could use the NRHP's own identification in nom forms, with Mid meaning having national-level importance asserted in the nom, rather than just local importance. We also can use other WikiProjects' ratings as a guide, too, tho they have different priorities.

By the way, where did the quality Ratings request page go to? There was a backlog of ratings requests, but I can't find it from our main page. Should be at, or linked from Wikipedia:WikiProject National Register of Historic Places#Collaboration and assessment. I recall previously thinking the quality Ratings request page maybe oughta be canned, because requests were getting no response, but i thot it still existed. --Doncram (talk) 19:06, 1 January 2011 (UTC)

I don't know if there is a way to compile the information, but I would suggest rating the importance of any article based on how many other articles link to it. Those that are linked to in lots of other articles are most important, because those are the ones that are most likely to attract readers. (I suppose I am saying that we should not worry about how important the building is... and instead worry about how important the article is to Wikipedia). Blueboar (talk) 19:47, 1 January 2011 (UTC)

Chicken and egg problem. I'd rather use the importance ratings to prioritize articles that should be developed. Note part of improving the higher priority ones would be to increase their prominence among wikipedia editors by our visiting town article and other related articles and adding links to them.

For architect articles:

those with 100 or more NRHPs (about 10) get Top importance rating by wp:NRHP
those with 10-99 NRHPs (about 245) get High importance
those with 5-49 NRHPs (about 650) get Mid importance
those with 1-4 NRHPs (about 30,000) get Low importance, if they get an article at all

Based on ranking of top architects by count (5 or more), reflected here, plus there being 39,218 distinct architects in NRIS 2009a version (but including many duplications differing merely by grammar/spelling). --Doncram (talk) 19:59, 1 January 2011 (UTC)

(EC)For ratings request page see the 1st item in the to-do list. I'll also suggest getting rid of it, and further suggest that if you really want a new rating, just remove the old one, which will result in it going straight to the unrated category. Looking at the priority table, my 1st comment in this section looks pretty ignorant. We'd really have to go with something almost completely objective, and then be satisfied that it won't fully capture what most of us think as "importance." Working from that table, the NRHP "mid" which I didn't previously know existed, here's a possible starting point (additions, deletions, and general editing encouraged):

Smallbones (talk) 20:01, 1 January 2011 (UTC)

When I started the project, lo those many moons ago, I deliberately avoided having "priority/importance" ratings. I couldn't see a fair way of judging that due to subjectivity, as Smallbones said. Would one state's be more "important" than another's. I thought it could lead to all sorts of contentiousness.

Having said that, what DMFB proposes has merit. It seems fair, and is at least a good starting point.

The idea of having to prioritize all the Florida articles, though, gaah. What I get for making so dang many of them. :) --Ebyabe (talk) 21:41, 1 January 2011 (UTC)

Priority scale

Label	Criteria	Examples
Top	Articles about the NRHP itself Sites agreed at wt:NRHP to have truly exceptional importance Top level list-articles of NHLs, of NRHPs	National Register of Historic Places National Historic Landmark Contributing property National Historic Preservation Act of 1966 State Historic Preservation Office National Park Service No specific sites yet discussed List of NHLs;List of RHPs
High	National Historic Landmarks NRHP-listed sites with NPS designations (NMON, NHS, etc.) Founders of NRHP program Architectural styles reflected in 500 or more NRHPs Architects of 100 or more buildings on the NRHP Top-level NHL list and NRHP list for each state A few exceptions to be discussed at wt:NRHP	Manzanar Frank Lloyd Wright;Tourtellotte & Hummel Greek Revival architecture List of NHLs in AL;List of RHPs in AL
Mid	Listed individually on the NRHP, plus some other distinction to be discussed at state list talk page (as long as less than __% of NRHP listings in state) Valid articles corresponding to focused Thematic Resources/MPS studies Architectural styles reflected in 50 or more NRHPs Architects with 50 or more buildings on the NRHP All other geography-organized list-articles; other top-level list-articles	Atalaya Castle Boston Public Garden Land of the Cross-Tipped Churches Herkimer Home State Historic Site Shingle Style architecture;Exotic Revival architecture Conde McCullough;Ward Wellington Ward National Register of Historic Places listings in North Side Chicago;List of Elks buildings
Low	All other NRHP sites or districts Architects with 10 or more buildings on the NRHP List-articles subordinate to other NRHP articles	Phoenixville Historic District List of contributing properties in the St. Anthony Falls Historic District
Bottom	Contributing properties All other architects All other architectural styles	(examples needed)

I just edited slightly the above table. I think articles about the register itself should also be top-importance, and I think sites that have also been recognized by other registers (NPS, state, or local organizations) should be included with the individually listed CPs. What do you guys think?--Dudemanfellabra (talk) 20:45, 1 January 2011 (UTC)

New Jersey NRHPs are all NJ State RHPs, and I'd guess this happens in many other states - seems like there will be too many "highs" in these states. Smallbones (talk) 21:58, 1 January 2011 (UTC)

In Lauderdale County, Mississippi, my home county, there are 46 NRHP listings. There are also 29 state landmarks. Of the NRHP listings, only 11 are also state landmarks. I haven't done the numbers for all the counties, but I would expect them to be largely the same. I'm not saying all states are like Mississippi (the country would truly be hell if they were), but if this is the norm in state landmarks, I think 20-25% of listings being high importance would be ok.

If, however, Mississippi is an exception, then state/local landmarks can be taken out. NPS designations and individually listed CPs could suffice I guess... I do feel like the vast majority of our articles, though, would be mid-importance without the state/local qualifier. Maybe we could just make exceptions to the state/local rule for states like New Jersey where the registers are virtually identical?--Dudemanfellabra (talk) 22:20, 1 January 2011 (UTC)

I'd really rather keep the top priorities small, to really allow us to focus effort, and to allow locally knowledgeable editors room to kick up the importance from Low to Mid of something they know is really important. I like setting default ratings objectively, as suggested by several, but how about then allowing the NRHP editors in a given state, discussing at Talk page of the state-level list-article, with invited participation of WikiProject STATE editors, getting to refine the objective rating system in their state, as long as they keep High % to less than 3% and Mid or higher to less than 10%. So as to disregard NJ and NY state designations, because there every NRHP is automatically getting that rating, but in NY using the handful of State Historic sites to some effect. And allowing for explicit discussion, like how in central New York the Niagara Mohawk Building, which should really be a National Historic Landmark but doesn't get a boost from any objective criteria so far discussed, has to get higher importance rating.

(Second table commented out)

I have to say I like the previous table better, yours is too tough. I'd think the basic listing should be "Mid-importance" - calling something "low importance" might be viewed as a put down by some people. The only buildings I really see it apply to is contributing buildings that aren't individually listed - in some HDs there are 100s of contributing buildings, and to give them all "Mid-importance" would degrade the importance classification. Could you describe the architect related distinctions a bit better? I'm wondering exactly how they fit in. Certainly NHL should be of top-importance - I don't think they are too numerous, even if they got up to 5% of the total NRHP sites. "Objective" default classification has to be better than potentially endless discussions on importance -level, but any specific property should be open to discusion. Please limit these requests however! Smallbones (talk) 23:53, 1 January 2011 (UTC)

I agree mostly with Smallbones. I've adjusted the above table to incorporate some of what you've brought up, but overall I think it's too strict as well. I think it's more clear for everyone if we just have one table on the page, so I've commented the second one out. Let's discuss down here and edit the table above.

I didn't think about architect articles and have now incorporated them, saying anyone with 50+ listed buildings should be high importance. I don't think they're really top-worthy.. this project is in fact about the actual places listed, and architects are kind of a byproduct.. if the architects are very prolific, however, they should be recognized. 25+ are now mid and <25 is now low importance. I've also distinguished between contributing properties to NHL districts and regular districts (those being mid and low importance, respectively). There is also a note about discussing the state/local qualifier on state talk pages. Good compromise?--Dudemanfellabra (talk) 00:37, 2 January 2011 (UTC)

Maybe I'm just misreading it, but it seems Doncram wants all the NRHP articles on the buildings of those architects to be given those importance-levels, not the articles on the architects themselves. I don't particularly object, but it does seem to be a new direction for the project to focus on the architects, rather than the building or the history. Perhaps it's overdue. Smallbones (talk) 01:02, 2 January 2011 (UTC)

I meant just the articles on the architects themselves to get Top rating. There are only about 10 architects having 100 or more NRHP places (not in order): Frank Lloyd Wright; Louis A. Simon; the National Park Service itself; Civilian Conservation Corps; Works Project Administration; James Knox Taylor; Charles L. Thompson; James A. Wetmore; Tourtellotte & Hummel; maybe McKim, Mead & White. These architect articles should be part of WikiProject NRHP, and these ones will be linked from several thousand NRHP places articles in total. On the criteria of which are the most-linked articles from other articles in our wikiproject, these are among the tops. Maybe only National Register of Historic Places and National Historic Landmarks will get more links from NRHP articles. Well, i guess Greek Revival architecture should get more, about 8000 links. So the top architectural style articles should also become high importance, in my view. I did not mean for every James Knox Taylor-designed post office to be given Top rating.

About putting all the NHLs in Top importance, that is crazy. We cannot have 2,400 articles being rated of Top importance. That is not prioritizing at all. There should be more focus put on the really important articles which IvoShandor worked on in the past, namely National Register of Historic Places and Historic district (United States) and a couple others.

I don't see why a tougher set of criteria, with several nuances like handling the Atalaya & Brookfield Gardens-type situation, and conveying how/where WikiProject STATE editors can be invited in, should be commented out. We can handle seeing more than one alternative proposed set of priorities.

"Tougher" is a misnomer if u think that is bad; it is not an insult to call all regular NRHPs "Low" priority. All of our articles are currently Low priority, by default, from the Elkman generator, with regards to WikiProject STATE importance levels given. Just which ones do we want to say are more important for us to focus our efforts on? It is not particularly helpful to distinguish between regular individual NRHPs vs. mere contributing properties. There are too many of both of them. What we need is separation of a few much smaller classes which are plausible real priorities that could be addressed in an article drive. We did an article drive on NHLs in the past; that is feasible. To do an article drive on all the individual NRHPs vs. all the contributing properties is not useful, as an article drive cannot be done on either one of those, so it is not worth making the distinction. The priority classes should be reserved as candidates for article improvement drives.

Also, there needs to be separate criteria for prioritizing our list-articles. --Doncram (talk) 04:22, 2 January 2011 (UTC)

P.S. Grumblethorpe should not be High priority. I have never heard of it and can't believe it is important. It is merely NRHP-listed plus being a contributing property in an NHL district. It is NOT a key contributing property in that district; it is not the namesake of the district; it is just one of about 10 separately-NRHP-listed properties in the district; it is not even described near the top of the NHLD article. Only like if the NHL district is named "STATE Capitol Historic District" and includes a bunch of buildings, should the STATE Capitol building itself be considered a key property individually. --Doncram (talk) 04:38, 2 January 2011 (UTC)

P.P.S. It's interesting there is some sentiment in above comments for giving some priority to the NMON and NHS and other higher-than-NHL-rated-by-our-national-government-itself ones. By the way only the NMONs which are NRHP-listed should be included, not the ones having no historical significance and not being NRHP-listed, to be clear. The Mount Rushmore article and many others of these are really crummy, from the perspective of NRHP and what we know now. They were created early on by WikiProject Protected areas (from which Ebyabe split off Wikiproject NRHP) and still sadly show made up / completely inaccurate stuff, like that Mount Rushmore is an "IUCN Category V (Protected Landscape/Seascape)" place, which is in fact completely false. It was a good guess back then, but it was a wrong guess. The WikiProject Protected areas has been too long dormant, perhaps. A year or so ago there is a complication there with Orlady and one or a few others having gotten into battle mode about what they wanted "Protected area" to mean, contrary to what it does mean by any organization. So there is a bit of a mess. But i do agree we should ourselves put some higher priority onto fixing those up; many of them are more important than NHLs and only don't have NHL status because they ALREADY had the higher designation; e.g. Independence Hall in Philadelphia. --Doncram (talk) 04:54, 2 January 2011 (UTC)

And just why can't we have 2400 top importance articles? When you compare that to the 80,000+ total article count this project will eventually encompass (if it doesn't fly apart at the seams first), it's a very small and focused effort comprising of roughly 3% of the articles. It seems to me that you want to focus on something like one tenth of 1 percent of the articles in the project and group all the others together into one big lump. If that's what we are going to do, why bother to rate importance at all? Why not just pick a few articles and start collaborating on them?

Separating articles into bigger, more inclusive categories with less strict requirements is much more desirable and accurate IMO than only selecting a few articles to stick out from the rest, which are all clumped into one big group. I see no reason 2400 articles is too big of a starting group when compared to the breadth of the project. In your comments, you yourself claim a drive on NHLs is feasible in one paragraph.... but two paragraphs before that, you say the same thing is absurd! What sense can one make of that? Either we can or we can't; make up your mind. If we went with what you want and only elected very few articles to stand out as top importance, the drive would be over in no time, and there would be pressure to open up the importance ratings more so that another drive could be completed. That or there would be a drive moving down to high importance, then mid importance, etc. By the time we got down to low importance, we'd be slapped with a group of 70,000+ articles in one importance bracket, and we'd be looking for a way to split them up anyway. In other words, we'd be continually revising and upping the importance of more and more articles in the long run, so why not just make them big now and be done with it? The drive(s) will last longer, and I think in the end the importance levels will more accurately reflect the actual importance of historic sites.

About splitting up the lower importance articles more, I think it is very helpful to distinguish between CPs and sites that are actually listed. A site that is actually on the register (and thus deemed to be important in its own right) is more important and deserves more attention than a site that must be grouped with others into a district to reveal its historical significance. I can see where you're coming from with "pivotal" contributing properties, and I could agree to move them to a higher importance than regular contributing properties, but any contributing property that is listed on its own IMO trumps even pivotal CPs that are non-listed.--Dudemanfellabra (talk) 06:51, 2 January 2011 (UTC)

I'd like to reserve Top importance for the really important articles that we should come back to again and again. The ones linked from hundreds of our own articles. Maybe i misspoke otherwise. There was one article creation drive on the 2,400 NHL articles to start all the missing ones and to add at least the NHL webpage reference for each one, but there's never before or since been an article improvement drive on that scale in this wikiproject. Trying to work on all 2,400 is too many for a drive, unless it was really minimal like an AWB-supported campaign to make just a few edits about categories or something. It would be useful to know, though, out of those 2,400 and a relatively few other High importance articles, which are rated Stub vs. Start vs. other, to allow for smaller collaborative drives to improve them.

You ask "Why not just pick a few articles and start collaborating on them?" Great question. What does it take to get even a few members here willing to cooperate in a real article improvement drive. I suggested working on National Register of Historic Places listings in Grand Forks County, North Dakota recently, for various reasons, but don't see a lot of joiners on that. Is there any different set of articles that you would pick for a collaborative article improvement drive? --Doncram (talk) 21:50, 2 January 2011 (UTC)

A problem I have with the scale as proposed is that it's top-heavy. There ought to be way more "low-importance" articles than anything else, but right now, in the proposed scale above, "low-importance" is reserved for buildings and people that are typically not notable enough to warrant an article anyway. (There are specific exceptions, obviously, but not, I think, enough to even be near the 80,000 NRHP-listings.) This effectively produces a three-level importance scale rather than a four-level scale.

Furthermore, I don't think importance levels should be assigned based on the need for an article drive, or the perception that "low-importance" articles are somehow "no-importance" articles. Importance levels should be assigned only to differentiate the articles that are of greater importance to the average reader. Greater importance means an ever-narrowing pyramid of numbers.

As an example, look at WikiProject Biography. They have a short, resonable statement on the relativity of their priority scale. And, specifically, they've chosen to restrict the number of "top-importance" to 200 out of a current article list of over 800,000 (and a potential article list of far more than that). That's a little extreme, but Michigan has only 23 "top-priority" articles out of 7000-ish written articles, and Connecticut (the example given above) has 100 top-priority out of 5000-ish written articles. Presumably, for both states, the "top-priority" (and most of the "high-priority") articles are already written, so further articles would be primarily in the mid- and low-importance areas.

Break (importance ratings)

Looking at both Michigan and Connecticut, I see article-count dropoffs in the range of 3X to 10X as the importance increases. Note these are for articles already written; presumably, for both states, further articles would be primarily in the mid- and low-importance areas, increasing the dropoff rate there. This NRHP project has a relatively limited scope of "historic places" articles: the 85,000+ listings (plus some additional contributing property articles) for something like 100,000 up to maybe 150,000 articles total, ignoring architects and lists and other things for the nonce.

Given that, I think the project should shoot for a priority scale that leads to dropoff rates in the 5X to 10X range. For example, at 7X, that gives:

Low priority: ~120,000 "historic places" articles
Mid priority: ~18,000 "historic places" articles
High priority: ~2500 "historic places" articles
Top priority: ~350 "historic places" articles

Obviously one can argue about exactly what dropoff rate is appropriate, but I think placing most articles in the "mid priority" category defeats the purpose of a priority scale. Andrew Jameson (talk) 13:00, 2 January 2011 (UTC)

Is there any way to make the ratings more than 4 levels? How about 5? I see what you're saying about all the example projects that you've given (I've never really paid much attention), but I still stick to the idea that CPs should not be on the same level as places that are actually listed. Maybe we could add a level (see Wikipedia:WikiProject Comics/Assessment) where we put CPs? Then we could spread out the above to give us more breathing room.

I've done so above. I demoted NHLs to high importance and upped NPS designations to the same level as NHLs. Everything else was shifted down a level so that most of the articles will now be in low importance. How does that look?--Dudemanfellabra (talk) 17:36, 2 January 2011 (UTC)

I don't see any additional rating level at Wikipedia:WikiProject Comics/Assessment. I revised the single proposal table to reflect fact there is no bottom priority level currently possible, and for other sensible-i-think adjustments. --Doncram (talk) 20:50, 2 January 2011 (UTC)

That seems approximately OK to me. I assume Dudemanfellabra is thinking of "bottom priority" articles as something like "articles that have an indirect connection to the NRHP" (cribbing from the comics page). This seems to me like a slightly seperate category, so that there isn't necessarily a dropoff in numbers from "bottom" to "low" because "low" is not a more-important subset of "bottom" in the way that "mid" is a more-important subset of "low." Discussing only historic places articles (ignoring lists and architects and project pages once again), that sorts out, to me, like:

Bottom importance: Articles that have an indirect connection to the NRHP, for example contributing properties.
Low priority: Most NRHP sites and districts.
Mid priority: Greater-importance NRHP sites and districts, to be hashed out by regional editors. Should be approx 10% to 20% of the number of "low priority" NRHP sites and districts.
High priority: Most NHLs. Possibly some truly exceptional NRHP sites as determined by all NRHP editors.
Top priority: Greater-importance NHL, to be hashed out by all NRHP editors. Should be approx 10% of the number of "high priority" NHLs.

I'm a bit reluctant to advocate purely objective sorting ("Listed individually on the NRHP, plus some other distinction to be discussed at state list talk page" or "Architects with 10 or more buildings on the NRHP" or "Architectural styles reflected in 50 or more NRHPs"), other than as a rough guideline, because I don't think a checklist of criteria is condusive to determining something as elusive as architectural and historical importance. Other WPs seem to make do with awfully squishy definitions ("This article is very important to the project, as it covers either a general area of knowledge or provides information about a significant topic" from WP:Comix, frex), and the hard-line distinction between NRHP-listed and NHLs is already far a more objective criterion than most WPs get. Andrew Jameson (talk) 00:48, 3 January 2011 (UTC)

If you can't see the Bottom category at WP:Comics, you must not be looking very hard? It's part of the table that looks exactly like ours. Perhaps you should look again? Here's a direct link to the section... right there in the table.. plain as day.

Yeah i saw it later, after i had added a correct link to the page u meant. You're welcome. :) --Doncram (talk) 01:58, 3 January 2011 (UTC)

I've modified the above slightly but only semantically and readding the bottom importance level. I still don't think CPs and regular listings should be on the same level. I've also spread out architect/style articles more so that most of them will be bottom importance and only the really prolific ones will be any higher. We should focus on the actual register.. not biographies.--Dudemanfellabra (talk) 01:43, 3 January 2011 (UTC)

I don't think "Low importance" is at all anything negative, but "Bottom importance" sounds sorta like an insult to me. The Comic wikiproject uses it seemingly derisively for articles that are trivial / merely trivia collections. Currently i am not familiar with the term, but I don't think i'd want it assigned to any building article i developed. So what if a building is a CP in a historic district, I don't think it can be established from just that, that it must be labelled as less important than many low-real-importance NRHPs which i happen to know about. If a historic district is defined, a component building is usually not going to become separately NRHP-listed, because there's no further tax or other financial benefit, and there is real cost to getting a separate NRHP listing through. So there are many worthy buildings that are not going to be separately listed. --Doncram (talk) 01:58, 3 January 2011 (UTC)

I don't think it's really that bad to insult CPs.. I mean we won't be the only people giving them importance ratings.. and even if we were, they aren't really that important anyway (at least from a historical pov) if they aren't on some register. Importance ratings are project-specific; something we rate bottom importance may be extreme-super-top importance to some other project (i.e. a project focusing on the city). The fact of the matter is, though, that this project is about sites on the National Register of Historic Places... and contributing properties, though close, are not on that register, so we shouldn't be giving that much attention to them. I'm not saying they're of ZERO importance, but they should definitely be at the bottom of the list.

On top of that of course there is always wiggle room, and I mean honestly there's no real way to enforce the importance ratings (or quality ratings either for that matter) we make here, so bumping something up from bottom to low or low to mid isn't going to cause that much drama in the whole scope of things. These criteria are more like a rough starting point; I don't imagine it's going to turn into a checklist or anything like Andrew Jameson is fearing. There may in fact be some NHLs that are less important than ordinary NRHP listings, and there may be some contributing properties that are more important than even NHLs (pushing it...). Nothing is set in stone here. There are no "rules". Let's just come up with a basic outline so we can begin assigning articles to their places and if later on we decide to reassess some of them, then that's no biggie. I think what we have now is pretty good. Why don't we go ahead and start?--Dudemanfellabra (talk) 04:49, 3 January 2011 (UTC)

I don't think that we should get hung up on terminology. Dudemanfellabra is proposing an additional category for articles that "have an indirect connection" with the NRHP--let's call it "related" rather than "bottom"--and that such things as contributing properties be tagged in this fifth category--"related"--rather than as "low." Personally, I don't care all that much whether CPs are tagged "low" or "related": it's reasonable to differentiate contributing properties from individually listed sites, but I see the main importance of a priority scale as distinguishing the small percentage of higher priority articles, whereas distinguishing the small percentage of lower priority articles isn't so important. Andrew Jameson (talk) 12:52, 3 January 2011 (UTC)

If this Lowest level can be termed Related or perhaps better Other, or something other than Bottom, then it's fine/good to have one more level as a bonus type. Can use for Owner objection cases, where deemed eligible but not actually Listed, and perhaps for Delisted cases, too. Please note i do feel that creating minimal stubs for Delisted and probably also never-listed Owner objection cases is worthwhile (if only to correct the public record about what places are NRHP-listed), and this might end up being done by some mass creation drive. Having an "Other" rating tag to apply to such minimal stubs would be a good clarification that even the articles' creator judges them not worth not much more development. It's gonna be seen and used as a judgement, though, that nothing more than a stub article is wanted, which will conflict with local users' views in some cases. I would want for there to be no absolutely no guff given, if any one of us or any new local user wants to claim a given CP or former, demolished building is more important and "deserves" a higher rating. No consensus required: if anyone says a place "deserves" better because "it was important when the town still existed" or any other-possibly-lame-sounding-reason, okay? If anyone wants to develop such an article to Start or higher, or to bring to DYK, they probably do not want a "Lowest"-type rating attached to it, and no one should battle them to force that upon them, and they should not have to get anyone else's approval to just rate it higher. --Doncram (talk) 13:48, 3 January 2011 (UTC)

Maybe "Low priority" could be "Basic importance" if there is objection to "low". I like basic, implying the bones or basis of the structure, upon which the rest is built. I definitely like either Related or Other better than Bottom. Lvklock (talk) 15:34, 3 January 2011 (UTC)

Thanks. How about naming it simply "Lower"? --Doncram (talk) 15:45, 3 January 2011 (UTC)

"Minimal"?--Dudemanfellabra (talk) 16:50, 3 January 2011 (UTC)

"Related" works for me, other variants might also work. Somehow the whole thing looks like a lot of work. Smallbones (talk) 04:08, 4 January 2011 (UTC)

"Related" works. No one picked up on "Lower" and maybe that would be confusing out of context of seeing it placed on a scale. Is it meant as being above or below "Low"? "Basic" seems positive about anything with that label, maybe too positive though if applied to never-listed Owner objection cases. "Minimal" can seem as negative as "Bottom". Dudemanfellabra, at this point i for one would be happy if u could just Boldly turn it on, if u want. Not sure what it takes to create the new priority level but i assume u have that figured out.

About it being a lot of work, I'm pretty sure a bot could take care of assigning High for all the NHLs and assign Low as the default for all others, then we could make further changes. We can certainly just rate the non-Low ones, and do that gradually. I agree, though, that it's "low priority" to place these priority ratings, even for a bot. Maybe it could be included with some other bot's edits eventually. I would be interested to see how the National Historic Landmark articles are spread across Stub-Start-C-B-GA-FA quality ratings, but not enough to AWB them all myself. Eventually I hope it would help us focus some future efforts and to mitigate some past concerns. --Doncram (talk) 01:19, 5 January 2011 (UTC)

If people are worried about insulting other people's articles then perhaps you should take the approach of the maths project and use the term "priority" instead of "importance". Low-priority does not reflect on the article itself (which some may regard as important) but simply says it not a priority for the project. — Martin (MSGJ · talk) 16:43, 5 January 2011 (UTC)

Template:National Register of Historic Places/sandbox has now been modified to allow importance ratings. Though we can change the ratings to be customized, User:WOSlinker has suggested that maybe we should just use "NA-importance" – a category that already exists by default. Personally, I like "related-importance" better, but if we want to use NA, no further edits have to be made to the template; it can be copied over as-is.

I also like MSGJ's suggestion to use priority levels instead of importance levels. I think if we did that, "bottom-priority" doesn't sound as insulting, so it would be fine to use. What do others think?--Dudemanfellabra (talk) 17:18, 5 January 2011 (UTC)

I don't think putting importance=NA works for the lower/lowest priority articles being discussed. We already use that for 60 or so WikiProject NRHP pages like the new article announcements. I wouldn't want the category to mix mainspace and really non-mainspace type stuff.

I don't see the value in changing wording to use "priority" rather than "importance". In all of the STATE wikiproject banners put into >20,000 NRHP talk pages already, "importance" is used. I imagine there would be errors when editors would not know whether it is "priority" or "importance" to use in whichever cases. I don't personally see the distinction, and believe it is not worth making the distinction.

If it is not immediately easy to create a custom importance level, that is okay. All the lower levels can be put into Low, if anyone is actually categorizing, to be split out later perhaps. I would only bother with adding Top and High and maybe Medium categories for a long time. To enable us to really look at the Top and High articles and consider improving the Stubs and Starts among them. We've done without importance ratings forever, and there is no immediate use (like for an article drive) to the Low vs. Related distinction, anyhow. --Doncram (talk) 17:49, 7 January 2011 (UTC)

Importance levels have been created and are now in use. See Category:National Register of Historic Places articles by importance. Rate away!--Dudemanfellabra (talk) 03:04, 8 January 2011 (UTC)

NRHP level of significance codes

For what it's worth, the National Register database assigns a level of significance to each property. The levels are "Local", "State", "National", and "International". I'm not sure about the ones marked "International" (see below), but the national level ones look to be accurate, like Chrysler Building, Lever House, Fort Snelling, James J. Hill House, and Washburn A Mill. Properties significant at the state level, like in Minnesota, are places like Cedar Avenue Bridge, Foshay Tower, Hesper (shipwreck), Intercity Bridge, and so on. Finally, the ones significant at the local level, like in Minnesota, are places like the Lumber Exchange Building, Jackson Hotel, Great Northern Depot (Princeton, Minnesota), and Soo Line Depot (Crosby, Minnesota). I would tend to think that the ones important at the national level should be ranked higher than the ones important at the state and local levels.

As an aside, there are only 30 with a level of "International", but I'm wondering if those are accurate. Forestville Baptist Church is marked International, but I'm not sure if that's accurate. --Elkman ^(Elkspeak) 22:27, 2 January 2011 (UTC)

That seems interesting to me, but only as a rough suggested guideline. I'd be curious how the significance correlates with my own subjective feeling of the importance of sites I'm familiar with. Andrew Jameson (talk) 00:53, 3 January 2011 (UTC)

Rejected nomination?

On nationalregisterofhistoricplaces.com, there is an entry for the Chilhowee Park Historic District in Knox County, Tennessee, supposedly listed in 2005. The reference number given is 00501039. However, the Chilhowee Park HD is apparently not on the Register, and a search for the reference number on the NRHP's database brings up no matches. Furthermore, I found a nomination form for the Chilhowee Park HD here (starts on page 4), dated January 2009, with the exact same streets given as the 2005 listing. Does this mean it was nominated in 2005, but the NRHP rejected it? Bms4880 (talk) 21:07, 6 January 2011 (UTC)

This comes up often, e.g. User talk:Royalbroil#Brown County NRHP listings recently. "NRHP.COM" is inaccurate, showing places as listed on the date that they were in fact delisted, or date nomination was received (but not yet accepted and possibly never to be accepted), or other National Register entry other than listing. U can check status by Elkman's generator, per advise in that other discussion. --Doncram (talk) 23:57, 6 January 2011 (UTC)

What's the actual status here then? It coulda been nominated in 2005 but not accepted then. It's been nominated in 2009 and still not accepted? I notice no links to redlink Chilhowee Park Historic District and no NRHP list-article linking to Chilhowee Park. If it was accepted i would think it woulda gotten added to the List of RHPs in Knoxville.... --Doncram (talk) 16:22, 8 January 2011 (UTC)

The Elkman generator shows it with 2005 date and having status "DR" which means "DATE RECEIVED/PENDING NOMINATION", consistent with it having been nominated but not accepted. Since NRHP.COM and ARchiplanet.com and other NRIS mirrors will incorrectly be stating it is NRHP-listed, i think it is now worthwhile for us to have an article on this one, just to set the record straight, even tho it seems not to be listed. And to have it mentioned in a lower table on the Knoxville NRHP list-article, like delisted items. --Doncram (talk) 16:29, 8 January 2011 (UTC)

I'd be very careful about adding extra items to the County lists. If we were to be complete about "sites not listed" - well that could be a very big list! If we were to try to correct everybody else's mistakes - well that might be an even bigger list! Please excuse my attempt at humor, but I'd like to know how we are to know these things, how big the extra lists might be, and what would be the real benefit of expending the extra effort. I have no problems, btw, with somebody writing an article about the non-listed site, I'm just interested in keeping the list pages fairly pure and simple, when I don't know about the cost/benefit tradeoff. Smallbones (talk) 17:19, 8 January 2011 (UTC)

I'd suggest our centrally generating county tables for the delisted ones, to be added to the county list-articles, in full tables like lower table at National Register of Historic Places listings in Grand Forks County, North Dakota, which can accommodate pictures and coordinates and descriptions. Listed and later demolished cases are interesting, fully valid topics IMHO. That would add maybe 2 items for each 100 item NRHP list-article. In the new version of NRIS, the relevant tallies are:

LISTED 85923
PENDING/LISTED 2403
REMOVED/LISTED 1651
OWNER OBJECTION 1057

I'm not sure what "Pending/Listed" means, but think it includes cases like Chilhowee Park Historic District, not yet listed. I think those plus Owner Objection cases should be added into a more streamlined table that is perhaps not much more than a list. --Doncram (talk) 17:40, 8 January 2011 (UTC)

Categories 1) and 3) are in the tables now - no objection. Category 2) - we might as well wait until we find out exactly what it is. Category 4), I'd rather not bother. One point against it is that the owners have asked to not be listed presumably for reasons of privacy or perhaps to avoid vandalism, looting, etc. Smallbones (talk) 18:57, 8 January 2011 (UTC)

Maybe i should have said i don't know why in the recent 2010a NRIS the status code is labelled "Pending/Listed" now. I think it is grammatically bad way of saying "nomination pending, not listed". I believe these are just the places that got labelled "DR" which means DATE RECEIVED/PENDING NOMINATION, before. See the counts by status type for 2009a NRIS version vs. 2010a NRIS version tabulated for you at wp:NRHPhelp.

There are a few owner objection cases listed in tables in some NRHP list-articles i think, but those are few. Another editor or two has previously objected to them being there, i think not in any central discussion but rather by edits removing them where they had been added. I would like to just have it established in central discussion that these are welcome.

About your suggestion that we might need to protect the privacy owners of owner objection cases, i think that train has left the station. The NRIS data is public domain. Those sites are detailed in other webpages like NRHP.com and countless "local attractions" type pages, which disregard the actual fact of non-listing of the place and actually suggest it should be visited. And there has seemed to be no consensus, or not a very full consensus, even for us to keep locations of address restricted archeological sites private. The owner objection cases are, from what i recall knowing about, not ever like that. These include cases where, for example, a large corporate owner's lawyers seem to have come down in the end with a judgment that a local division's wish to list a historic building doesn't improve the larger corporation's bottom-line interests (i recall some dam in Oregon or Washington; and one of the Charles Scribner Buildings in NYC). I don't know of any case like what you supposed here.

It is also theoretically possible that an Owner Objection case is one where local preservationists tried to get a threatened building preserved in part by getting a National Register listing of eligibility at least, against the owner's wishes. I think in some cases like those the building has been demolished by the owner (or "accidentally" burned) by the time we notice it showing in NRIS, and then NRIS would not ever be updated i bet, since the building was never listed due to the owner objection. If they weren't burned, those buildings would be subject of local newspaper coverage already, and there is no privacy issue for them.

Anyhow for all of these i think it would be useful to get the places listed on the county list-articles, as red-links, in order to allow for process of article name conflict discovery and resolution, which takes time. A future bot to create simple objective articles on the delisted and owner objection ones would do better once the place names had been validated by existing for a while. --Doncram (talk) 21:51, 8 January 2011 (UTC)

Note

I noticed on the CPHD nomination form that one of the areas of significance is 1956-1959. It's possible they decided to wait until 2009 in order to meet the less-rigorous 50-year rule. In any case, I've split off and expanded the neighborhood info to Chilhowee Park (neighborhood). Bms4880 (talk) 23:47, 8 January 2011 (UTC)

Neighborhood article looks good. Hopefully my adding a mini-section on the NRHP nomination is useful and not too much there. I redirected the HD name to there, too. --Doncram (talk) 18:23, 9 January 2011 (UTC)

[1]