User talk:Scs

Keeping threads together: If you post here, I'll reply here. If I post on your talk page, I'll look for your reply there.

Peace award

Congratulations. --Dweller 12:33, 9 January 2007 (UTC)[reply]

Did you see this? --Dweller 09:23, 11 January 2007 (UTC)[reply]

Pastafarian Award

[[:Image:Touched by His Noodly Appendage.jpg|thumb|right|This Pastafarian award goes to Steve Summit in recognition of his elaborations on building a spaghetti bridge. Unceremoniously presented by Sluzzelin ]] Hello Steve Summit. I found your insightful elaborations on building a spaghetti bridge interesting and appetizing. I prepared myself some pasta and decided to award you one Flying Spaghetti Monster for your answer at the Science Reference Desk. ---Sluzzelin 23:13, 28 January 2007 (UTC)[reply]

Heh, this gave you away. :) ---Sluzzelin 20:23, 29 January 2007 (UTC)[reply]

Scsbot

Here is a little reward for creating the ever-so-great archival bot for both the help desk and reference desk. — E ^talk_bots 06:26, 17 August 2007 (UTC)[reply]

Why, thank you! —Steve Summit (talk) 11:25, 17 August 2007 (UTC)[reply]

		The da Vinci Barnstar
		For creating the ever-so-great archival bot for both the help desk and reference desk. — E ^talk_bots 06:26, 17 August 2007 (UTC)[reply]

The Butterfly Effect

		Have a Pie!
		You are hereby awarded ONE PIE for having the line of the day on the Science Desk!

Arakunem^Talk 19:52, 28 October 2007 (UTC)[reply]

Yum! Thanks! :-) —Steve Summit (talk) 19:57, 28 October 2007 (UTC)[reply]

Help desk date headers

Hi - I noticed this edit and thought to myself "surely there's a bot that used to do this". Looking into this I gather Scsbot does this, but you run it semi-automatically. Is there some particular reason this task can't be done fully automatically, or is the issue that you don't have a machine you can use for scheduled tasks? Just curious. -- Rick Block (talk) 03:58, 4 January 2009 (UTC)[reply]

It's a combination of several things. I do have a machine I could run it on in a scheduled way, but it's not ideal. I used to do the semi-automatic invocation just about every evening without fail, but changes in RL lately preclude that. But I haven't worried too much, because I have this feeling that there are probably plenty of people who enjoy the opportunity to make an extra little contribution manually. (But on the other other hand, I may yet fully automate it, somehow.) —Steve Summit (talk) 03:14, 6 January 2009 (UTC)[reply]

I run a scheduled bot and could likely do this. It generally pains me to see folks doing things by hand that look automatable. If you think you're actually not likely to get to this anytime soon (I mean, like months) let me know and I could take a whack at it. -- Rick Block (talk) 03:39, 6 January 2009 (UTC)[reply]

It's up and running (on the "less-than-ideal" machine) now. Thanks for the prod, and the offer. —Steve Summit (talk) 13:35, 12 January 2009 (UTC)[reply]

No problem. Thank you for running it! I have thought for a while there should be a public, extensible, framework for scheduled periodic bots that would allow folks to upload their bot source. Rough sketch: some machine somewhere (m:toolserver maybe) runs cron jobs that kick off at various periodicities (hourly, daily, weekly at least) and downloads a protected set of files containing bot tasks written as Unix shells (and/or using any of some set of bot frameworks, like pywikipedia). Then, the tasks are run and the log files uploaded to a log file. The basic point is that many of the automated periodic tasks are more or less necessary to keep en: running (see Wikipedia:Maintenance/tasklist, which is far from complete or current). This seems like a much more wiki-like solution than what we do now (all the bot source would be publicly available, with suggestions for changes going through talk pages since the files would be protected). I have intended to write a generalized periodic bot for a while, but haven't gotten around to it. -- Rick Block (talk) 14:44, 12 January 2009 (UTC)[reply]

Interesting you should mention Unix shell scripts. I suspect there are those who would opine that shell scripts are hopelessly old-school, and have no place in a "modern" environment like mediawiki...

...but not me, because sh is precisely what I chose to implement most of Scsbot in. :-)

I don't know what the best (or even a good) centralized infrastructure for Wikipedia bots might be, but I do agree that they're far more important than most people likely realize, and probably deserve something cleaner than the current ad-hoc distributed welter. —Steve Summit (talk) 00:21, 13 January 2009 (UTC)[reply]

user:Rick Bot's basic pattern is "curl | awk | replace.py" (it's not literally a pipe all the way through, but it's certainly the general idea). I've thought about adding "transform using an external executable" to replace.py, but haven't gotten around to it yet. I have submitted a patch allowing an entire replacement file to be given to replace.py (seems significantly easier than using the api directly). If I get around to writing a generalized periodic bot I'll let you know. -- Rick Block (talk) 05:22, 13 January 2009 (UTC)[reply]

Bot error

FYI, User:Scsbot ate part of someone's post while adding the new date header here. Algebraist 00:12, 14 January 2009 (UTC)[reply]

Gaahh! That's not supposed to be possible. Thanks for catching and fixing that.

The comment it truncated was very long, so clearly the bot has a line-length problem somewhere. (Although 3805 is an odd sort of limit...) —Steve Summit (talk) 01:44, 14 January 2009 (UTC)[reply]

Ref Desk archive bot Q

Steve, please see our discussion here: Wikipedia_talk:Reference_desk#Transcluded_Q_talk_pages. StuRat (talk) 21:30, 23 January 2009 (UTC)[reply]

File:Walschaert stamp.jpg listed for deletion

An image or media file that you uploaded or altered, File:Walschaert stamp.jpg, has been listed at Wikipedia:Files for deletion. Please see the discussion to see why this is (you may have to search for the title of the image to find its entry), if you are interested in it not being deleted. Thank you. Calliopejen1 (talk) 14:42, 4 February 2009 (UTC)[reply]

Potential enhancement for scsbot

I recently stumbled across this page which appears useful but deprecated, in part because it exceeds the maximum transclusion size limit. In an attempt to fix that, I'd like to wrap <noinclude> tags around the transcluded date pages at the beginning of each desk. That would prevent each desk's header and old questions/responses from being double-transcluded onto the 'All' page (which, I'm led to believe, counts double against the transclusion size limits). From my tests, excluding double-transclusions has reduced the total transclusion size to less than the maximum limit (and appears to provide a snappier response as well). While I can manually apply these tags (and have already done so), it would be preferable to have scsbot automatically place newly-transcluded content inside the closing </noinclude> tag. Is this something you would be willing to consider? Thanks! – 74 19:14, 21 February 2009 (UTC)[reply]

Certainly! I'm on vacation now, but if I don't get back to you on this within the next week, remind me. —Steve Summit (talk) 16:07, 22 February 2009 (UTC)[reply]

Thanks! And I see you had to manually fix the transcluded pages because of the extra tag; sorry about that. We can move the </noinclude> tag until scsbot can handle it correctly; I was just verifying that they would fix RD/ALL before I bothered you with the request. – 74 16:59, 22 February 2009 (UTC)[reply]

Right. And I was about to say, yes, until I can tweak the bot's editing heuristic, we'll have to undo your change. But I should be able to get to that tweaking within a few days. —Steve Summit (talk) 14:24, 25 February 2009 (UTC)[reply]

Okay, all done. (Bot now noinclude-aware; no need to undo anything after all.) —Steve Summit (talk) 02:34, 26 February 2009 (UTC)[reply]

I noted it appears to be working quite well, and WP:RD/ALL is fixed. Thanks! – 74 03:06, 26 February 2009 (UTC)[reply]

Small error of the bot

An IP encountered and reported on WP:VP/T the following broken link, which I fixed. --TheDJ (talk • contribs) 00:27, 3 March 2009 (UTC)[reply]

It's a known bug, which I have been dragging my feet on fixing, based on the following lame excuse.

Before this bot took over, those items were not links to different points within that page. (See, for example, the older Wikipedia:Reference desk/Archives/Humanities/January 2007.) Making them all "hot" was an idea I had for an improvement, which ended up being easy for the bot to do. (Or, seemed to be easy -- but it was easy because it was incomplete.) But, figuring that having 99% of the links "hot" was better than nothing, fixing the bug that causes the remaining 1% to be broken has (alas) never been at the top of my priority list... —Steve Summit (talk) 01:19, 3 March 2009 (UTC)[reply]

Liability paranoia

Regarding your comment at WT:RD here, I wanted to clear something up.

I can't speak for other editors, but to my mind, liability concerns are not the only reason why we might want to remove a request for medical advice. Indeed, I consider such concerns to be relatively minor, and – with some caveats – the effects of a lawsuit more a matter of nuisance than of serious liability risk.

I've plugged my essay on this topic before, but I'll hit it again: User:TenOfAllTrades/Why not? Briefly summarized, I'm interested in preventing harm to the OP, to the responders, and to Wikipedia's reputation. (If pressed, I'd probably put those in that order of priority, too.) By responding to questions about the poster's health in this forum, we do everyone involved a disservice.

I'm not trying to browbeat you into a change of your !vote (I think that voting on these matters is generally unproductive in any event), nor would I want you to take this as intended as an attack on your judgement. I just wanted to be clear that I don't – and I don't think we should – view these questions through the lens of legal paranoia. Cheers! TenOfAllTrades(talk) 01:51, 5 March 2009 (UTC)[reply]

Fair enough! Thanks for replying. —Steve Summit (talk) 02:00, 5 March 2009 (UTC)[reply]

In the middle of nowhere there is nowhere to hide

Happy Vacation! :-) ---Sluzzelin talk 13:09, 20 May 2009 (UTC)[reply]

Bug in scsbot?

Any idea what happened here [1]? Buffer overflow because my comment was too long? Nil Einne (talk) 17:55, 8 June 2009 (UTC)[reply]

Ouch! Yes. It's a known bug, but I thought I'd taught the bot to detect and not commit a change if it realized it was going to damage some unrelated text like that. Thanks for pointing this out -- I'll have to figure out why the bot's double-check didn't catch it. (Or, better yet, figure out and fix the stupid buffer overflow...) —Steve Summit (talk) 23:25, 8 June 2009 (UTC)[reply]

Bot removed a bunch of today's comments on the science desk...

Diff, FYI. I'm guessing that it wasn't supposed to do that... --Kurt Shaped Box (talk) 07:50, 17 June 2009 (UTC)[reply]

EDIT: I've restored all the 'lost' replies. --Kurt Shaped Box (talk) 07:56, 17 June 2009 (UTC)[reply]

Arrgh. My mistake. (Not really the bot's.) Thanks for fixing. —Steve Summit (talk) 13:29, 17 June 2009 (UTC)[reply]

Today's Date section removed from Help_desk

I noticed it get added this morning, now it's gone! Also, it showed June 22nd as being archived, but when I looked at Wikipedia:Help_desk, 22nd June entries are still there! I'm assuming someone's later edit removed them? PhantomSteve (talk) 07:28, 25 June 2009 (UTC)[reply]

Yes. The bot inserted it here , someone accidentally edited and removed it here, someone else fixed the header level here, and finally some third person reinserted the date header here. (Oh! That last was you. :-) )

When a day's entries get archived, they're always transcluded back onto the desk for another day (or for three days on the reference desks). This trades off how long the entries stay visible, versus how large the page is to load to edit. —Steve Summit (talk) 11:09, 26 June 2009 (UTC)[reply]

Yes, I thought I'd better re insert the date header! Thanks for explaining this - I didn't realise that they're transcluded! PhantomSteve (Contact Me, My Contribs) 11:27, 26 June 2009 (UTC)[reply]

Mathematics reference desk oddities

I'm not quite sure what's going on, but something strange is happening with the date section headers and archiving on the Wikipedia:Reference desk/Mathematics page. -- Tcncv (talk) 04:08, 7 July 2009 (UTC)[reply]

Do you mean the empty days for July 5 and 6? ~~Those are empty because no one ever posted anything on those days.~~ (Not because the bot archived them prematurely or anything.) —Steve Summit (talk) 02:30, 8 July 2009 (UTC)[reply]

There were several threads started on the 6th. Algebraist 02:39, 8 July 2009 (UTC)[reply]

Oh! Duh. So there are. Hmm. —Steve Summit (talk) 02:41, 8 July 2009 (UTC)[reply]

The bot definitely screwed that up a bit, though I'm not sure how, and I don't have time just now to fully investigate. But I fixed up the headers, and it looks like other desks were not similarly affected. —Steve Summit (talk) 02:54, 8 July 2009 (UTC)[reply]

It looks like the same thing happened again, and I believe I see the pattern. After adding the July 11 header a few days ago, no new topics were added for either July 11 or July 12. I guess nobody thinks about math on weekends. For some reason, the lack of new topics prevented Scsbot from adding the July 12 and July 13 date headers in its normal processing. Finally on July 13, a couple of new topics were added, appearing under the July 11 header. The next time Scsbot came through, it played catch-up and added the July 12, 13, and 14 headers in reverse chronological order. (I have since fixed these headers.) This same scenario occurred over the July 4 weekend as you can see here. -- Tcncv (talk) 01:45, 14 July 2009 (UTC)[reply]

Scsbot issue: parsing snafu

Howdy. In this edit scsbot seems to have become confused (I guess with that rather complex signature markup) and mangled things up slightly. I've fixed it for that page. -- Finlay McWalter • Talk 00:15, 4 September 2009 (UTC)[reply]

It's repeating the same error with every visit to that page [2], [3] -- Finlay McWalter • Talk 00:21, 6 September 2009 (UTC)[reply]

RD troll

In response to your question here, yes — that was a troll. Any messages posted to the Ref Desks (or particularly to the Ref Desk talk page) from IP addresses resolving to Tiscali UK are almost certainly the banned troll Light current (talk · contribs · deleted contribs · logs · filter log · block user · block log) socking away. It's weird, actually. He's got a reasonable grasp of electronics, and I suspect he's at least university-educated, but he has a penchant for asking childish questions involving Uranus puns, and he's willing to age socks for a year or more to do it. Moreover, he'll use that sock drawer to violently attack anyone who removes his trolling. It's quite a remarkable phenomenon; I don't know entirely what's wrong with his brain, though I do hope he eventually gets the help he needs.

Nominally Tiscali owns the IP addresses 79.64.x.x through 79.79.x.x (though I don't see edits from all of those /16 subnets). A second range from the same ISP (which also turns up from time to time, but which I haven't seen for a while) is 88.108.x.x through 88.111.x.x. In general, if you see any of those addresses making any sort of borderline post on the Ref Desks, it's safe to revert and block. Used to be I could also ask Alison to run a checkuser to empty out any new socks he'd salted away, but alas — she is retired. TenOfAllTrades(talk) 21:22, 21 September 2009 (UTC)[reply]

Please refrain from making personal attacks such as stating a contributor has something "wrong with his brain" and hoping he "gets the help he needs," even if the contributor in question has been banned. Edison (talk) 15:08, 21 October 2009 (UTC)[reply]

Oh, that was Light Current? That explains everything; sorry you had to spend time with the longer explanation.

(I know all about Light Current, and Tiscali. I used to be able to recognize a Tiscali IP address immediately. I actually started out on his side, back in the day, but then eventually I was one of the ones who was getting violently attacked for removing his trolling, as a skim through my 2007 talk page archives will show.) —Steve Summit (talk) 18:21, 22 September 2009 (UTC)[reply]

Your comment was the most succinct representation of the core issue here - Many people (myself currently included) would much rather spend their limited Wikipedia time actually contributing to the encyclopedia, rather than wasting time on (trolls). I concur strongly. Nimur (talk) 20:17, 16 October 2009 (UTC)[reply]

A canned comment would be preferable to venting spleen. A goal of a troll is likely to get someone so upset they start ranting. That is why I find the standard canned warnings useful for vandals, rather than any angry scolding I might think of, which is likely what they want. There would be far less reward in getting the same canned reply each time. Edison (talk) 15:19, 21 October 2009 (UTC)[reply]

You cited an essay calling for reverting. blocking and ignoring vandals. Please study the part where it says "Long-term vandals will quickly grow tired when all of their "work" is quietly reverted, their accounts/IPs blocked, and their cries for attention ignored, with no fanfare whatsoever." When someone repeatedly attacks the vandal, in violation of a core policy against personal attacks, it has exactly the opposite effect intended by "revert-block-ignore." The essay definitely does not say "Revert, block, ignore, then post amateur psychiatric diagnoses that they have 'something wrong with their brain' and 'they should get help.' " Edison (talk) 15:32, 22 October 2009 (UTC)[reply]

You are responding to two people at once. I was the one who cited WP:RBI in this comment on your talk page (a comment which I later regretted, and was thinking of deleting this morning if you hadn't yet responded to it). TenOfAllTrades is the one you're accusing of posting amateur psychiatric diagnoses. —Steve Summit (talk) 16:36, 22 October 2009 (UTC)[reply]