User talk:Oleg Alexandrov

Write a new message. I will reply on this page, under your post.

Mathbot and disambiguation pages

Hi, I've noticed that Mathbot adds pages to the "List of mathematics articles ..." articles based on transclusion of the {{mathdab}} template. Problem is, that leads to these articles having links to disambiguation pages. For instance, List of mathematics articles (S) links to no less than 41 disambig pages, as you can see here. This is a problem for the WP:DPL project; not only does it give us false positives for which articles need fixing, but eventually we want to use a bot to tag articles with excessive number of dablinks (the {{dablinks}} template that was on the article a few days ago); this would lead to a once-a-day edit war as one bot adds the tag and the other removes it.

Since all of these dablinks are intentional, there's an easy solution: WP:INTDABLINK. It's a convention we use to identify dablinks that don't need "fixing", and our Toolserver scripts are instructed to ignore links with the INTDABLINK format.

For instance, say you have a link to [[Secant]]. To mark it as intentional, pipe it through the (disambiguation) redirect: [[Secant (disambiguation)|Secant]].

The logic is, if you're linking directly to a (disambiguation) page, it must be intentional. (For instance, the {{otheruses}} template.) So we tell our reports - the same ones that will eventually be used by the bot - to ignore links like this. This works even when you link to a redirect with (disambiguation) in its title.

Do you think you could modify your bot to use this syntax when you get articles that transclude the {{mathdab}} template? Thanks, --JaGa^talk 03:33, 14 January 2011 (UTC)[reply]

I don't think that would work. I routinely use this page to do joins in the toolserver database: the canonical way to get a list of all mathematics pages is to make a list of pages linked directly from one of the subpages of list of mathematics articles. If we changed so that it doesn't link to the dab pages, then those dab pages won't show up in our lists based on pagelinks.

I appreciate the maintenance script problem, but it would probably be simpler to just tell the tool that lists pages with too many dablinks to skip the lists of mathematics articles, since we know they will have lots of dablinks. — Carl (CBM · talk) 03:14, 15 January 2011 (UTC)[reply]

It would still link to the dab page; Secant (disambiguation) redirects to the dab page. --JaGa^talk 03:27, 15 January 2011 (UTC)[reply]

So... any decisions made about this? --JaGa^talk 04:59, 5 February 2011 (UTC)[reply]

It won't work to link to a redirect instead of linking to the page itself, because that will not put the correct entry into the pagelinks table in the mediawiki database. The actual target of the link matters here, not just the eventual destination you would get to by clicking on the link. The best solution is to have the maintenance script skip any page whose title starts with "List of mathematics articles". These pages are intended to link to dab pages if the dab pages are in the scope of the math project. — Carl (CBM · talk) 14:37, 5 February 2011 (UTC)[reply]

I'm not going to hard-code article titles into the scripts, for obvious maintenance reasons. I've got a Toolserver account myself. Could you let me know more about what you're doing? I could probably write a solution for you. BTW, why are you answering instead of Oleg? --JaGa^talk 17:37, 5 February 2011 (UTC)[reply]

I help maintain the bot that updates the list of math articles. It runs from the 'wpmath' multi-mainatainter project on toolserver. Oleg is not as active anymore, although it's usually possible to get him by email. So it's faster for me to respond.

The bot goes through a list of mathematics categories and makes a list of all articles in those categories. The list is useful for people who want to browse through math articles, and also useful for maintenance.

In its maintenance capacity, I use queries something like this, which makes a list of all math articles that have prod tags on them:

select art.page_title
from page as art 
join categorylinks on art.page_id = cl_from 
join pagelinks on pl_title = art.page_title  and art.page_namespace = 0
and pl_namespace = 0 
join page as src on src.page_id = pl_from 
where src.page_namespace = 0 
and src.page_title regexp '^List_of_mathematics_articles' 
and cl_to = 'All_articles_proposed_for_deletion'
order by art.page_len asc;

If the list of math articles linked to a redirect instead of linking directly to the article, this query would fail to find those articles, because the necessary entries in the pagelinks table for the third join wouldn't be there.

I use similar queries for many other things, for example to make a list of all non-free images on math articles. The key point is that if you join the pagelinks table against all the subpages of List of mathematics articles, you get a complete list of mathematics articles that can itself be joined with other tables in the database.

This is an example of a time when it is perfectly correct to link to the dab pages, because we want the links to the dab pages to be in the pagelinks table.

I don't see it as worth a significant amount of effort to completely redesign our system simply to avoid linking to dab pages. It's trivial to just skip the lists of math articles in a dab-fixing tool. Or not; if someone changes them accidentally, the math bot will fix it the next time it runs, no real harm done. — Carl (CBM · talk) 12:56, 6 February 2011 (UTC)[reply]

(Sorry for the long absence.) I started looking at the data today, and noticed there are already many redirects on these pages. For instance, List of mathematics articles (M) has 39 redirects that I can see (for instance, Mann's theorem and Metra Potential Method). I'm still going to take a stab at the query, I just wanted to let you know it would be an improvement beyond its usefulness to the WP:DPL project, since there are already redirects on these pages that are slipping through your queries. --JaGa^talk 20:37, 18 April 2011 (UTC)[reply]

OK, here's the new query:

SELECT art.page_title AS title,
       art.page_len AS len
FROM page AS art
JOIN categorylinks ON art.page_id = cl_from 
JOIN pagelinks ON pl_title = art.page_title
AND art.page_namespace = 0
AND art.page_is_redirect = 0
AND pl_namespace = 0 
JOIN page AS src ON src.page_id = pl_from 
AND src.page_namespace = 0 
AND src.page_title REGEXP '^List_of_mathematics_articles' 
AND cl_to = 'All_articles_proposed_for_deletion'

UNION

SELECT art.page_title AS title,
       art.page_len AS len
FROM page AS art
JOIN categorylinks ON art.page_id = cl_from 
JOIN redirect AS red ON art.page_title = red.rd_title
AND art.page_namespace = 0
JOIN page AS r ON r.page_id = red.rd_from
AND red.rd_namespace = 0
JOIN pagelinks ON pl_title = r.page_title
AND r.page_namespace = 0
AND r.page_is_redirect = 1
AND pl_namespace = 0 
JOIN page AS src ON src.page_id = pl_from 
AND src.page_namespace = 0
AND src.page_title REGEXP '^List_of_mathematics_articles' 
AND cl_to = 'All_articles_proposed_for_deletion'

GROUP BY title
ORDER BY len ASC;

It looks like a lot of changes, but there's really little to it. I'd be happy to rewrite all your queries if that would help. Also, I'd recommend you use queries like this regardless of whether you decide to support WP:INTDABLINK or not; as it stands, the many redirects already on the bot-maintained pages slip through the original query. --JaGa^talk 06:51, 20 April 2011 (UTC)[reply]

The Signpost interview

"WikiProject Report" would like to focus on WikiProject Mathematics for an upcoming edition of The Signpost. This is an excellent opportunity to draw attention to your efforts and attract new members to the project. Would you be willing to participate in an interview? If so, you can find the interview questions here. Just add your response below each question and feel free to skip any questions that you don't feel comfortable answering. If you have any questions, you can leave a note on my talk page. Have a great day. – SMasters (talk) 04:09, 7 February 2011 (UTC)[reply]

Mathbot and Wikipedia:Articles for deletion/Old

I just had to revert mathbot's last edit to this page because it showed no open AFDs which is incorrect. It probably has something to do with the recent upgrades. --Ron Ritzman (talk) 14:30, 16 February 2011 (UTC)[reply]

I'll take a look soon. Oleg Alexandrov (talk) 17:32, 16 February 2011 (UTC)[reply]

Uh, the html code of the AFD page changed dramatically. This won't be a simple fix. It may take a few days for me to fix it. Mathbot (talk) 04:41, 18 February 2011 (UTC)[reply]

I think I fixed the bot. It was a somewhat big change, so there is some likelihood that I may have missed something. If you notice something odd then please let me know.

Thank you for acting as mathbot in the meantime, that was a lot of counting to do. :) Oleg Alexandrov (talk) 04:43, 22 February 2011 (UTC)[reply]

On an unrelated note, the bot is listing arbitrary breaks in AfDs as separate open AfDs. Any way to prevent that from happening? Logan _Talk ^{Contributions} 00:16, 1 March 2011 (UTC)[reply]

Thanks. I fixed this. Oleg Alexandrov (talk) 18:06, 1 March 2011 (UTC)[reply]

Would you take a look at Beeblebrox (talk · contribs)'s concern at Wikipedia talk:Articles for deletion/Old#I don't like the new interface? Thank you, Cunard (talk) 05:05, 6 March 2011 (UTC)[reply]

I commented there. Oleg Alexandrov (talk) 17:06, 7 March 2011 (UTC)[reply]

Thank you for the revision. Cunard (talk) 10:30, 9 March 2011 (UTC)[reply]

Mathbot seems to be glitching now for titles with parentheses in them. It replaces them with random numbers, causing the articles to display as red links. Any ideas why? Logan _Talk ^{Contributions} 01:59, 9 March 2011 (UTC)[reply]

I fixed this (hopefully). This is due to the fact that Wikipedia handles special characters inconsistently. It sometimes encodes "(" to ".28", but in other contexts it fails to recognize this as as an encoding of "(". Please let me know if there other problems. Thank you for monitoring the tool. Oleg Alexandrov (talk) 01:24, 11 March 2011 (UTC)[reply]

Will do soon the "Carryl Varley" bug as well (here article name is not the same as AfD subpage name). Oleg Alexandrov (talk) 01:29, 11 March 2011 (UTC)[reply]

Fixed that one too. link Oleg Alexandrov (talk) 07:17, 11 March 2011 (UTC)[reply]

How to award a barnstar?

Hi Oleg, I have a simple question on how to award a barnstar to the guys who have added the fields "mr" and "zbl" to the {{Citation}} template: did you know them? Who are them? Thank you for your help. Daniele.tampieri (talk) 12:36, 23 March 2011 (UTC)[reply]

I guess you can look at the edit history of Template:Citation to see who added the fields. To award the barnstar, there is no formal procedure, just paste the text/picture on their page. Oleg Alexandrov (talk) 18:36, 23 March 2011 (UTC)[reply]

How to report a possible conflict of interests?

Hi Oleg, I'm disturbing you again since I have placed in the "Solomon Mikhlin" entry a link to his MacTutor biography which is authored by Vladimir Maz'ya, Tatyana O. Shaposhnikova and myself. I remember that in such cases it is safe to report a possible conflict of interests, but I don't know how to do. Can you help me? Daniele.tampieri (talk) 06:51, 11 April 2011 (UTC)[reply]

I think a conflict of interest is something worse than that. :) You have nothing to gain personally from linking from one article partially authored by you on Wikipedia to another article partially authored by you on another web site. And the external article is relevant to the Wikipedia article. So I think everything is OK. Oleg Alexandrov (talk) 16:00, 11 April 2011 (UTC)[reply]

Thank you very much. Precise and helpful as ever. :D Daniele.tampieri (talk) 18:18, 11 April 2011 (UTC)[reply]

Colors of animation for Heat Equation article

Oleg, hello. I was just doing some minor proofreading fixes to the Heat Equation article, when I noticed that your nice animation is actually showing a 2D system, and that the caption for the animation did not match the system being simulated. I made some changes to the caption, which I think now better explains what is going on.

Along the way I realized that in the animation right now green=hot and red=cold. This is opposite to the more usual coloring, where red typically means hot. The pictures would be easier for a beginner to understand, I think, if that assignment of colors could be changed, maybe using blue for cold, red for hot.

I see your Matlab code to generate the pictures. I would try to change the coloring myself, but I don't have Matlab. So I was wondering how you might feel about making that change and re-uploading the animation -- or maybe suggesting some way I could run Matlab online, or something like that.

Thank you.

Regards,

Ralph Dratman Dratman (talk) 15:54, 21 April 2011 (UTC)[reply]

Thanks. I'll reverse the colors one of these days. Oleg Alexandrov (talk) 16:16, 21 April 2011 (UTC)[reply]

I think you just have to change the colormap from Autumn to Jet. Dratman (talk) 22:36, 21 April 2011 (UTC)[reply]

I reversed the colors (Image:Heat eqn.gif). Thanks for the suggestion. Oleg Alexandrov (talk) 20:00, 24 April 2011 (UTC)[reply]

Can you redraw image?

Hi! I see that on 16 July 2007 you did a revision for the colors in the image File:Kepler-first-law.svg . I wonder if you could alter the image to make it more accurate. Right now the foci are drawn way too far apart from each other for the ellipse shown. This can easily be seen by looking at the point at the top of the minor axis: the sum of its distances to the two alleged foci is obviously much greater than twice the semi-major axis. Could you move both foci so that each one is at least twice as close to the center as they are now? Thanks very much! Duoduoduo (talk) 19:08, 22 April 2011 (UTC)[reply]

You are right. I am short of time now, but I'll get to it in a couple of days. Oleg Alexandrov (talk) 17:34, 24 April 2011 (UTC)[reply]

Surprisingly enough, the image is correct. I measured the distances in Inkscape. The equation of an ellipse is

x^{2}/a^{2}+y^{2}/b^{2}=1.

In this picture

a=120,\ b=84.

Then the distance from the center of symmetry of the ellipse to any of the two focii is

{\sqrt {a^{2}-b^{2}}}=86.

What I measured was 84 or so. So all it takes is a little squashing of an ellipse for its focii to diverge a lot. Oleg Alexandrov (talk) 03:39, 28 April 2011 (UTC)[reply]

Thanks for looking into it -- you're right. Now I've used a ruler to measure the sum of distances to the foci from the top of the minor axis, and it does indeed equal twice the semi-major axis. It really doesn't look that way to me -- if you take the segment from the sun to the top of the minor axis, and slide its right endpoint down to the center while sliding its left endpoint to the left, it looks to me like the left endpoint ends up well outside the ellipse; but that's not true. An interesting optical illusion, apparently! Duoduoduo (talk) 14:11, 28 April 2011 (UTC)[reply]

Article rescue: Rexer survey

Hello. I am new to Wikipedia and would like to dispute the notability tag added by User:Melcombe to the Rexer's Annual Data Miner Survey article. Can you help? Thanks. --Luke145 (talk) 20:31, 22 April 2011 (UTC)[reply]

Try posting this at Wikipedia talk:WikiProject Statistics. That way hopefully we'll have a few opinions about whether the article is notable. Oleg Alexandrov (talk) 17:33, 24 April 2011 (UTC)[reply]

Will do. Thanks! Luke145 (talk) 20:03, 6 May 2011 (UTC)[reply]

MathBot April 22

I found an AfD listed that is no longer open. I attempted to remove it manually, and your bot seems to think I am mistaken. CycloneGU (talk) 23:55, 22 April 2011 (UTC)[reply]

(talk page stalker)Actually there's no need to manually update the number of open discussions. Mathbot does that periodically. However, if for some reason you need it done now then you can kick him. (shown as "Refresh the number of open discussions") If after kicking mathbot it still show closed discussions as "open" then we have a problem. --Ron Ritzman (talk) 17:33, 23 April 2011 (UTC)[reply]

Yes, I think this was when I asked it to update as it had been a few hours. It put the closed discussion back out as open per the edit above. I thought it quirky myself. =) CycloneGU (talk) 20:21, 25 April 2011 (UTC)[reply]

The bot was confused by the fact that Wikipedia:Articles for deletion/Log/2011 April 11 had a section which was not an AfD, see "Proposed Compromise involving MLB Rivalry page" in there. I refined a bit bot's search pattern to distinguish this from a true AfD. Parsing html is inherently error-prone, hopefully there could be a better solution at some point. Thanks for the bug report. Oleg Alexandrov (talk) 06:07, 27 April 2011 (UTC)[reply]

Please teach your bot to stop adding redundant disambig links to lists of mathematics articles.

Please teach your bot to stop adding redundant disambig links to lists of mathematics articles. bd2412 T 02:03, 8 May 2011 (UTC)[reply]