User talk:CBM: Difference between revisions

Browse history interactively ← Previous edit Next edit →Content deleted Content addedVisual WikitextInline

Revision as of 11:35, 8 August 2008 editCBM (talk \| contribs)Extended confirmed users, File movers, Pending changes reviewers, Rollbackers55,390 edits →fetch_articles_cats2: r← Previous edit		Revision as of 13:01, 8 August 2008 edit undoMiszaBot III (talk \| contribs)597,462 editsm Archiving 6 thread(s) (older than 2d) to User talk:CBM/Archive 9.Next edit →
Line 6:		Line 6:
	\|archive = User talk:CBM/Archive 9		\|archive = User talk:CBM/Archive 9
	}}		}}

	== August PR cat ==

	Hi Carl - when you get back there's another cat for VeblenBot to watch: ]. '']'' 12:30, 30 July 2008 (UTC)

	== Misplaced Pages:Role of Jimmy Wales has been marked as a policy ==

	{{lw\|Role of Jimmy Wales}} has recently been edited to mark it as a ]. This is an automated notice of the change (]). -- ] (]) 18:49, 2 August 2008 (UTC)

	== Displayname template ==		== Displayname template ==
Line 41:		Line 33:

	::::::::Basically, it boils down to the fact that it's not a good idea to have the displayed title of an article be significantly different than the actual title of the article. It would lead to confusion - people would copy-and-paste the title to make a wikilink to the article only to discover that "]" is really just a redirect to "]". Isn't there a way that you can do your categorization of these articles without changing the article titles? —] <sup>(])</sup> 01:19, 8 August 2008 (UTC)		::::::::Basically, it boils down to the fact that it's not a good idea to have the displayed title of an article be significantly different than the actual title of the article. It would lead to confusion - people would copy-and-paste the title to make a wikilink to the article only to discover that "]" is really just a redirect to "]". Isn't there a way that you can do your categorization of these articles without changing the article titles? —] <sup>(])</sup> 01:19, 8 August 2008 (UTC)

	== {{tl\|Merge-Class}} ==

	Check the template again; needs an extra } for the protection template. My fault. ] (]) 01:45, 3 August 2008 (UTC)

	== musings on an AN comment ==

	''That's a good argument for using the image on the article about Malcolm X, but not so much for using it on four different articles (!) as the image page ] tries to justify. The interest in the picture is not in the particular gun, but in the person who is holding it. — Carl <small>(] · ])</small> 00:14, 3 August 2008 (UTC)''

	:This is a flawed post of yours for a number of reasons. Two of the justifications on that page are distinct from or go further than Georgewilliamherbert's argument. One of them was written by myself for inclusion in an article about a period in which political concerns are a defining characteristic; this image ties into a moderately complex allusion. When we attempt to write articles on aspects of culture from a serious historical and socio-political perspective, we should expect others to at least read our rationales and judge them against the article. Otherwise we should just write facile content-light articles, I guess. So it seems in your comment you have already decided the image use is unjustifiable based merely on the number of articles in which it is used. This is not a good mindset to start from, and it's possible it might poison discussion initiated on this image use.

	:I think the bare minimum reading required for evaluating image use in "my" article is the first two paragraphs of the lead and the img rationale, with the relevant parts of the img talk page also very helpful. My understanding is that no one who has questioned the img use in this case (Bkell, Howcheng, and, very briefly, Ricky81682 so far) has first done this reading, but, conversely, my understanding is that once they have done so, they have let the img use stand. I consider this a testament to both the intelligence and the temperament of the average editor working in this area. But my fear is that sooner or later someone will start from such an entrenched default position that no further considerations will be allowed impact on that. ] (]) 21:26, 3 August 2008 (UTC)

	::The image page claims use in four articles. However: the use on ] isn't needed, since that article can just link to the article on ]; and the use on ] is an exact duplicate of , and so it too can be replace by a text link. Our goal is to keep the use of non-free images to a minimum. If we can link to another article that contains the image, then we don't need to use it a second time. — Carl <small>(] · ])</small> 21:35, 3 August 2008 (UTC)

	:::In terms of chronology, the use on ] is a duplicate of the use on ]. What the img is doing on New school hip hop is specific to that article, as the lead and the article in full hopefully makes clear. ] (]) 21:51, 3 August 2008 (UTC)

	::::As far as I can tell, there is exactly one sentence in the new school hip hop article that refers to that album: "The next album By All Means Necessary (B-Boy, 1988) left that element behind for political radicalism, with the title and cover alluding to Malcolm X.". That text, along with the link to the album article, conveys the information perfectly well even if the image were only on the album article. — Carl <small>(] · ])</small> 21:56, 3 August 2008 (UTC)

	:::::Yeah, there's a little bit about this at img talk, of whatever worth. You're also discounting the caption, which adds a very definite nuance concerning this (self-defense, this ties into other aspects, young black males considering themselves under threat from a white establishment; as I said, this is a moderately complex allusion, it also relates obviously and specifically to to a certain moment in social history), and disconnecting the img use with one of the main concerns of the article—the artists' socio-political commentary as a defining trait, moving from reportage to radicalism to an afrocentricity twinned with inclusiveness. While more critical commentary regarding these specific images would be nice, and I intend to add some, a summary style is called for in an article such as this, and the pictures impart the import of this allusion in a way that, it seems perfectly obvious to me, cannot be adequately imparted by text, even if we wrote at length and tracked down the suitable sources limning the ramifications of this choice of cover art (it's somewhat paradoxical that such critical commentary would alleviate concerns over img use even as it would attempt to diminish the need for it).

	:::::While it may seem to you against the grain to have this image use now in two articles, it is a logical error to conclude therefore that usage in New school hip hop is untenable. The usage relates to the article as a whole. A debate about which of the two articles the imgs should best be restricted to in order to satisfy minimal usage goes against common sense, don't you think? Either the use is justified on its own merits in each article or it is not. And in the case of New school hip hop, this is a textbook non-cosmetic, pedagogical use of non-free media.] (]) 22:34, 3 August 2008 (UTC)

	::::::Just because each use could be justified on its own doesn't mean that both can be justified simultaneously, since the first use can make the second use superfluous, as it does here. — Carl <small>(] · ])</small> 22:40, 3 August 2008 (UTC)

	::::::: And which is the first use that makes the second superfluous here, and, if not a simple matter of editing chronology (in which it would default to New school), how and why do we make that editorial decision? Do we follow an instinct left over from treatments of discography articles? Or do we assume an album article is base camp for all information regardless of context? Do we decide that a political reference is less important in a historical context than in an atomized one?

	:::::::To me this is a strange debate. No, i still think you're disconnecting the reason for the img use from the img use itself. We cannot hope that the reader will select from all the article text the album link, that he will click that link, that he will then pay special attention to the images, that he will then link that back to the section he was reading, and that he will hold in his head the relation to the overall arc of the article. He may do so, but this goes against the pedagogical purpose of an encyclopedia, against the NFCC, against the NFC. ] (]) 23:39, 3 August 2008 (UTC)

	== Scope of math wikiproject ==

	Is cryptography considered outside the scope of the math wikiproject entirely, or where is the line drawn? Best, ] (]) 20:32, 4 August 2008 (UTC)

	:There isn't a firm line. There are lots of things that involve mathematics in some way - ], ]s, ], etc. My personal opinion is that topics that are primarily studied and taught in the context of computer science shouldn't be tagged for the math project. There is some overlap in theory of computation, but things like cryptography, sorting algorithms, etc. are of much more interest to computer scientists than mathematicians. — Carl <small>(] · ])</small> 20:45, 4 August 2008 (UTC)

	::Gotcha. Sounds good. ] (]) 20:50, 4 August 2008 (UTC)

	== WP 1.0 bot v2 ==		== WP 1.0 bot v2 ==
Line 116:		Line 72:
	* I'll be learning Perl on the go, but sure, I'll try. ]<sup>(] - ])</sup> 20:06, 7 August 2008 (UTC)		* I'll be learning Perl on the go, but sure, I'll try. ]<sup>(] - ])</sup> 20:06, 7 August 2008 (UTC)
	*: It doesn't actually seem that complicated if you have a book next to you (which I rushed to get from the local library), but the roadblock for now is getting Apache/Perl to recognize MySQL (apparently support for it is not installed on a vanilla Leopard installation, nor is mod_perl), so I might end up just getting a toolserver account to test this on. ]<sup>(] - ])</sup> 06:36, 8 August 2008 (UTC)		*: It doesn't actually seem that complicated if you have a book next to you (which I rushed to get from the local library), but the roadblock for now is getting Apache/Perl to recognize MySQL (apparently support for it is not installed on a vanilla Leopard installation, nor is mod_perl), so I might end up just getting a toolserver account to test this on. ]<sup>(] - ])</sup> 06:36, 8 August 2008 (UTC)

	==error in placement of remarks on PR page==
	On ], the bot placed its comments before the introductory (opening) comments for the peer review, instead of after them. Please fix this. Thanks. ] (]) 14:20, 5 August 2008 (UTC)

	:Bot comments were misplaced ], too. ] (]) 14:24, 5 August 2008 (UTC)

	::As you can see in this diff , the bot simply replaces a placeholder with its text. The bot cannot control where that placeholder appears on the peer review page. It may be that the "preload" page can be tweaked to put the placeholder below the nominator's text. I'll ask the PR people about it. — Carl <small>(] · ])</small> 14:49, 5 August 2008 (UTC)

	:::I responded at ], but wanted to note that every single peer review has the SAPR before the comments currently. ] ''']''' 16:30, 5 August 2008 (UTC)

	== Invented words ==		== Invented words ==

Revision as of 13:01, 8 August 2008

m:User:CBM

Please leave new comments at the bottom of the page, using the "new section" button at the top of the page.

I will respond on this page unless you request otherwise.

Archives

1	2	3	4	5	6
7	8	9	10	11	12
13	14	15	16	17	18

Displayname template

(Continued from Geometry_guy's talk page)

Sorry to intrude here. The main feature you need (to change the bold text above the article) is already in MediaWiki, as the DISPLAYTITLE magic word. It would be possible to make a template somewhat like {{wrongtitle}} that would take the common name and latin name as parameters and would display them appropriately. At least it wouldn't take any software changes. — Carl (CBM · talk) 23:47, 2 August 2008 (UTC)

Interesting. However, judging from what it says in the manual, it looks to me like DISPLAYTITLE can only be used to change a title's case (i.e. IPod => iPod). Are you sure it would work? --Jwinius (talk) 01:06, 3 August 2008 (UTC)

You're right - it would require a software change to remove the "normalizes to the same title" check. That's not as hard as adding the displaytitle functionality in the first place, though. I need to figure out (or ask around) why that check was added in the first place. — Carl (CBM · talk) 01:30, 3 August 2008 (UTC)

Although I'm not optimistic that this check will be easy to remove, I'll be very interested to hear what you find out. --Jwinius (talk) 01:57, 3 August 2008 (UTC)

Yes, I looked into this yesterday. DISPLAYTITLE doesn't even handle subscripts and superscripts, so for that reason there is code at MediaWiki:Common.js which does. Actually, the code is fairly straightforward and would be easy to change if there were consensus to change it. It is maintained by User:Remember the dot. Geometry guy 10:52, 3 August 2008 (UTC)

Well, let's see what he has to say about this. Cheers, --Jwinius (talk) 11:49, 3 August 2008 (UTC)

Please try to avoid making article titles different from their displayed titles. The JavaScript code we currently use should be used sparingly. Ideally, we would not use it at all because it causes problems for screen readers, text-only browsers, and search engines.

I'm afraid that I don't know what the best solution to your problem is. I personally prefer common naming ("lion" instead of "panthera leo"). Both names are given in the lead section anyway to avoid confusion. Plus, templates like {{Felidae nav}} seem to do a good job of organizing articles already, without the need to move articles to uncommon names.

I know this answer was probably not the one you were hoping for, but all the same I hope you find a good solution to the problem. Please let me know if there's anything else I can do for you. —Remember the dot 03:17, 5 August 2008 (UTC)

That's exactly what I was hoping to do with it: change the titles of articles with scientific names to common names. That way, I figured readers could remain comfortable with the titles while the valid scientific names would remain inextricably linked with the articles, making it much easier for people like myself to maintain large collections of them. The problem is/was that DISPLAYNAME does not allow this.

Your answer was predictable, unfortunately, but what I find most depressing is that the people on the long end of this consensus seem not even to want to consider the possibility that the opposition may have a point that is important enough to warrent a search for a possible solution in this matter. --Jwinius (talk) 23:30, 6 August 2008 (UTC)

Basically, it boils down to the fact that it's not a good idea to have the displayed title of an article be significantly different than the actual title of the article. It would lead to confusion - people would copy-and-paste the title to make a wikilink to the article only to discover that "Lion" is really just a redirect to "Panthera leo". Isn't there a way that you can do your categorization of these articles without changing the article titles? —Remember the dot 01:19, 8 August 2008 (UTC)

WP 1.0 bot v2

Hi, Carl. When do you think we could start coding the new bot? My availability will decrease two weeks from now (the school year starts again), and I'm really interested in working on the bot at least a little bit while I still can. Should we advertise User:WP 1.0 bot/Second generation more widely so we can get more opinions as to what to do? Titoxd 04:37, 5 August 2008 (UTC)

Yes, I think we should advertise is more widely now. I do have some pre-alpha but functional code that can:

Download ratings into a mysql database
Make summary tables from the database
Run simple web-based queries against the database (including intersecting two projects)
Extract the old log information from wiki pages and add it to the database

My goal with that code was to see what issues will actually arise in practice. I'll clean up that code today and put it into my svn repository on toolserver. I think most of the coding work will be in the query interface. — Carl (CBM · talk) 13:37, 5 August 2008 (UTC)

You can download the proof-of-concept code on svn at https://svn.toolserver.org/svnroot/cbm/wp10.2g/alpha — Carl (CBM · talk) 17:48, 5 August 2008 (UTC)

Whee, I'll have to learn Perl. :) That said, from a cursory design-level review, it looks good. What I didn't understand was what $Extra did. And also, wouldn't it be more efficient (from a SQL perspective) to generate a bunch of SQL queries and execute them all at once, instead of executing them one at a time? It would also help to prevent updates to the database failing halfway and leaving the database in an inconsistent state. Now, a small-scale test with limited live data (I as usual volunteer using WP:WPTC's assessments) would be nice. Titoxd 21:09, 6 August 2008 (UTC)

The $Extra var, and get_extra_assessments(), was a successful test of one way to let a wikiproject specify extra rating values. Since then, I changed my mind about how to do that. I am thinking that the project will put a template on their category (Category:Mathematics articles by quality for example) that the script will parse. The template might look like this:

{{WP10params|
|homepage=Misplaced Pages:WikiProject Mathematics
|extra1-name=Bplus
|extra1-type=quality
|extra1-category=Bplus mathematics articles
|extra1-ranking=400
}}

That would tell the bot that about a new quality rating "Bplus" that is used by the project. The template could also be used to track which WikiProjects are task forces of larger projects, or to track other per-project data in a way that can be configured on the wiki.

The way to prevent updates from failing halfway through is to use database transactions. I will look into how to accomplish this with Perl's DBI class. My initial research says that I just have to add the right "start transaction" call at the beginning and a "finish transaction" call at the end of the script. I didn't want to use them yet because I like to kill it halfway through and look at the database manually.

I've been doing some initial testing on my local computer. I should be able to set up a demo on toolserver for you to look at. The web interfaces I wrote are all very very basic. — Carl (CBM · talk) 21:34, 6 August 2008 (UTC)

Here is a live demo:

I think you'll find this query interesting: . — Carl (CBM · talk) 22:28, 6 August 2008 (UTC)

Indeed, that query is interesting. :) One question, though: if we allow B+ to be 400, and another project wants to use B+ and GA, (which is also 400), how will the script know which one to put ahead? Also, if a project thinks that GA should be above A, would they be able to modify their wp10params declaration to set GA to 500 and A to 400? How would we deal with those? Titoxd 06:13, 7 August 2008 (UTC)

I just made up the 400. All the details about project-specific ratings are up for discussion at this point. — Carl (CBM · talk) 12:20, 7 August 2008 (UTC)

I'll be learning Perl on the go, but sure, I'll try. Titoxd 20:06, 7 August 2008 (UTC)
It doesn't actually seem that complicated if you have a book next to you (which I rushed to get from the local library), but the roadblock for now is getting Apache/Perl to recognize MySQL (apparently support for it is not installed on a vanilla Leopard installation, nor is mod_perl), so I might end up just getting a toolserver account to test this on. Titoxd 06:36, 8 August 2008 (UTC)

Invented words

Please verify your comment. And no, a blog fails miserably as a verifiable source. The New York Times does not, in fact, call it's lead section "lede". The term is archaic, and is used by some, and I mean some, Wikipedians to sound, well, snobbish. The word had a meaning at one time for newspapers (let's not even discuss an encyclopedia), but not now, according to several reliable sources, mainly the Oxford English Dictionary. To be honest, of all the problems with Misplaced Pages, snobby editors using "lede" doesn't rank in the top 100, but let's not come across as elitist, especially when we want to be a democratic project. OrangeMarlin 23:21, 6 August 2008 (UTC)

As I commented on the talk page (where you didn't respond...) the Associated Press actively uses the term. — Carl (CBM · talk) 01:12, 7 August 2008 (UTC)

Misplaced Pages:Featured topics/Candidate list

This doesn't strictly work, as the main article name doesn't always match the nomination name. For example, there's one at the moment. What's this page used for, anyway? Seemingly nothing - rst20xx (talk) 16:06, 7 August 2008 (UTC)

Yes, that doesn't work. The solution is to create the redlink as a redirect to the appropriate page. I think that the FTC list is used by some people to watch the lists of FTC nominations without all the text. I made it because of this request — Carl (CBM · talk) 16:19, 7 August 2008 (UTC)

Hmm, that's not ideal at all, as I guess that would have to be done manually. Oh well - rst20xx (talk) 20:30, 7 August 2008 (UTC)

fetch_articles_cats2

Hi Carl. Preparing to migrate away from query.php I took a look at fetch_articles_cats2.pl. I wonder, is there way a simpler way to fetch articles and categories than done there? I tried to adapt it to mathbot, and I got the error:

XML PARSING ERROR 1 $VAR1 = ; Error parsing XML - truncated response?
$VAR1 = 'Died at /home/mathbot/public_html/cgi-bin/wp/modules/Mediawiki/API.pm line 1361. ';

Did you see something of this kind before? Any help with moving from query.php will be very much appreciated. Thanks! Oleg Alexandrov (talk) 03:29, 8 August 2008 (UTC)

Ah, I figured it out. My XML::Simple.pm was out of date. I did not do Perl programming for a while, I start feeling as if I've been living in the woods. :) Oleg Alexandrov (talk) 03:35, 8 August 2008 (UTC)

Once you get the right libraries on your machine, fetch_articles_cats2.pl should be almost a drop-in replacement. I had thought about moving away from XML, but the other encodings that the API can use would also require some library to parse them, and XML was the first one I tried. I think that the version of Mediawiki::API in my svn is more recent than the one on kiwix.

The main issue I have always run into is utf8 encoding. The general principle is that you pass utf8-encoded arguments to Mediawiki::API and it returns native (not encoded) results. But it's a continuing nuisance, especially since api.php was not stable when I was writing my code.

Please do let me know if you run into problems with Mediawiki::API, I will be glad to fix them. — Carl (CBM · talk) 11:35, 8 August 2008 (UTC)