Wikisource:Scriptorium/Archives/2009-06

Announcements

Votes of confidence

Wild Wolf

Wild Wolf (talk • contribs) • activity • Global who is up for confirmation and qualifies as active per policy has had three people oppose his confirmation referring to his low level of activity as the reason. This activates the requirement for a vote of confidence in his adminship. A simple majority supporting his adminship will see it retained. Please examine the issue before the end of the month.--BirgitteSB 14:09, 6 April 2009 (UTC)

I've alerted the user here. Jude (talk) 01:18, 7 April 2009 (UTC)

Thank you I got sidetracked and then forgot I hadn't done it.--BirgitteSB 23:34, 8 April 2009 (UTC)

When do these votes of confidence end? —Anonymous Dissident^Talk 23:50, 25 April 2009 (UTC)

They're open throughout the month, so editors can drop by to comment whenever they have time. —Pathoschild 03:58:07, 26 April 2009 (UTC)

DarkFalls

DarkFalls' confirmation has accumulated three oppose votes, which is the minimum required by the restricted access policy for removal. His continued access will be decided by a simple majority of established voters. (Since the third vote was placed near the end of the month, the discussion should be kept open a couple of weeks from now to make sure everyone has a chance to chip in their opinion.) —Pathoschild 04:17:03, 26 April 2009 (UTC)

Votes were struck to drop below three editors opposing, so keeping this open longer is not necessary.--BirgitteSB 18:29, 4 May 2009 (UTC)

Dropinitial template changed

I've changed {{dropinitial}} to use the code-fork that was at {{dropinitial-span}}; I've also moved the fork's talk page (there was none in the way).

This is a solid change and presages wider use of spans in lieu of divs to improve the actual markup generated from the wiki-text and template usages. See the talk page for more explanation.

I'd like folks to watch for any issues that may arise from this change that is used on hundred of pages. Please don't be too quick to revert the change as I've seen cases where a local fix is the correct fix. If there is any serious impact that I missed, do revert, of course. Cheers, Jack Merridew 09:08, 7 May 2009 (UTC)

I've bypassed the redirects the uses of the -span version were going via; all usages now use the new implementation directly at its new home. Along the way I did notice a few issues, which I've fixed locally. These issues were preexisting, not something the switch has caused and thus they are likely issues elsewhere. See these for my tweaks;

• [1] [2] [3]

The second two diffs are a bit disconcerting; tables wrapping the text were causing footnotes to be radically mal-positioned. You can see it by loading the old versions; example. I'm wondering of there is some other recent change involved. Let me know if I'm missing anything. Cheers, Jack Merridew 11:42, 7 May 2009 (UTC)

Per a discussion on my talk page, I've undone one of the local tweaks, I gave above; the stricken one. I've also determined that the other two issues only occur at higher resolutions. Cheers, Jack Merridew 14:59, 7 May 2009 (UTC)

{{translations}}

I made a new disambiguation header for pages that list English translations of a foreign language work. Template is {{translations}}. See it in use at Analyse du fruit. At present it categorises into Category:Disambiguation pages by default; I'm not sure if that should be changed to something more appropriate. Hesperian 05:21, 12 May 2009 (UTC)

• I fail to see the benefit of this approach, especially in cases where there is only one translation of the work. A simpler approach is to add a link to the translation, indented under the listing of the work in the original language, on the author page. The original title can then link to the Wikisource of the original language. Thus:
  • fr:Titre original
    • Translation (Smith)
    • Translation (Jones)

Eclecticology - the offended (talk) 08:22, 12 May 2009 (UTC)

One wonders, in this scenario, which English translation fr:Titre original should interwiki to. It seems to me that it ought not to favour one translation over another. I favour having it interwiki to a page that lays out the available English translations.

But the main issue for me is internal linking. I work on papers that often refer to foreign material; see for example Page:Miscellaneousbot01brow.djvu/463. Our ability to cross-link between texts is, as I've said many times before, our basic value proposition compared to Project Gutenberg et al., and we should be making full use of it. So here's a hypothetical for you:

I transcribe an English work that refers to the French work Titre original. The French Wikisource currently does not have the original, but we have two translations, Translation (Smith) and Translation (Jones). How, then, should I link the reference to Titre original? Leave it unlinked?—We would be discarding an opportunity to direct readers to the translations that we host. Link to Translation (Smith)?—I think that misrepresents the source, by implying that the author is actually referring to an English translation. And on what grounds do we favour one translation over another? Link to Translation (Jones)?—Ditto. It seems to me that the best thing is to link to a page that lays out the various English translations available.

I'll plead no contest on cases where there is only one translation. In those cases the above hypothetical is solved by linking to the original title, but redirecting it back to our sole translation.

Hesperian 23:41, 12 May 2009 (UTC)

My impression from what you say is that you are making your own work more difficult than it should be. If "fr:Titre original" does not exist, there is no need for a link to it. Unless we have a very long bibliographic record for the author it is still enough to sort all the possibilities on the author page. Could we even use an anchor to the proper line on the author page. It would be nice to have all the botanical works that you are referencing hosted on Wikisource, but realistically it may be a very very long time before we have them. Having to choose between two translation may be a rare luxury, and perhaps your approach may be more appropriate when we have that luxury. Failing that we should favour the translation that we actually have. Eclecticology - the offended (talk) 08:49, 13 May 2009 (UTC)

I'd prefer linking to a translation/disambig page on the English wikisource when there are no translations or at least two translations, and only to the original language if there are no English translations in the public domain. But I don't think it's necessary to create a disambiguation page if we only have one out of several translations; just link to it directly and add the disambiguation if another translation is added later. --Spangineer^wp (háblame) 15:22, 13 May 2009 (UTC)

We likely have easy accommodation when we host two or more translations. Even there, continuing on the Analyse du fruit example, we need to consider what was available to the author who cited the reference. If only the original language version was available to the author, then that is the version that influenced his writing, not any translation. If that writer's knowledge of the original language was weak his own work could include undocumentable misinterpretations of the original language. If there was a contemporary translation available to the author, that could be what influenced him even though now, with the benefit of 20/20 hindsight, we know that a later translation is far superior. This may not be a big problem with a botanical work such as Analyse du fruit, but in more controversial areas there will be ample arguments. The Bible is an extreme case. The King James translation played a big 17th century role in establishing English literary standards, but linguistic scholarship would suggest that more recent translations may more accurately represent the original language. Even so, we cannot ignore the fact that the King James version has been so frequently quoted over more than three centuries of English literature. Of course, few authors can rival the Bible in matters of controversy, but I have from the beginning seen author pages as the focus to the lens between what we have and what we could have. Eclecticology - the offended (talk) 16:36, 13 May 2009 (UTC)

The Bible is indeed a extreme case; in most cases, the best thing we can do for most of our readers is to point them to the translation or translations we have, contemporary or not. If the reader knows that the author worked from the original language and is interested and capable in working through undocumentable misinterpretations of the original, then I'm sure they'll know where to find it.--Prosfilaes (talk) 19:40, 13 May 2009 (UTC)

I share Spangineer's desire to keep links local rather than cross-wiki where possible, and Eclecticology's concern that sources not be misrepresented by linking foreign language titles to English translations. These concerns can be reconciled by always linking to the foreign language title (either locally or cross-wiki) but sometimes redirecting the foreign language title to the English translation. This has the added bonus that links need not be repaired if we get another translation; we need only convert the redirect to a {{translations}} page. Hesperian 23:51, 13 May 2009 (UTC)

There remains the question of how much of this can be resolved on an author page. Needing to repair links later when we get a second translation is a luxury that we should wish for. Eclecticology - the offended (talk) 15:47, 14 May 2009 (UTC)

Maybe I am missing something, but isn't the traditional means to interwiki still the preferred means? Example being Ionitch.

As we now have a number of Chekhov's works in multiple translations, it would seem that we also need to consider nomenclature. The use of (Author-Translator) works well, hence TITLE (Chekhov-Garnett) cf. (Chekhov-Fell). I am not sure how well it is going to be when we have a translations page, and a disambiguation page, however, we can cross that bridge when we required. -- billinghurst (talk) 11:59, 15 May 2009 (UTC)

I wouldn't make too much of it but I think that (Author/Translator) might work a little better; it avoids the confusion that will arise when any of these people have a compound surname. The Chekhov page is a work in process with much being done by a person who is also able to deal with the Russian text. That puts it into a better position than those where no-one is currently at work. That said, it does raise interesting questions about how the material on the author's page or at Author:Constance Garnett would be best organized. Eclecticology - the offended (talk) 18:25, 15 May 2009 (UTC)

Dmitri does similar, so I will happily amend to your suggestion. I may even find some place to add some notes within the help pages. -- billinghurst (talk) 16:45, 16 May 2009 (UTC)

Proposals

Autopatrolled permission

Once more, I'd like to propose that we request the autopatrolled permission be enabled on Wikisource. I proposed this in November of last year, and it was suggested that User:JVbot/patrol_whitelist was sufficient. While I agree that this is useful and quite flexible for specific projects, in instances where known good users, who are not administrators, are contributing regularly to a wide variety of pages, it gets to the point where patrolling their edits seems to be a pointless exercise. This would allow administrators to give known good users the autopatrolled permission, and obviously, users would be able request this permission from the community.

1. Support. Jude (talk) 03:01, 31 March 2009 (UTC)
2. Support, so long as we create no bureaucracy for this. If an admin can vouch for an established user, there's no problem autopatrolling without further policy or voting. —Pathoschild 03:17:29, 31 March 2009 (UTC)
3. Support. fwiw, Wikisource:Scriptorium/Archives/2008-11#Autopatrolled is the earlier discussion. I dont object to this being done automatically by the software. Bots need sleep too. Even if created without any bureaucracy, I suspect it will inevitably appear. But not yet. We can deal with that problem when we are looking at it. Right now, we have a good many users who are not diverse enough to be admins, but are very trusted to work meticulously within their chosen area. John Vandenberg ^(chat) 04:01, 31 March 2009 (UTC)
4. Support. Maybe all we need is a page somewhere listing the users who have been given this permission, when they got it, and which admin vouched for them? Or not even that? --Spangineer^wp (háblame) 04:07, 31 March 2009 (UTC)

   Special:Listusers for that group, and Special:Log for that user would have this information, I believe. Jude (talk) 04:13, 31 March 2009 (UTC)

5. Support. Sounds like a great idea. You have my support. --Mattwj2002 (talk) 05:48, 31 March 2009 (UTC)

• Comment/question are we talking mw:Extension:Makepatrol? -- billinghurst (talk) 06:04, 31 March 2009 (UTC)

  Either that or simply allowing sysops access to Special:UserRights, though only the ability to assign and revoke the permission. However the devs implement it these days. Jude (talk) 06:45, 31 March 2009 (UTC)

Update: This permission has been enabled, however, it is currently on usable by 'crats. I've commented on the bug to ask to allow sysops to assign and remove the group, though I'm not sure what the ETA on that is. Jude (talk) 09:47, 8 April 2009 (UTC)

Implemented, as per this bug. Administrators may now assign and revoke this permission. Alterations may be viewed at this log, and a complete list of people with this right may be viewed at this log. Jude (talk) 23:11, 15 April 2009 (UTC)

Nice work Jude, thanks for steering this through the process. -- billinghurst (talk) 00:16, 16 April 2009 (UTC)

• Perhaps we should erect a section on requests for Adminship where trusted users may request to have the autopatrolled permission given to them? It'd formalise the process and ensure that an archive were kept of all assignations. —Anonymous Dissident^Talk 01:14, 16 April 2009 (UTC)

The consensus above seems to be a system without bureaucracy, and I must admit, it's not something that I imagine requires that much bureaucracy. There are not that many people active and that much changes of user rights that the rights log wouldn't be useful in identifying who assigned what and when, and the user list should suitably list those who have the autopatrolled permission.

Further to that, there is no special features granted access to by having this permission, and I imagine it'd work best as an ad-hoc system where any administrator doing patrolling on recent changes can see a trusted user and assign them the permission. In fact, I think bureaucracy and having to go through a vote to achieve consensus would make the permission pointless and little-used by anyone.

Finally, I'm not averse to a section where trusted users can request the permission and an administrator can review their request and, if they deem it valid, grant the permission, but I'm not sure that it would ever be used nor would it really be required, unless of course recent changes patrollers who are not administrators can request that other people be assigned this permission. Jude (talk) 02:18, 16 April 2009 (UTC)

As our criteria for who can patrol is not extensive, having a rigid autopatrol criteria sounds of less benefit, so maybe we can establish some criteria for assignation of patrolling, and then possibly modify. Presumably the sysop will be comfortable that the person has

• an extended edit history (do we set a number?)
• across a diverse number of topics
• demonstated knowledge of Help:Patrolling

Note: if users are working solely on singular works, we probably prefer their addition to the whitelist

I don't think that the user will request to have the status applied as they are less aware of patrolling, and its implications. The only people who physically benefit from autopatrolling are those who are actually patrolling, which is the more involved in the process. -- billinghurst (talk) 04:09, 16 April 2009 (UTC)

I thought the patrolling right was assigned to the autoconfirmed group? There's no requirement for patrolling beyond being autoconfirmed, as far as I can see. Jude (talk) 11:29, 16 April 2009 (UTC)

• I should say that I never meant for anything bureaucratic. Something where a request is made and a neutral administrator makes an informed judgement call. —Anonymous Dissident^Talk 14:59, 17 May 2009 (UTC)

Wikimania 2010

Please start your proposals at http://wikimania2010.wikimedia.org now. For Wikimania 2010, --Wikimania2010-roleaccount (talk) 10:38, 22 May 2009 (UTC)

Questions

Exporting as pdf

The "PDF version" link only gives the pdf of the current page as far as I can tell, so how does one get a pdf version of an entire book from wikisource? http://en.wikisource.org/wiki/Beyond_Good_and_Evil is the specific book I was looking at. Kinkywhynot (talk) 00:40, 17 May 2009 (UTC)

One builds a book of the wanted pages, and then one can download as PDF with the Book tool. The tool is still a work in progress, however, I have constructed User:Billinghurst/Books/Beyond Good and Evil -- billinghurst (talk) 01:55, 17 May 2009 (UTC)

awesome, thanks Kinkywhynot (talk) 01:17, 22 May 2009 (UTC)

Use of <ref> and <ref name="X">

I used <ref name="X"> in a work, and it just blanks the ref tag, and definitely doesn't use the tag when later called as <ref name="X" />. What is different from our rendition of TAG and en.wp's rendition? I would hope that such a basic tag would be identical in its use between wikis. -- billinghurst (talk) 14:22, 7 April 2009 (UTC)

Test^[1] What about now?^[1]

1. Does this work?

Works for me. Angr 18:08, 7 April 2009 (UTC)

Yep, it seems that with <ref> if you give the name as a number it fails, eg. name="4" gives a nada. Seeing that I am doing tidying duties, I wasn't going for inventive names. <shrug> -- billinghurst (talk) 21:35, 8 April 2009 (UTC)

The name attribute maps to an 'ID' attribute and there is a level of MediaWiki manipulation that occurs here. Basically an ID should start with a letter and does not allow many funny characters; no spaces, apostrophes or anything too weird; and they should be unique on a page. MW does encode some, but that makes for horrid names. If at all possible, folks should just use a simple name. Cheers, Jack Merridew 09:00, 10 May 2009 (UTC)

Should I publish my Master's in Mathematics here?

Are you even interested in it? Do people do that? 38.100.141.136 09:44, 19 April 2009 (UTC)

Normally no. The first criterion for acceptance is that it must have been previously published elsewhere. This criterion saves us from having to evaluate your thesis in ways that are beyond our competence. There are other criteria, which may or may not be more onerous, but these are moot if you can't deal with the first. Eclecticology - the offended (talk) 21:36, 19 April 2009 (UTC)

OK, that is less feedback than I'd expected. 38.100.141.135 21:31, 19 April 2009 (UTC)

Our criteria for what should be published is at WS:IO, see whether you think it falls within that. In general the answer would be no, andyou may find that Wikibooks:Main Page is more for what you are looking. -- billinghurst (talk) 22:57, 19 April 2009 (UTC)

Contrary to the other two, I would say that Masters' Theses fit within our inclusion criteria, and would be welcomed here. They meet the definition of a published, peer-reviewed work. Sherurcij ^{Collaboration of the Week: Author:Carl Jung.} 19:22, 23 April 2009 (UTC)

I'm not sure they're technically published, but I wouldn't object to thesis from accredited universities being published here.--Prosfilaes (talk) 22:52, 12 May 2009 (UTC)

If masters' theses are to be accepted they must at least have resulted in a degree. The last thing we would want is original research in a thesis that was rejected by the university. They must be verifiable and available for any other individual to independently verify its contents without needing to contact the author of the thesis. This could be by deposit in the library of the degree-granting university, so that someone can find the thesis by looking in the library's catalogue. Eclecticology - the offended (talk) 07:48, 13 May 2009 (UTC)

Warning: Template include size is too large. Some templates will not be included.

Can anyone assist me in diagnosing the foregoing error message? It appeared when I saved this edit to Copyright Law Revision (Senate Report No. 94-473). Looking at the saved page, it appears that everything after p. 130 is cut off, indicated only with redlinks labeled “Page.” Is this a {{Page}} problem, or a problem in the underlying pages that are being transcluded together, or something else entirely? I’m puzzled. Tarmstro99 (talk) 01:21, 23 April 2009 (UTC)

You've maxed out your "post-expand include size"; see w:Wikipedia:Template limits. Hesperian 03:29, 23 April 2009 (UTC)

Thanks. Any ideas why {{Page}} would appear to be so much more wasteful than #lst?

• Copyright Law Revision (House Report No. 94-1476) transcludes 368 pages without using {{Page}}, and does not come anywhere close to hitting the 2MB post-include size limit.
• Copyright Law Revision (Senate Report No. 94-473) uses {{Page}} and hits the post-include size limit after transcluding only 130 pages.

Why the big difference? I have not dug deeply into the template code, but my ill-informed impression was that {{Page}} was basically a wrapper around #lst, with friendlier syntax. Tarmstro99 (talk) 00:32, 24 April 2009 (UTC)

Due to a known glitch in the MediaWiki parser, the size of transcluded content is counted repeatedly for each transclusion level, including ParserFunctions. This is an example of how a single page is counted towards the 2MB limit:

• Page: 277 bytes
• Page in template code: 9990 bytes
• Page in template code in article: 13750 bytes

Thus, using the {{Page}} template counts the original page size 61 times towards the limit. The solution is simplification: minimize the number of times the text is included, and minimize the number of ParserFunctions involved. For example, "{{#if:page text|page text}}" more than doubles the page size. —Pathoschild 09:03:57, 24 April 2009 (UTC)

But #lst is a ParserFunction, too, is it not? I’m afraid I’m still not grasping why using {{Page}} results in hitting the post-expand include size so much faster than using #lst. (It is entirely possible that there is no explanation for this phenomenon that is both (1) accurate and (2) comprehensible to me.) Tarmstro99 (talk) 00:26, 25 April 2009 (UTC)

I may be mistaken, I think Pathoschild may be referring to the code of {{Page}} where there are many Page calls with(in) IF statements. billinghurst (talk)

Whenever anything is retrieved from the database (or cache) to be transcluded, its length is added to the post-expand size. This includes arguments to parser functions; the double-transcluding #if Pathoschild mentions above isn't a glitch; that is the fact that even branches that shouldn't expand, i.e. {{#if:|page text}} would still count is. #lst works at a lower level; it gets the page from the DB, then cuts out the desired section, and caches that, so only that section counts towards post expand size. -Steve Sanbeg (talk) 17:30, 28 April 2009 (UTC)

But contrary to what is being said here, the page template does not and never has contained any if statements with transclusions in the test condition; and in fact it contains very few transclusions at all. Most of those hideous brace sequences are parameter invocations not page transclusions. I don't know where Pathoschild got the figure of 66 from, but if it is correct then I am just as flabbergasted as Tarmstro99. Hesperian 06:11, 4 May 2009 (UTC)

It doesn't matter which part of the #if it's in. I don't think the 66 figure is right, but the part where you're multiply charged for each transclusion for each level of transclusion looks correct. After your change to the template (unaffected by your change) each call to the template will include 2 full transclusions (one for the #section-x, and one for the default in the #if) in addition to the desired section, so you'd expect the post-include size from using the template to be double that of a direct transclusion; but I did see that it was a 4 times, like Pathoschild said. I've since simplified it to remove the extra transclusion. With this fix, if you want the full page, you get one transclusion of it (with a 2x penalty per Pathoschild), while if you ask for a section, you should only transclude that section, so the size shouldn't depend on the page size. My test of a full page transclusion verifies that this drops from a 4x penalty to 2x, but it should be better with section transclusion. -Steve Sanbeg (talk) 19:05, 4 May 2009 (UTC)

Deprecated components

The only work that uses the deprecated form of PAGE is Jane Eyre, is it now worth looking to update that work to the more modern form? That would allow for a minor simplification of {{Page}}. -- billinghurst (talk) 05:35, 25 April 2009 (UTC)

Yes; I've removed the old form from the smattering of other pages that still used it. It would be good to finish that and simplify the template. -Steve Sanbeg (talk) 17:30, 28 April 2009 (UTC)

All main namespace pages that use {{NewSection}} are now converted to use <section> so you should be able to make any further update to {{page}} without that being an inconvenience. -- billinghurst (talk) 07:06, 29 April 2009 (UTC)

[4] Hesperian 04:45, 4 May 2009 (UTC) plus page/doc page amended to remove mention, and Newsection template deprecated. -- billinghurst (talk) 05:18, 4 May 2009 (UTC)

Wikisource Books, a place to collect books created with the book tool.

Hi guys, I just thought I would bring to everyone's attention that we have a new place for books created with the new book tool. I think that WS:B should only have proofread and featured content. Please help with this section. Hopefully, we can get these bugs with the book tool fixed shortly. :) --Mattwj2002 (talk) 03:13, 28 April 2009 (UTC)

Here is what is being built elsewhere w:Wikipedia:Books. I like the idea, and I think that restricting it to featured and fully validated books (at least initially) would be wise. -- billinghurst (talk) 05:19, 28 April 2009 (UTC)

I hate to be a downer, but when I download the pdf for Wind in the Willows, I get a fourteen page document—presumably because the books extension doesn't understand the page template. Unless I'm doing something wrong, this seems to be a dealbreaker for many of our validated texts, until this bug gets fixed. --Spangineer^wp (háblame) 22:41, 28 April 2009 (UTC)

Absolutely. It actually seems sporadic, as we created Equitation and went through it the other night, and it didn't have the [Page] rendering that it generated today. Weird. The print preview ORDER BOOK copy is fine, going out to 44 pages. We can turn off the PDF from the template, though it would be nice to determine whether it is a load issue or otherwise for the PDF generated version. -- billinghurst (talk) 00:59, 29 April 2009 (UTC)

The name "books" is too confusing, since Wikisource is already all about books (well, and journals). Maybe call it "printable volumes" or something. --LA2 (talk) 05:43, 29 April 2009 (UTC)

I created Wikisource:Books/A specimen of the botany of New Holland, and the PDF doesn't contain a single page! I must say I am disinclined to invest any further energy in this until the bugs are ironed out. Hesperian 04:15, 4 May 2009 (UTC)

I'm personally unlikely to use this tool since it would be easier to just go to A specimen of the botany of New Holland if I were interested in this book. The proof of any idea is that people use it beyond the few days after its introduction. Theoretical benefits will just get people to say "That's nice" before they carry on with whatever they were doing beforehand. Eclecticology - the offended (talk) 08:54, 4 May 2009 (UTC)

I guess the advantage I saw is that it would be easily printable. You may counter that people can always print the scan, but in this case the scan is horrid.[1][2]. Hesperian 01:58, 7 May 2009 (UTC)

After creating a book for a user, there are some thoughts about how we might be able to get the product refined.

• ability to automagically include subpages to a construction
• allow a chapter to start on a new page, rather than to follow on

-- billinghurst (talk) 02:03, 17 May 2009 (UTC)

Searching a category

Is it possible to do a full text search but only in a certain category?Ferrylodge (talk) 19:00, 29 April 2009 (UTC)

Unfortunately no. --Spangineer^wp (háblame) 04:26, 4 May 2009 (UTC)

Too bad. Thanks for the info. It seems like it could be very useful.Ferrylodge (talk) 00:39, 7 May 2009 (UTC)

One option would be to use your browser's Find feature. It'd be limited to just titles (and only to 200 items at a time), but it's better than nothing. EVula // talk // ☯ // 05:44, 8 May 2009 (UTC)

I've wished for this too. I'm assuming from the above that adding Category:Name doesn't work, although you can at least, I think, add that name to refine the search. Again, not an ideal solution. Cygnis insignis (talk) 11:32, 8 May 2009 (UTC)

Nice idea. I played around with it, and found that if you search the category name only, and put it in quotes, it will return items from that category—so search "1850 works" + whatever the text is that you're looking for. But text transcluded from the page namespace won't be found. Overall, the search function needs some work. --Spangineer^wp (háblame) 13:08, 8 May 2009 (UTC)

Adding the Popular Science Monthly template to each issue

Hi guys,

I was wondering what you guys thought about adding the Popular Science Monthly template to the main page of each issue? It would make for easy navigation. Please let me know what you think. --Mattwj2002 (talk) 09:35, 2 May 2009 (UTC)

Just to clarify, I am asking about adding the template to the indexes for each issue. --Mattwj2002 (talk) 09:40, 2 May 2009 (UTC)

Sounds like a great idea. Suicidalhamster (talk) 14:06, 2 May 2009 (UTC)

Yes, very useful and pretty at the same time :) --Zyephyrus (talk) 16:15, 2 May 2009 (UTC)

Featured text

We're five days into May and still running April's featured text. Can someone experienced in FTC procedure please choose a new FT? Hesperian 03:45, 4 May 2009 (UTC)

Updated. I think I did it correctly. --Spangineer^wp (háblame) 04:25, 4 May 2009 (UTC)

Ta. Hesperian 04:47, 4 May 2009 (UTC)

Knew I forgot to do something! I glanced at it earlier this week and was going to update it, but then got distracted with something and completely forgot. There are quite a few text candidates on WS:FTC that have only one or two comments, so if anyone has the time or the inclination, remember to drop by and comment on texts and point out problems etc. Jude (talk) 07:14, 4 May 2009 (UTC)

Good fonts for proofing

Since I spend so much time here comparing the OCR text in the edit box with the text in the page image, I decided it would be a big help if I were using a font in which characters are very easily distinguished; i.e. a font in which typical OCR errors stick out like the proverbial. Obviously I can't change the font in the page image, but I do have control over the font in the edit box.

On looking into this issue, I found that there are standard OCR fonts, of which one, OCR-B, is designed to be optimally human-readable. I found a free OCR-B font, installed it, and set my browser to prefer it in monospace contexts. (In Firefox: Tools > Options > Fonts & Colors > Advanced) It was pretty quick and easy to do.

I haven't yet driven it for long enough to form a firm opinion; but I am disappointed to find that it doesn't make it any easier to distinguish between ndashes and mdashes, and makes it harder to distinguish apostrophes from backticks. It is very good on I versus l though.

Does anyone have any advice on this?

Hesperian 02:15, 5 May 2009 (UTC)

w:Distributed Proofreaders has a font called DPCustomMono2, designed to make OCR errors clearer. It's engineered for a Latin-1 character set, though, so while it makes it easier to distinguish between apostrophes and backticks, I don't think it helps with en-dashes and em-dashes.--Prosfilaes (talk) 02:46, 5 May 2009 (UTC)

Easy solution to ndash/mdash problem is to just type them out all the time, rather than hoping that the one you're looking at is what you want. DP's OCR font is definitely nice for proofing, but I wonder if there's a better way to implement it—it'd be great if there were a way to make it appear only when editing in the Page namespace... could something like that be done via a user's monobook.js/css? --Spangineer^wp (háblame) 03:47, 5 May 2009 (UTC)

Typing them out all the time is exactly what I do. I don't call that an "easy solution". But I am glad to learn I am not the only person here perverse enough to do so.

I wonder how hard it is to manually tweak a typeface.... —unsigned comment by Hesperian (talk) .

In theory, all you should have to do is add a:

.pagetext {
   font-family: XYV;
   etc...
}

To your [[User:<myname>/<mystyle>.css]] (where the default is monobook). I've just tested this out by playing with the font size, and it seems to work fine for me. Jude (talk) 06:12, 5 May 2009 (UTC)

"pagetext"? I assume that applies the font to the entire page, right? If you only want to change the font in the edit box, then maybe

.textarea {
   etc...
}

This is just speculation; I'm about to rush off and don't have time to check it myself. Hesperian 06:30, 5 May 2009 (UTC)

Neither seems to work. Any suggestions, again, for making only the edit box on only the page namespace appear in a different font? Or is it a pipe dream... --Spangineer^wp (háblame) 01:23, 6 May 2009 (UTC)

My mistake, I wasn't looking at the edit window, rather the actual Page window. In that case, you might need to define a separate style for the edit window, and the use a JavaScript (if namespace==Page, get the edit box and change the .class value). Shouldn't be all that difficult to do, either. Jude (talk) 04:12, 6 May 2009 (UTC)

DjVu page numbering

The pages of my DjVus created in DjVu Solo are always numbered from 1 to number of pages. Is there any way to correct this, if this does not match the numbering of the original book? For example, some books also count the white pages at the start and the inner title page, while others start the numbering with the first page of real content. Some books have prefaces numbered with Latin letters and the regular numbering starts with 1 after the preface. And some books have illustrative plates that have a separate numbering from the main content of the book. Is there any way to bring the numbering of the DjVu in line with the original numbering? --Slomox (talk) 23:52, 7 May 2009 (UTC)

You can't change the numbering in the DjVu itself; and therefore you cannot avoid beginning your transcription here at Page:File.djvu/1.

However, you can change the numbering in the list on the index page, using arguments to <pagelist />. I don't know where the documentation is but if you have a look at the source of Index:Diary of ten years.djvu I'm sure you can work it out.

Summary of Pagelist -- billinghurst (talk) 00:15, 8 May 2009 (UTC)

Also, if you use {{page}} to transclude pages, you can pass it whatever you want in the |num= parameter. Hesperian 00:05, 8 May 2009 (UTC)

You can see how pagelist is used here and how the num= parameter is used here. --Zyephyrus (talk) 05:17, 8 May 2009 (UTC)

a template to transclude a series of pages

Template:PageNum - ThomasV (talk) 20:58, 9 May 2009 (UTC)

I've wanted such a thing, I tried it out here. Cygnis insignis (talk) 13:10, 11 May 2009 (UTC)

New page status for incomplete pages

I have added a 5th status for pages 'without text' : it will be used for blank pages, pages with an image, or with text that we do not want to transcribe. Possibly, its associated colour should be gray.

Some people have expressed interest in adding another state for 'incomplete' pages : That is, pages where text transcription is unfinished. If possible, I would like to get more feedback from the community about this. Please participate in the discussion at oldwikisource:Wikisource:ProofreadPage ThomasV (talk) 16:03, 10 May 2009 (UTC)

A poem annotated by the poet

How do I handle a poem that includes abundant footnotes by the poet, and which are an essential part of the text? They are indicated with numerals that reset to "1" on each page, and were printed at the bottom of each page, occasionally overflowing to a facing page. — Alarob (talk) 16:45, 15 May 2009 (UTC)

From what you say, I am assuming that the original author did not start renumbering footnotes on each page (Some do.), but that there was continuous numbering over the entire work. Can you simply hard-number the footnotes using {{ref|}} and {{note|}} in combination. Eclecticology (talk) 18:37, 15 May 2009 (UTC)

You're right, there is continuous numbering, except that the first note is numbered "*" instead of "1". I'm new to this, so am wondering whether I can avoid shoving all footnotes to the end of the poem. I suppose subpages are the answer. — Alarob (talk) 19:52, 15 May 2009 (UTC)

You could use the standard footnotes extension, i.e. place each footnote in a <ref> tag, and use <references/> where you want the footnotes to appear. That would avoid putting them all at the bottom of the page, but it would also reset the numbering to 1 after each call to <references/> so that should be fine just to get the footnote text in the page, but the extension is difficult if you need to emulate a specific numbering. -Steve Sanbeg (talk) 20:37, 15 May 2009 (UTC)

I personally don't consider it to be a problem to put all the footnotes at the end of a long document—remember that links are built in when you use the <ref> tag, both from the text to the footnote and from the footnote to the text. That solves the problem of figuring out where to artificially divide the work to add the references tag. --Spangineer^wp (háblame) 21:18, 15 May 2009 (UTC)

Inter-language bots?

Hello everyone.

I was under the impression that if you added a language link on one language wiki, a bot would run around making sure that all languages had links to all other languages. I tried this with Author:Cato the Elder, adding the la: link, but no bots seem to have picked up on the fact that there is an fr: link on the la: page. Should Wikisource not have inter-language bots like w: does? It Is Me Here ^{t / c} 12:07, 16 May 2009 (UTC)

For that to work the system would need to know that Author:Cato the Elder is la:Scriptor:Marcus Porcius Cato, and fr:Caton l'Ancien (without an author namespace) in French. Without standardized names across languages the bots can't be expected to pick this up. Eclecticology - the offended (talk) 16:32, 16 May 2009 (UTC)

It would pick it up the same way it does in other Wikimedia Wikis; if an author page on en links to one on la, and the one on la links to fr, the author page on en should also link to fr. It's more clumsy here, as work pages can get complicated, but author pages should be simple, if someone wants to pick it up.--Prosfilaes (talk) 03:31, 17 May 2009 (UTC)

Transclusions limit

Take a look at this page: Gesenius' Hebrew Grammar/Second Part/Chapter 2/The Strong Verb.

Why is it broken at the bottom? Did it hit the high limit of transclusions?

I'll gladly split it into several pages - i am just curious about the reason for the breakage.

Thanks in advance. --Amir E. Aharoni (talk) 15:27, 22 May 2009 (UTC)

Yes, you can see that if you look at the comments in the resulting html:

Post-expand include size: 2048000/2048000 bytes

As described previously, text included with the page template is charged double a normal inclusion; the template used here would double that again, so all the transcluded text would be counted 4 times towards the limit. -Steve Sanbeg (talk) 16:52, 22 May 2009 (UTC)

Thanks! --Amir E. Aharoni (talk) 17:22, 22 May 2009 (UTC)

Qur'an links

I was just proofreading History of What Befel the Fowl-let with the Fowler#p.156. In the footnote for that page I had to make a link to the Qur'an: The Holy Qur'an/Adh-Dhariyat, being Chapter 51 of the Qur'an. To find this chapter {{Qur'an chapters}} is of little help because it only gives chapter names. The link to The Holy Qur'an in the header asn't very helpful either because it redirected to the page Qur'an which itself is more a general list of different editions of the book. Furthermore, the "Adh-Dhariyat" page shows the translator as Abdullah Yusuf Ali, but only some of the pages linked from the template do this.

I have restored an older version of The Holy Qur'an to make things work. Before it was converted to a redirect the page had had "incomplete" tags (probably no longer applicable), and a copyvio notice. If the latter applies, it should relate to the various chapter pages, none of which were deleted. I am not, and do not want to be a student of the Qur'an. That said, it would be good to know the best place to go when one wants to link to a referenced passage of the Qur'an. Eclecticology - the offended (talk) 21:00, 23 May 2009 (UTC)

New index page

Based on a suggestion from Jayvdb (talk • contribs) at Template talk:New texts, I went and created a new index page for Health-related documents, at Wikisource:Health. Check it out, any help with expansion, organization, etc., would be most appreciated. :) Cirt (talk) 06:57, 27 May 2009 (UTC)

Wouldn't "Medicine" be better as a general topic? Library of Congress Subject Headings treats Health as a sub-topic of Medicine used for "works on optimal physical, mental, and social well-being, as well as how to achieve and preserve it." Eclecticology - the offended (talk) 08:40, 27 May 2009 (UTC)

That page already exists, so I changed the index to point there for more general stuff [5]. Cirt (talk) 08:54, 27 May 2009 (UTC)