Wikisource talk:Featured texts

Some questions

I have a few questions I'd like to ask. In the section "criteria" I wonder what it means that a work must cite at least one online source. Is this to say that some work must cite another page on the internet that contains the work we have, or that it must link some page on WS itself (like the author page)? I don't think I quite understand this criterion.

Also, this is just a minor one, but will the star and the locked template at all interfere with each other if both are placed on the same page?—Zhaladshar ^(Talk) 18:11, 9 May 2006 (UTC)

That criteria is meant to allow verifiability of our content by citing an external source, since editors will be able to compare our text with the source. I've tweaked it slightly to make it clearer.

Regarding the templates, only one can be used at once. If they are both used, one will be placed on top of the other, thus making the back one nearly or totally invisible. The fact that the work is featured necessarily means that it's also been locked to protect its integrity; perhaps we could add a prominent note to that effect on the Featured text guidelines page for readers who click through. // [admin] Pathoschild (^talk/_map) 19:36, 9 May 2006 (UTC)

Ah, that's what I thought. So, would the featured text template continue to be used after it's been replaced by the succeeding week's featured text? If so, making an explicit note of the implied protection meant by the featured text template would be very beneficial.—Zhaladshar ^(Talk) 19:39, 9 May 2006 (UTC)

I've added a new section, "What is a featured text?", which states that they are protected. // [admin] Pathoschild (^talk/_map) 19:57, 9 May 2006 (UTC)

I've got a few questions:

1)"Featured texts are selected through consensus once per week" - Does this mean that we'll have a vote every week? I think it might be easier & more efficient to just vote on them as they come up & set up a list of featured articles pre-agreed ready to just roll out at the appointed time (as wikipedia does). This would avoid the risk of us running out of featured articles on a slow week (school/university holidays or a week the wikimedia servers are down/slow) or having featured articles created by & voted on by only one user.

2)Shouldn't we create a seperate page for nominating/voting for featured articles rather than doing so on the talk page. I think this would be neater & would avoid cluttering up the talk page with regular votes, archiving would be easier too.

3)"Source: The work must cite at least one external (preferably online) source, either in the {{header}} or on the talk page. Exceptions may be made for rare works with no other suitable online source. " I can see the reasons for this, but just want to make the point that we shouldn't be too strict about non-online works as some of the more obscure/interesting works have been added by wikisourcerers transcribing/scanning from original or re-published sources. I'd say that as long as we can verify a source exists external to wikipedia (& it has been sourced & proofread by a wikisourcerer in good standing) then we shouldn't refuse to allow it to be nominated, or vote against simply because it isn't on-line elsewhere (if we do it takes away one of the main strengths of wikisource - that we have things not available elsewhere on the internet).

4)"Proofread: The work must be completely proofread by multiple editors to ensure that it matches the original as precisely as possible." This is a problem if the source isn't on-line or widely available in written format. Basically what we're saying with this bit is that it must conform to 100% Wikisource:Text quality(incidentally I don't think we even have any 100% text quality articles at the moment). I think this, while a good standard to aim for, is too high a bar & will mean our featured articles are just the same old stuff available all over the internet, assuming we can get enough people bothered to proofread against the original a text a week. AllanHainey 11:55, 15 May 2006 (UTC)

You bring up some good points. I think I agree that weekly featured articles might be too difficult to do. I would be more agreeable to do bi-weekly texts. We need to have time to be able to hunt a work down, format it, proof it, etc., that one week might not give us enough time. This might really bite us when we are all busy for one reason or another.

However, I do think that the texts should conform to the 100% mark. Since we are protecting the texts and not unprotecting them and because we are presenting them to the public, they should have attained that status. The problem won't be having multiple editors proof a page, it will be finding accessible, reliable sources to use in proofreading. In many cases, though, I don't think it should be too hard to find a trustworthy source to use. But I don't think we can waiver on this one, given the status featured texts will have.

About your third point. I, too, agree that "must" might be a bit too strict. It will keep us from adding high quality transcribed texts. For example, the remaining works of Arthur Conan Doyle have no online source (as far as I can find--and I've been looking for a while). That doesn't mean someone can't transcribe them, and that they can't attain a very high quality (even possibly 100%). But even though they have no online source, they should have as much right to be a featured text. However, 95% of all our works I don't think will have a problem finding an online source, what with Google Books, Gutenberg, and a few other quite large databases of texts. But, I think that linking to an online source is a very good concept which should be kept as a strong guideline.—Zhaladshar ^(Talk) 14:46, 15 May 2006 (UTC)

I wasn't saying that we couldn't or shouldn't try to do weekly featured articles, just that we should have a sizeable bank of featured articles checked & agreed beforehand that we can use rather than having to find a new one every week, or fortnight.

I'd say 75% text quality would be fine as these can be protected & in most cases its requirements that

• "the text has been proofread and corrected by one user, who has checked it with a reference edition.
• the text has been imported from a reliable source (e.g. Project Gutenberg, Gallica) and double-checked for formatting and scanning errors. " would be just as reliable & easier to achieve. AllanHainey 15:09, 15 May 2006 (UTC)

I agree with most of the points raised. As the guidelines already state, up to 10 nominations will be kept active as long as their support ratio is above 75%. This would theoretically allow us to fall behind for a full nine weeks before we ran out of texts to feature. I have no problem softening "must" to "should" regarding online sources, although we should strongly encourage these.

I agree with Zhaladshar that having a featured text once per fortnight may be more sustainable. Switching every week may be a little too fast until we have a larger base of established users. I also strongly agree that works must conform to the 100% mark. These works are very prominently displayed as the very best we have available, which should not simply be good enough. If the work is not finished, it shouldn't be protected in the first place.

On the other hand, I disagree with the point concerning multiple proofreaders. Most digitised texts are from optical character recognition software, in which case a scan can be uploaded to Wikisource for proofreading. Manually transcribing texts is likely to result in typos, making it even more important that at least one other editor verify the text. It seems unlikely that someone would manually transcribe a very large text, while short texts can easily be scanned and uploaded. // [admin] Pathoschild (^talk/_map) 23:39, 16 May 2006 (UTC)

I don't think I quite understand what you mean with you comment in the last paragraph. In order to get a text to 100% quality (see Wikisource:Text quality) what is required is that multiple editors check what we currently have against a reliable edition. This is what I meant by multiple proof readers; just have many editors double check our works. I realize many of the problems stem from OCRing mistakes and inadequate double checking before uploading works onto the internet, so it would make sense to have many proofers for each work if we are going to say this is some of our best works.—Zhaladshar ^(Talk) 00:26, 17 May 2006 (UTC)

I agree; I was arguing in favour of having multiple proofreaders. // [admin] Pathoschild (^talk/_map) 00:49, 17 May 2006 (UTC)

Then I completely misunderstood. Darn reading comprehension...—Zhaladshar ^(Talk) 18:33, 18 May 2006 (UTC)

Text quality templates

I'm wondering if the Text quality templates are beneficial, since they are geared towards coordinating efforts on an incomplete text. I don't think the information they provide is relevant to a featured text for the individual reasons outlined below.

Field	Comments
Original edition	This is important information about the text that should be placed on the main page in the {{header}} editorial notes.
Contributors	Although this is useful during collaboration, it becomes harmful afterwards. Because it is limited to major contributors, it exerts ownership of the article—which is usually considered harmful in a wiki environment. if we don't exclude minor contributors, it is redundant with the page history.
[...] by	Same comment as above.
Level of progress	This is unneeded; the progress for any text locked to protect its integrity, in particular featured texts, is assumed to be 100%.
Notes	Editorial notes should be placed on the main page, collaborative notes are no longer needed upon completion.

In my opinion, the source is the only useful piece of information. As such, it should be placed seperately from the text quality template, as is done on Talk:The Time Machine. I'd like to know what others think on this, though. // [admin] Pathoschild (^talk/_map) 23:15, 16 May 2006 (UTC)

I think these templates are incredibly useful & informative. The noted sections provide a lot of supplementary info & organises it well.

Original edition - while this should be noted in the header template there may in some circumstances be additional info noted here which wouldn't be noted in the header such as, translated by X in from original foreign work Y.

Contributors - I think the main use of this is to enable us to identify easily those who originally added or worked on a text (not always possible from history as when we switched to the English language sub-domain we lost all this info). For example this allows us to identify who has access to a rare original source so any queries/discussions can be cleared up. While it does introduce some degree of ownership I feel this is not harmful. I thik it could also help to track down any subtle vandal or history re-writer (eg if user:X is adding fake documents). I think the intention is to include only those who've played a big part in bringing the text up to its noted text quality (eg adding, proofreading)

Level of progress - You miss the point in saying all locked texts need to be 100% (infact they only need to be 75% to be locked), this template isn't just added to texts which are locked but should be added to all texts to enable us to judge their level of completeness/how much they've been checked & verified as accurate. If we don't know the level of quality of a text we don't know whether it needs checking/sourcing/proofreading, etc & we can't get it to the level where it can be protected or used as a featured text (atleast not with any degree of confidence).

Notes - THese aren't always appropriate for the main page, for example the notes for Through the Brazilian Wilderness:
"Reasonably reliable. No preface by TR & no dedication. Some problems in adding to wikisource - Drawings of native American signs can't be correctly rendered in ch 9, noted on talk page. In Appendix 2 The outfit for travelling the tables of provisions don't display right at all (tabulation all wrong & not in tables), don't know enough to fix this. No illustrations are included as they are in the original book. Bartleby had converted "degrees" used in the book to ° & "minutes" converted to '. I've corrected this to match the original book version."

I don't accept that all editorial comments should go on the document page anyway, some on the accuracy/validity of the source itself are better suited to the talk page (as well as technical info on transcription problems, uncertainties in the text, etc). I've raised this on the Scriptorium in relation to Talk:Cast off the Yoke of Bondage but there wasn't a great deal of response.

I can't remember when we agreed on the text quality templates but I think there was a fair bit of discussion at the time and it was agreed they were a useful step in ensuring verifiability & accuracy of our sources.

AllanHainey 07:45, 17 May 2006 (UTC)

Those are details that apply to the collaborative process. Featured texts are the very best we have, and should be entirely complete. At the point a work reaches featured status, there should be no questions remaining about accuracy, transcription problems, or verifiability. 100% should be the absolute minumum level for a featured text; although works can be locked at 75%, they should not be featured at 75%.

I've just created {{featured talk}} (below) to replace the text integrity templates upon reaching featured status.

Featured texts was the featured text for May 2006 (discussion). It is considered among the most complete works available on Wikisource. Since it has been locked from editing to protect its integrity, you should propose any changes on this talk page.

Any notes that remain should be specified seperately, perhaps in a {{messagebox}}. Note that any questions or doubt remaining will prevent the work from being featured, since it has not reached 100%. Most of the examples you've given do not apply to a featured text.

Featured texts was the featured text for May 2006 (discussion). It is considered among the most complete works available on Wikisource. Since it has been locked from editing to protect its integrity, you should propose any changes on this talk page.

This work is based on a physical copy of Billy's "Annoted Book of stories", 2006. See Joe to verify details.

// [admin] Pathoschild (^talk/_map) 15:06, 18 May 2006 (UTC)

Although the best reasons for Template:Textinfo are more applicable to less than 100% quality works, I don't agree that there are any strong reasons to remove it. We hardly need to worry about ownership over texts that should not be edited. The Edition info usually contains more complete info than is in the header. Publisher ISBN etc. There also may be be notes about issues unresolved due to technical limitations that could be overcome in the future.--BirgitteSB 15:35, 18 May 2006 (UTC)

I'm apparently the only user who disagrees with their placement, so I'll go with the otherwise consensus. // [admin] Pathoschild (^talk/_map) 17:39, 18 May 2006 (UTC)

I disagree with their placement on texts, too. I see no purpose for having them, and do not see what they really achieve. Up until the 100% level, they can be useful, but once there, I think they should be removed for utility and aesthetic purposes. What does it matter who proofread or contributed the works once it reaches our level of "perfection"? The source might be a useful bit, but I strongly stress "might." At 100% level, the source should not matter because there should be absolutely no mistakes. The level of progress is a given at this stage: it's super-100% quality (as it is a featured text). Why must we keep them at this point? If we really must have them, they should have their own tab; they should not clutter the article page or the talk page, however. —unsigned comment by Zhaladshar (talk) 18:38, 18 May 2006.

Why do you hate the templates? Is it something a redesign could address? I like having the information more than the template itself. We hardly fill them out in a standard manner, but I always list edition as you would when citing a book with Publisher etc. The source is especially worthwhile when there is an online scan of the book availble. (See Talk:Poems (Donne)) The rest of the info is less important to me, but I see no reason to erase it. The talk pages will hardly be used when an article reaches this point, so I am not worried about them being cluttered. The way I see it if the text really is never to be edited again it doesn't hurt to have this on a page that is not likely to be used and if we make a mistake and we need to do further editing the info would be useful. I do not mind having the info in a different form however (tab, etc.). The only thing with the tab is it would show up on every page, where this template is generally not used on sub-pages etc. Actually I would like to see this or something similar modified for the enclyopedia projects. There could be another field for checking the Wikipedia article and adding any missing information, while removing the field for edition.--BirgitteSB 19:22, 18 May 2006 (UTC)

The edition is useful information, and should be placed on the main page. For example, see The Book of the Damned, whose {{header}} editorial comments state: "This edition is from The Books of Charles Fort, first ed. 1919, sixth printing, February 1957." The template is useful during collaboration, but not upon reaching total completion. // [admin] Pathoschild (^talk/_map) 01:38, 19 May 2006 (UTC)

I don't think that just because we've certified a work as complete we there is any need or benefit in removing the talk page info box template that's already there & either replacing it with something else or expanding the header description (if indeed expansion is needed - as sometimes there will be fuller info in the text info box edition than would be feasible/easy/relevant on the header, eg sub-title of book, ISBN, specific title of individual volume of multi-volume work, translation info, etc) . This seems to me to be pointless additional work which would also remove an element of structure, consistency & transparency from our validation process. AllanHainey 12:43, 19 May 2006 (UTC)

It would not remove any information. I see every single textinfo template on every single featured text talk page looking like this, with only the source being useful:

Information about this edition
Original edition	n/a (see text)
Source	Project Gutenberg
Contributor(s)	n/a (complete)
Level of progress	100%
Notes	n/a (complete)
Proofreaders	n/a (complete)

The template simply isn't useful anymore once the text has reached the pinnacle of completion. Contributors and proofreaders are not experts on the text (with the scarce exception of works copied from a rare text with no online source), and there is no longer any need to note them once the text is featured.

Further, it's not pointless additional work for the majority of works which are currently without the infobox, and upon which proofreaders can easily collaborate without the need of a registry of major editors. // [admin] Pathoschild (^talk/_map) 16:34, 19 May 2006 (UTC)

I think you are oversimplifing things. I must continue to disagree with you about the uselessness of the information. I know that if we adopted your idea and there was a text I was interested in (that I had not worked on myself), I would be acessing the history of the talk page to find this information. At the same time, I understand that some people, perhaps most, would not care one way or the other if they had access to this information. You say that any useful info should be in the header notes. The key problem here is you and I do not have the same threshold for what information is useful. There are things I find useful that I do not want in the header to distract any casual reader. If you would have me keep these things somewhere besides this template I am open to ideas.--BirgitteSB 04:22, 20 May 2006 (UTC)

As a comprimise, would you like me to look into finding a Javascript that will automatically generate a tab when the textinfo template is placed on a page? On this tab would be all the information currently on the template, but it won't be on any page that people frequent. That is, they will have to go out of their way to look for that information. This means only editors interested in looking for this information will find it, and it won't require the drudge work of sifting through the history list.—Zhaladshar ^(Talk) 13:52, 21 May 2006 (UTC)

I don't think a new tab would be necessary, I really don't see what the problem is in keeping this information on the template on the talk page. It isn't going to be used by or affect most users (though it will be useful for some & in some particular circumstances). I really think that either creating a javascript tab or amending the textinfo template as suggested by Pathoschild is unecessary work. While the template info isn't really all that necessary for completed texts there doesn't seem to be any real reason to dispense with it, most folk won't care one way or another whether it's there or not but for a few the information could be useful (& in a lot of cases better there than on the header); most especially the notes info, source & original edition info. The template gives an indication of professionalism & traceability, more so than if all the boxes were filled with n/a, and I think we're really just making work for ourselves - once the text is verified & the infobox updated to reflect that what real reason is there to remove the information which has been noted? AllanHainey 07:20, 22 May 2006 (UTC)

Talk pages should be used for discussion about a particular work. It should not house metadata about that work, and it definitely shouldn't house such data simply because we've become accustomed to it being on the talk page.—Zhaladshar ^(Talk) 14:20, 22 May 2006 (UTC)

I would be happy with be any compromise which keeps the info somewhere. I don't understand exactly what you mean by "automatically generate a tab when the textinfo template is placed on a page", but a technical solution sounds great. That said, I believe the disscussions that take place on the talk page are pretty much "metadata" at Wikisource (since we are not actually creating the content). Perhaps you have something else in mind for the talk pages that I am not understanding.--BirgitteSB 14:50, 22 May 2006 (UTC)

Go to fr:Charte des Nations Unies and you'll notice a tab up top. It is created by a template on the page (you can find it on the bottom of the edit page) which does not show up on the actual article. Now, look at fr:MediaWiki:Monobook.js and you can find the actual script that creates the new tab. I believe I can modify the script so that it will create a link to a page that houses the template with all the information on it without too much trouble. There are some things we will have to clear up before I modify the script, but I want to see how receptive the community is to this.—Zhaladshar ^(Talk) 15:47, 22 May 2006 (UTC)

The template might not be so bad if it didn't take up so much screen space; we could right-align it and implement a Javascript show/hide feature. For an example of this, see the "Confirmed Wikipedians" list at right on w:User:Essjay/Verifications. // [admin] Pathoschild (^talk/_map) 15:24, 23 May 2006 (UTC)

Even better. But where would the template go? The reason I proposed the new tab was to get it off the talk page. Would we still keep it there, just with the option of making it disappear (or I guess have it disappeared by default and making it appear on demand)?—Zhaladshar ^(Talk) 01:30, 24 May 2006 (UTC)

I think placing it on the talk page would be preferable. The tab option would require a new page for each work that has the template. With the Javascript, it would be hidden by default and made visible by clicking the [show] button. We could integrate the overall quality into the heading, so that it would initially be a one-line box stating something like "Text quality information (currently 50%) [show]". // [admin] Pathoschild (^talk/_map) 03:59, 25 May 2006 (UTC)

Sounds good, then!—Zhaladshar ^(Talk) 13:40, 25 May 2006 (UTC)

Featured Text Nominations page

Rather than having votes on this talk page I've created Wikisource:Featured text candidates for this specific purpose. I've copied the Time Machine discussions over to there. I've basically copied the wikipedia featured articles nominations page & edited it but I suspect our procedures for archiving, etc will differ so the page will need to evolve further. I think we might also want to create a template saying "this text has been proposed as a featured text - vote at this link", etc for the talk page of any nominated text. AllanHainey 12:34, 19 May 2006 (UTC)

I'm indifferent to which page we do the nominating on. But I do like adding a template which encourages people to discuss a possible featured text. I think the template should be on the article page (who's going to view the talk page?) instead. After all, it's only a temporary placement.—Zhaladshar ^(Talk) 13:12, 19 May 2006 (UTC)

I thought the talk page as it is likely that the article page could already be protected, though I agree if it can be applied the article page is the best place for it. AllanHainey 13:16, 19 May 2006 (UTC)

The text page should not be protected while it is being discussed, according to the current guidelines. // [admin] Pathoschild (^talk/_map) 16:07, 19 May 2006 (UTC)

I meant pages which were already protected & then nominated for featured texts, I suppose we'd have to request unprotection for them before nominating for featured text status. AllanHainey 07:05, 22 May 2006 (UTC)

Feature intervals

The guidelines currently provide a week for the selection of featured texts, but that may be somewhat too fast-paced for the small community here. As demonstrated by the sluggish progress towards consensus on the featuring of The Time Machine, this is exacerbated by new questions being raised and uncertainty over the guidelines. Perhaps the interval should be increased to two, three, or even four weeks. This could easily be decreased again if we develop a backlog of nominations. // [admin] Pathoschild (^talk/_map) 15:15, 23 May 2006 (UTC)

What about something like a "text of the month"? Wikibooks has a Book of the Month, and I think for this community, something a bit longer than a week (maybe even longer than two, given how busy many of the heavy users here can get) would definitely be better.—Zhaladshar ^(Talk) 01:38, 24 May 2006 (UTC)

That sounds fine, although I think we should keep the tagline 'featured text'. // [admin] Pathoschild (^talk/_map) 12:25, 26 May 2006 (UTC)

A potential stumbling block

I think in our discussions concerning featured text candidates we have come upon a potentially serious problem: getting the text up to 100% text quality. The problem here is finding a work which has a reliable reference edition with which multiple editors can use. I see this as mainly meaning we have to find reliable online editions of a work so that users can coordinate their efforts and can use the same edition when double checking.

The only reliable online editions we could get would be scanned images of pages (since any OCR or transcription could possibly be riddled with errors). Unfortunately, there is a great dearth of this kind of book. I can only think of Google Books as a source which acts in this capacity. The problem is, many PD books in the world are not on GB (well, they are, but we can't view the full book because the edition of the work is later than 1922). This severely limits the number of works we can get off GB to cross-check with what we have. For example, there are no fully viewable books by Jane Austen and very few by Louisa May Alcott and H. G. Wells.

Can anyone else think of a possible solution to how to get works proofread?—Zhaladshar ^(Talk) 13:53, 25 May 2006 (UTC)

I've never seen google books so I can't comment on it. On scanned images of pages another potential problem is that it is difficult to get the actual text off scanned images of pages so in a lot of cases we could be proofreading a text from one source against another source (and of course the editions/printing errors between editions, etc may not match).

I don't think we need to require on-line sources to be the only ones which can be checked/used to proofread as I feel this is unnecessarily limiting wikisource to those populist or POV/cause supporting texts which can be easily & consistently found on the net (& restricting vastly the range of texts we can verify as accurate). I think in a lot of cases it would be possible to verify accuracy & proofread against printed works (although of course this only broadens the scope slightly) as, for example, most libraries carry works by Jane Austen, Louisa May Alcott and H. G. Wells. The main problem here is having 2 users with the same text (& an inclination to proofread a text).

I agree a problem is finding a reliable on-line source to proofread against as Gutenberg & Bartleby both have variations from how works are rendered on their sites & how they appear in the books (quite dramatic variations sometimes). I take the view that if I have a print edition I edit the on-line obtained version to match the printed text (cutting & pasting & editing's a lot easier than manually transcribing) but if all I have is the on-line text I just ensure that it matches it (except for in a few cases where editing is required to remove later editors changes to the original text (eg notations of cheers on speeches or changes of notation of added names, de-americanisation of spelling altered by later editor, etc). Any changes I make or problems with the source I note on the talk page in the notes part of the text info template.

I think that with sources obtained from websites they will mostly be copied & pasted onto here (though its necessary to proofread them too; inorder to correct tabulation, spaces between paragraphs, rendering of text features, foreign letters, etc) & if the original contributor proofreads them (& notes on talk page any issues) it is sufficient to have them at 75% accuracy & protected (I also think we should allow these to be featured texts but that's a separate issue), though unless we want to accept having another user proofread it against the on-line source it was copied from as sufficient to move it up to 100% (which has its problems) then we're frankly unlikely to get a great number of 100% texts verified against a printed work or other on-line source (& then we have no way of knowing whether that on-line source was copied from the original on-line source) & those we do get aren't likely to be too impressive interms of scope or range. AllanHainey 14:44, 25 May 2006 (UTC)

We don't need to require that online sources be the only ones we check against, and I wasn't suggesting it. I'm saying that for proofreading purposes online works are by far the easiest to coordinate with Wikisourcerors, as it might not be likely that our local libraries have the same edition of a text (especially since many of the texts are so popular that every publishing house has been trying to get a share of the market for hundreds of years). And that is why I see a problem: because library contents are entirely unreliable in terms of coordinating the same edition of a text, but online scanned images of pages are also unreliable because of the great lack of scanned literature.—Zhaladshar ^(Talk) 15:11, 25 May 2006 (UTC)

In order to proofread H.G. Wells The Time Machine - I found a second internet source. That source typed it in from a different paper edition. The changes are subtle, but real enough and consistent enough that I know they typed in from different sources. Best example, civilization vs. civilisation. I then copied out the two files and did a file compare. I didn't record all the differences because it got tedious, but samples are on my talk page. I found one mistake in the Wikisource edition & several in the other source. I believe working off of two different sources - where they were obviously typed ins eparately, is one method that can work fairly quickly. Based on this, I upped the quality of text for The Time Machine itself to 100%. (Not sure how many is multiple users???) Banjee ca 02:52, 26 May 2006 (UTC)

I usually proofread works this way by replacing them with text from an alternate source, and previewing the difference in MediaWiki. (Sometimes I need to use regex to eliminate the arbitrary linebreaks that are present in plaintext sources.) That's a major advantage to online sources, since comparing printed texts is much more tedious. If other users adopt this technique, it'll be much easier to achieve the 100% minimum. // [admin] Pathoschild (^talk/_map) 12:23, 26 May 2006 (UTC)

That sounds like a much more practical method. It's not likely I'll ever use it (sorry, but I sort of enjoy scanning lines of printed text and comparing them to online text), but it can be a quite efficient method for boosting texts up in their text quality percentages.—Zhaladshar ^(Talk) 17:25, 26 May 2006 (UTC)

When becoming featured

I think we should automatically protect texts when they are promoted. The candiate process should qualify as a request for protectin IMHO. Also it would be nice if the template on the talk pages of featured texts linked to the archived disscussions.--BirgitteSB 15:07, 25 June 2006 (UTC)

Featured texts are already protected; see the Protection policy's "Preservation of integrity and featured texts". Linking to the discussion would complicate the template usage; perhaps we could link to the archives, where finding a particular discussion is quite easy. // [admin] Pathoschild (^talk/_map) 01:36, 26 June 2006 (UTC)

The protection is fine. When I wrote that I thought the Gettsburg Addres was giving mi the option to protect it. But I must have been mistaken. About the archives it would fairly simple as a parmeter. Sine everything is archived immediately when it becomes featured we know what the link will be. It would not have to be automatic. I have just found that very useful WP, but not a big deal if you are against it.--BirgitteSB 10:32, 26 June 2006 (UTC)

If we standardise the headings to the pagename, all that would be missing would be the archive date. if we specify the month and year, the template could automate the rest:

{{featured talk
 | month = {{subst:CURRENTMONTH}}
 | year  = {{subst:CURRENTYEAR}}
}}

// [admin] Pathoschild (^talk/_map) 21:34, 26 June 2006 (UTC)

That would be great! --BirgitteSB 21:57, 26 June 2006 (UTC)

Done.

Gettysburg Address was the featured text for June 2006 (discussion). it is considered among the most complete works available on Wikisource. Since it has been locked from editing to protect its integrity, you should propose any changes on this talk page.

// [admin] Pathoschild (^talk/_map) 22:48, 26 June 2006 (UTC)

Only one a month?

Why is it that these texts have to compete with each other to get promoted? That just doesn't seem reasonable to me; all texts that meet the criteria should be featured. Its my impression that this is also used to chose the text that will appear on the main page for the next month, but that could easily be chosen by a separate method. If you don't split it in two, you will never have enough featured texts to make it change more frequently than once a month. It just seems counterproductive to identifying the texts that meet the ideal standards for a Wikisource text. Atropos 05:08, 31 October 2007 (UTC)

They don't compete with each other. Whenever a text gets promoted it has the next slot on the mian page that is open. There is no need for competition for the main page, as we do not have enough qualifing texts to run out of slots.--BirgitteSB 16:33, 31 October 2007 (UTC)

You're missing my point. More than one text can be promoted in a month. If you have multiple texts being considered for featured quality and only one of them can be promoted, they will inevitably be pitted against each other. The entire description at the top of the candidates page is clearly of a competition. "Every month the nomination with the highest support ratio ... will be chosen as featured text." "The most promising nominations (up to 10) will be carried over to the next week..." Both of these clearly require texts to compete with eachother. Atropos 01:00, 2 November 2007 (UTC)

That isn't how it really works in practice. Look at Wikisource:Featured text candidates/Archives/2007: Darkness, Lights, and Resignation letter (Roosevelt) were all featured in Feb of 2007 and displayed in subsequent months on the main page. Right now we have not had a text qualify as featured for three months or so. Competition is hardly an issue :p However the page should probably be fixed to be more accurate. Unfortunately the editor who wrote this up this is on a wikibreak right now. I would like to get his opinion on a re-write as he is the most familiar with the actual practice here.--BirgitteSB 20:58, 2 November 2007 (UTC)

Alright. As long as its logical in practice, it doesn't much matter. Atropos 01:47, 7 November 2007 (UTC)

Some additions to the policy

I have some suggestions related to a few changes to featured texts. First, would be to create a secondary page like Featured text/interwiki that shouldn't be protected, and that could be edited by anyone, so that in case the text appears in another language, you wouldn't have to ask the admins. I also though it might be better if the talk page was left blank after a text became featured, and remove all except the {{textinfo}} template, all other discussion being archived. diego_pmc 13:34, 30 May 2008 (UTC)

Protected

I moved the following item from Criteria (preselection) to Procedures;

4. Protected: The work should be locked to protect its integrity after selection. However, it should not be protected during the discussion to allow improvement.

Britannica 1911 article

Can a text, which is a part of another work, but has autonomous value of its own, for example, a Britannica 1911 article, be proposed as a featured text if it fits all the other criteria?

I can think of couple of additional criteria, for example:

• It should not be too short (such as Anthon, Charles).
• If it has internal links, they must be blue and point to other proofread articles.
• If it has parts written in a foreign language, they must be proofread by someone who knows that language.

Any other thoughts? --Amir E. Aharoni (talk) 21:47, 9 May 2010 (UTC)

Source criterion

I think we need to update the "Source" criterion now that works that use ProofreadPage automatically get a "Source" link at the top. Perhaps something like:

Source: The work must either be sourced from a scanned document using the side-by-side proofreading mechanism, or it must cite at least one external (preferably online) source, either in the {{header}} or on the talk page. Exceptions may be made for rare works with no suitable online source.

Thoughts? - Htonl (talk) 02:01, 24 January 2011 (UTC)

 Support. No one comments on this regards. I think now we should move to people encourage to do proofread.Can we change the source section as said above? Jayantanth (talk) 14:34, 1 December 2013 (UTC)

I'm going to be somewhat bold here (as bold as is possible with a three year old proposal anyway) and implement this change. To be honest, this is already a de facto rule; it might as well be de jure. In future, we might want to downgrade the external source requirement to an exceptional circumstance, with only side-by-side proofreading normally acceptable for featured texts, but that's not important right now. - AdamBMorgan (talk) 12:52, 7 January 2014 (UTC)

Proposals

Why is there only one featured text every month? According to me a different text every week would be better :-D Moreover I think that the description is too big and the text itself is too small — the text should be bigger than/as big as the description, because the text is more important, isn't it? And in my opinion the image should be a little bigger.--Erasmo Barresi (talk) 18:43, 24 August 2011 (UTC)

I agree with the second point, the text could be bigger, though clicking through to the work itself resolves that concern. Regarding the third, images increase the size of the main page, as the main entry point it should be as light as possible. And again, a user choosing to view the work will get the image resized to their preference (in theory, unless it is set to 400 or 500 px to suit wide screens). On the first point I strongly agree, I'll add what might be the barriers to this below and hope they are added to and discussed. CYGNIS INSIGNIS 07:49, 25 August 2011 (UTC)

• Not enough validated works.

There are actually quite a few that are; there is some work done on these in POTM, and by users who focus only on this.

• Works are experimental, with regard to scope or formatting.

FTs should reflect widely accepted or unobjectionable practices, not set precedents.

• Users need to be encouraged to put forward candidates.

Many FTs have been nominated by those uninvolved in the creation of the transcript. Proposing one's own contributions is also acceptable.