Wikisource:Scriptorium/Archives/2021-03

Please do not post any new comments on this page. This is a discussion archive first created in March 2021, although the comments contained were likely posted before and after this date. See current discussion or the archives index.

Links to another wiki, in another language

Is there a method to create links to another wiki, such as French Wikipedia, Wikionary, etc., etc.? AnotherEditor144 ^{t - c} 18:02, 1 March 2021 (UTC)

@AnotherEditor144: w:fr:Main Page, wikt:ontology, c:COM:VP/T. See w:Help:Interwiki linking for some guidance. But please do not add such links to transcribed works here (except the occasional Wiktionary link). See WS:Links and WS:Annotations. --Xover (talk) 18:52, 1 March 2021 (UTC)

@Xover: Ok. I was planning to put it on my user page anyway. AnotherEditor144 ^{t - c} 19:18, 1 March 2021 (UTC)

Tech News: 2021-09

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Recent changes

• Wikis using the Growth team tools can now show the name of a newcomer's mentor anywhere through a magic word. This can be used for welcome messages or userboxes.
• A new version of the VideoCutTool is now available. It enables cropping, trimming, audio disabling, and rotating video content. It is being created as part of the developer outreach programs.

Problems

• There was a problem with the job queue. This meant some functions did not save changes and mass messages were delayed. This did not affect wiki edits. [1]
• Some editors may not be logged in to their accounts automatically in the latest versions of Firefox and Safari. [2]

Changes later this week

• The new version of MediaWiki will be on test wikis and MediaWiki.org from 2 March. It will be on non-Wikipedia wikis and some Wikipedias from 3 March. It will be on all wikis from 4 March (calendar).

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

---

19:08, 1 March 2021 (UTC)

Wikifunctions logo contest

Hello. Please help to choose a design concept for the logo of the new Wikifunctions wiki. Voting starts today and will be open for 2 weeks. If you would like to participate, then please learn more and vote now at Meta-Wiki. Thank you! --Quiddity (WMF)

01:49, 2 March 2021 (UTC)

pages using pixel widths?

It's generated by the file checker of the index namespace. Can someone tell me what it means? Thanks. — Ineuw (talk) 03:14, 5 March 2021 (UTC)

@Ineuw: It's added automatically by (for the moment) {{Auxiliary Table of Contents}} and {{letter-spacing}}. See H:PXWIDTH for details. --Xover (talk) 04:45, 5 March 2021 (UTC)

Much thanks for the links. — Ineuw (talk) 03:07, 6 March 2021 (UTC)

"Do not move to Commons" template slightly broken

The documentation of Template:Do not move to Commons states that the template is supposed to add the page it's used on to Category:Media not suitable for Commons (or one of its subcategories). However, it seems this feature only functions correctly when the "expiry" argument is used, even though this argument is supposedly optional.

When used without "expiry", there's a piece of categorization wikicode visible below the template (see File:Canadian poems of the great war.djvu for an example). The page also isn't added to the aforementioned (hidden) category. -- Veikk0.ma (talk) 23:02, 5 March 2021 (UTC)

@Veikk0.ma: It was a recently introduced bug. It's fixed now. Thanks for the headsup! --Xover (talk) 23:27, 5 March 2021 (UTC)

"rescuing" works no longer in publication

This is for far into the future, but is there some means by which works that are no longer being published can be made "public domain" in the last 20 years of their copyright? I was thinking about this in terms of the recent Dr. Seuss controversy. As it is, some of his books will be effectively mothballed for decades at the publisher's request. "If I Ran the Zoo" was originally published 1950 and would have been public domain 2025 under the pre 1998 copyright, but will effectively have no new printings for over 2 decades from now since the copyright ends on 2045 under current law. If the company doesn't want to publish it anymore, could it be made available on here in 2025, or is this another "Disney Vault" type scenario? (SurprisedMewtwoFace (talk) 16:01, 2 March 2021 (UTC))

What you as an individual do for an out of publication work is for you do decide. Copyright is the right to make a copy, and we don't have that right for works that are under copyright in the United States. The copyright holders are able to do as they wish, and are not restricted by copyright law by what a publisher decides. There may of course be other considerations that limit people's actions under other laws. — billinghurst sDrewth 10:30, 3 March 2021 (UTC)

The concept is called an "orphan work" and has been the subject of $$$ legal cases involving Google and Hathi Trust. I'm not sure there's a solid case for being able to slap "free for commercial reuse" on them (specifically under US law) as we would need to be able to do to host them here. The Internet Archive has the Sonny Bono collection, but, crucially, they don't claim that you're allowed to freely re-use the works, they just claim they have the right to host and show them. Inductiveload—talk/contribs 10:49, 3 March 2021 (UTC)

this is the same case as mein kampf, which time has now passed in german. you could do a renewal search, to check PD status. there is controlled digital lending at IA. doubtless the cottage industry of "canceled" books will be treated like scihub. Slowking4 亞 Rama's revenge 16:55, 7 March 2021 (UTC)

Advice on Complicated authorship

Hi everyone, in the series of papers I am creating one of them has 184 authors, yes I know a headache. I have not uploaded this paper yet to commons am doing the first paper right now just waiting for COMS:OTRS to update licensing. Anyway I am trying to think of how best to capture the authorship without spending 2 pages just doing the authors. The original paper is online here on my taxon authority page on wikispecies it appears like this with only first few authors and a reference linking to the template here that has all the authors. Another option I guess would be to make both the author list and affiliation lists expandable boxes that default to closed and those who wish can expand them. Or any other suggestions are welcome. I am happy to do all the setup for this just need some direction on what you prefer. Cheers Scott Thomson (Faendalimas) ^talk 14:59, 5 March 2021 (UTC)

I would suggest to put a reasonable number of authors on the front page, then add an et al. Add edition = yes and paste them as raw text to the work talk page. I would think that it is more appropriate to get them all in the WD item, than all active hyperlinks here. — billinghurst sDrewth 05:39, 7 March 2021 (UTC)

Error in Module:Age

While importing Template:User Wikisource For (among other connected templates and a module), I got a Lua error:

Lua error in Module:Age at line 651: attempt to call local 'Date' (a nil value).

Backtrace:

1. Module:Age:651: in function "getDates"
2. Module:Age:816: in function "chunk"
3. mw.lua:518: ?
4. [C]: ?
5. [C]: in function "getAllExpandedArguments"
6. mw.lua:187: ?
7. [C]: in function "pairs"
8. Module:Arguments:207: in function "mergeArgs"
9. Module:Arguments:320: ?
10. [C]: in function "pairs"
11. Module:Userbox:54: in function "chunk"
12. mw.lua:518: ?
13. [C]: ?

What is this and how can it be fixed? -- AnotherEditor144 ^{t - c} 20:07, 5 March 2021 (UTC)

@AnotherEditor144: That's rather a lot of heavy and complicated dependencies for a userbox. Just sayin'…

The dependency chain ends up requiring w:Module:Date, but our Module:Date is actually c:Module:DateI18n. As it happens, we need to migrate our module to the new name anyways (the name change was specifically so they could coexist so we'll run into this kind of dependency conflict elsewhere anyway), but that's a somewhat bigger job.

PS. Copying between projects (or pages within a project for that matter) requires attribution (the "BY" part of the license), at a minimum with a link to the source page. But it is generally better to ask an admin to use the import function so we get the revision history. --Xover (talk) 22:09, 5 March 2021 (UTC)

@Xover: The attribution will be provided soon. -- AnotherEditor144 ^{t - c} 08:38, 6 March 2021 (UTC)

@Xover: You were going to move, but now you have a second Date module in the way! AnotherEditor144 ^{t - c} 08:53, 6 March 2021 (UTC)

@AnotherEditor144: I'm not following? --Xover (talk) 08:57, 6 March 2021 (UTC)

@Xover: You said you were going to migrate your current module to the new name, and there might be a conflict. AnotherEditor144 ^{t - c} 09:12, 6 March 2021 (UTC)

@AnotherEditor144: Hmm, no. At least I don't think I said that anywhere. I said that we need to migrate our Module:Date to Module:DateI18n, because that's what they've done on Commons and we imported our module from there. Once that is done we can import English Wikipedia's w:Module:Date to Module:Date here, which would solve your dependency problem. And I said that we need to do this irrespective of your immediate problem, because there are lots of relevant templates and modules that we might want to import that rely on these modules living at a particular name.

But the issue with that is that it needs to be done carefully by identifying all the templates and modules that have a direct or indirect dependency on the current module, and finding a way to migrate them without causing disruption on every single page that transcludes one of those templates. Which means this is slightly more involved (and hence time-consuming) than simply making a small fix to the code or renaming a wikipage, so you shouldn't expect it to happen within any particular timeframe. --Xover (talk) 09:29, 6 March 2021 (UTC)

@Xover: Then you should do it. AnotherEditor144 ^{t - c} 09:51, 6 March 2021 (UTC)

@Xover: The 'date' module at w:Module:Date (without i18n) will also be stored in my user space for your convenience: User:AnotherEditor144/Module-Date-enwiki

@AnotherEditor144: Feel free to keep a local copy in your user space, but please do not create anything in other namespaces that depends on a page in your user space. --Xover (talk) 11:51, 6 March 2021 (UTC)

@Xover: Nobody will notice the page anyway, as the only link is here. Feel free to use it during importing. It will only be accurate as of today. -- AnotherEditor144 ^{t - c} 11:54, 6 March 2021 (UTC)

More importantly, why do you think that we need Template:User Wikisource For? Seems a pretty pointless template. Just because enWP has a ton of valueless user boxes, doesn't mean that we need to do so. Nor do we need to jump by command to make something of limited value work here. Bit rude IMNSHO. — billinghurst sDrewth 05:27, 7 March 2021 (UTC)

Install .djvu version instead of .pdf?

This is an installed .pdf version with some poor quality pages. Downloaded another copy and converted it to The Barbarism of Slavery.djvu. Can I install the 2nd file and leave both file types for the time being?— Ineuw (talk) 07:41, 6 March 2021 (UTC)

@Ineuw: Certainly. We have plenty of duplicates like that. You can also move the Page:-namespace pages and even the Index: to the DjVu file (turn off "Create redirect" if you don't anticipate needing the PDF versions for anything). Once you're done proofreading the DjVu you can delete anything remaining for the PDF. --Xover (talk) 08:48, 6 March 2021 (UTC)

Thanks for the clarification.— Ineuw (talk) 06:20, 7 March 2021 (UTC)

Cobra!!! ....the Valentino Film

@PseudoSkull: I have just finished uploading Cobra (1925, Rudolph Valentino) to Wikisource. You can REEETTTRREEAAATTTT over there and find it at https://commons.wikimedia.org/wiki/File:ValentinoCobra.webm I will continue trying to find the remaining 1925 silents. A lot of them are lost films and it may take a while. I DID find a decent file of Merian C. Cooper's Grass but it has a video company logo on the very beginnning of it, right before the movie itself starts. This print is completely silent and has no other blemishes or attempts to claim additional copyright. Would you like me to post a link to it? (SurprisedMewtwoFace (talk) 19:12, 7 March 2021 (UTC))

@SurprisedMewtwoFace: Thank you for uploading this film. Please upload whatever you may find, and feel free to add them to this list when you're done: Wikisource:WikiProject Film/Uploaded to Commons. On another note, I think it might be better to post further messages regarding new films found etc. to the talk page of WikiProject Film so as not to clutter up the Scriptorium, as the Scriptorium is for more general discussion. PseudoSkull (talk) 22:14, 7 March 2021 (UTC)

WikEd v. 2017 Loading Error

When I've clicked "Edit" on proofreading pages such as [[3]], usually I've had the page fail to load (and no editing field), but if I load again it works. Today it failed to load altogether after reload attempts, so when I waved over the icon in the very upper right of the window and saw I could and "Disable" WikEd, I did so, and then the Edit page loaded fine. The error message was "Loading Error, WikEd v. 2017..."

Perhaps this has to do with the djvu and the complexity of the page; I have no problem with the editor or WikEd on other pages. Nissimnanach (talk) 01:10, 8 March 2021 (UTC)Nissimnanach

Djvu

If I have a screenshot of a news article that I have stored as a *.png, can I combine it with the text that I transcribed and transform that into a *.djvu file to store at Commons? Or are all *.djvu files we store already formed as image+text when Internet Archive scans a document and performs OCR? --RAN (talk) 21:21, 6 March 2021 (UTC)

@RAN: For our purposes the pre-generated OCR text is usually what we need (it's a starting point to make it easier to transcribe). When we generate DjVu files ourselves it is usually in bulk to add OCR for a (several hundred page+ book). However, if you have some need of a DjVu with a proofread text in its text layer I can probably make one for you (not done exactly that before, but it shouldn't be a problem). For Wikisource's purposes the text layer stops being relevant the moment you hit "Publish" on the Page: (only the page image is relevant for comparisons at that point), so you could just as well just paste the text in there and index off the PNG. --Xover (talk) 21:44, 6 March 2021 (UTC)

I just wanted to create one, so I can understand the file format better. I see online file converters that can change a png to a DjVu, but they do not actually integrate text. Can you recommend a free tool that makes combining text and an image as easy as creating a pdf, where the ASCII text or Unicode text, is stored, hidden within the document? --RAN (talk) 21:59, 6 March 2021 (UTC)

@RAN: I don't know of any that are actually easy to use. But the reference implementation of the format is DjVuLibre which contains downloadable (command-line) tools to do pretty much anything you like. Depending on your level of computer geekery you may find the tools anywhere from mildly complicated to completely inscrutable. For creating a DjVu from a PNG you'll want to look at c44 (DjVuPhoto) or cjb2 (bitonal). For adding text you'll need to look at djvused (the "set-txt" command). You may also want to look at GraphicsMagic for command-line image format conversions (DjVuLibre can't work with PNG images directly). You may also find some of the (now badly outdated) information on Help:DjVu files useful. --Xover (talk) 22:41, 6 March 2021 (UTC)

• Yikes! I will take a look, looks like a good opportunity for someone to come up with an easy to use set of tools. --RAN (talk) 23:01, 6 March 2021 (UTC)

  Typically we upload to archive.org, and then download, and use IA-Bot to generate a djvu. If you only have a single file, then it is a bit of a waste of time, as that can be uploaded and an Index: page generated with Page: links. then you can just OCR the images anyway. — billinghurst sDrewth 05:19, 7 March 2021 (UTC)

  yeah, i put in for an integrated upload dashboard tool on wishlist, and it got slow walked, as they did some other things. Slowking4 亞 Farmbrough's revenge 21:10, 8 March 2021 (UTC)

• @Richard Arthur Norton (1958- ): I have an image → DJVU+OCR script, but it's a bit of a mess and can't really be posted as a single script (there are imports and things). I'm slowly trying to beat it into shape so that it can be used by others. But for now, you can upload the image ZIP to the Internet Archive and I'll run it through for you.
• The problem with a hands-free solution is that there are a few knobs to adjust, for example some pages need the threshold varying because the print is too light or the page colour too dark. One day I hope to get it up on Toolforge with a web front-end. Inductiveload—talk/contribs 21:18, 8 March 2021 (UTC)

Thanks! I will try the upload and download trick. Can you load text into the metadata for a png file? I know there is a comments field I can access from one of my digital cameras, I don't know how much ASCII text it can hold. I see several online EXIF data editors for fixing incorrect dates stored with images. --RAN (talk) 22:11, 8 March 2021 (UTC)

I don't know about PNG. There certainly are metadata fields that can hold, as far as I know, an unlimited amount of data, but I don't know if there's any standard for embedding OCR. What I can produce for you is a DJVU with the text layer built in to each page. Inductiveload—talk/contribs 22:14, 8 March 2021 (UTC)

Tech News: 2021-10

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Recent changes

• Section translation now works on Bengali Wikipedia. It helps mobile editors translate sections of articles. It will come to more wikis later. The first focus is active wikis with a smaller number of articles. You can test it and leave feedback.
• Flagged revisions now give admins the review right. [4]
• When someone links to a Wikipedia article on Twitter this will now show a preview of the article. [5]

Problems

• Many graphs have JavaScript errors. Graph editors can check their graphs in their browser's developer console after editing. [6]

Changes later this week

• The new version of MediaWiki will be on test wikis and MediaWiki.org from 9 March. It will be on non-Wikipedia wikis and some Wikipedias from 10 March. It will be on all wikis from 11 March (calendar).
• The New Discussion tool will soon be a new discussion tools beta feature for on most Wikipedias. The goal is to make it easier to start new discussions. [7]

Future changes

• There will be a number of changes to make it easier to work with templates. Some will come to the first wikis in March. Other changes will come to the first wikis in June. This is both for those who use templates and those who create or maintain them. You can read more.
• Reference Previews will become a default feature on some wikis on 17 March. They will share a setting with Page Previews. If you prefer the Reference Tooltips or Navigation-Popups gadget you can keep using them. If so Reference Previews won't be shown. [8][9]
• New JavaScript-based functions will not work in Internet Explorer 11. This is because Internet Explorer is an old browser that doesn't work with how JavaScript is written today. Everything that works in Internet Explorer 11 today will continue working in Internet Explorer for now. You can read more.

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

17:51, 8 March 2021 (UTC)

Better Current collaborations Section

I don't think that the Current Collaborations section on the front page is serving us to well. For one, the current Community collaboration has been done for week. Instead, I propose that we have a runnig list that automatically progresses to the next work once the current one is done. Furthermore, I propose that we divide the texts into four categories and diplay all four at once to allow users more choice and cater to more skill levels.

1. Easy - These texts will be proofread texts that need validation. They will serve to introduce users to wikicode through an immersive environement and provide a low barrier to entry.
2. Medium - These texts will require proofreading, but have fairly decent OCR. They can be novel or books imported from PGDP.
3. Hard - These books have more complex layouts and may have more garbeled OCR, but they should not present too great of a challenge. Perhaps a book with lots of images or a few tables or a pre-19th century works with long s and ligatures.
4. Challenge - These are probably mainly reference books or manuscripts. Lots of complex formating required.

In this way, users can select from the range of difficulty and have a choice of which book they wish to proofread. Since the list will automatically update, we will not have a situation where a finished book will sit for weeks in the spot reserved for Community Collaboration. Languageseeker (talk) 02:26, 5 March 2021 (UTC)

@Languageseeker: The better place for this discussion is WT:Proofread of the Month as that is where that project is coordinated. We have had a variety of means to progress to subsequent works at the completion of the primary work. When I ran the project for a number of years we did followup plans. There is the means coded into Template:Collaboration/POTM to build a list of works that can be quickly implemented in either an extra work, a validation push, or overflow works. — billinghurst sDrewth 11:05, 9 March 2021 (UTC)

Author:Stephen King

I noticed we have an empty entry for Author:Stephen King which I imagine is for the copyright statement, would it be within rules to add in the titles of his works, even if they would all be red links, or all only point to the Wikidata or Wikipedia entry for each work. Or say we have an author where some are PD and others still under copyright, have a listing of all works and distinguish between the two. Or is the listing of works only for PD works we host? --RAN (talk) 02:18, 7 March 2021 (UTC)

I don't think there's any rule against doing so. I've seen it done for plenty of other authors. Thanks for helping out. Languageseeker (talk) 03:01, 7 March 2021 (UTC)

For authors where some works are PD and some are not, a complete list is acceptable (e.g. E. F. Benson. However, the works we can't host are best left unlinked. For authors where no works can be hosted by us, it's better to not list at all. In both situations this reduces the temptation to add works that would then have be removed under the copyright rules. As you say, the author pages for Stephen King & J. K. Rowling are there to indicate that they are copyright authors. Normally, we would delete such pages, but as these two (amongst others) are likely to be repeatedly created, it was decided to leave them with the copyright notice. Beeswaxcandle (talk) 03:57, 7 March 2021 (UTC)

I would think this is a sensible idea but I've had a very stupid edit war with an admin about this very issue so watch out. —Justin (koavf)❤T☮C☺M☯ 04:29, 7 March 2021 (UTC)

Per Beeswaxcandle, the community determined that we would not host pages where we could not host works, and/or they have no works in the public domain. Apart from the element of false advertising, once we open that door, we have very little ability to stop the conflict of interest additions. — billinghurst sDrewth 05:10, 7 March 2021 (UTC)

• When you say "the community determined" it would make the most sense to link to the RFC where that debate took place and consensus was formed, so we can all see it. What is a "conflict of interest addition"? --RAN (talk) 17:25, 7 March 2021 (UTC)

  In Wikisource:PD somewhere maybe 2010, 2011 or 2012, the conversation started with dealing with someone who was adding (semi-)modern Canadians, some of whom were politicians, though it reflected other pieces. Conflict of interest is someone, or their agent, who is publishing a book, and thinks that they can create an author page and list their works. — billinghurst sDrewth 23:41, 7 March 2021 (UTC)

• Again multiple people are reading the same policy are coming to the exact opposite conclusion. Stephen King & J. K. Rowling get empty author pages but Canadian politicians do not get pages. The rules should be so precise, and objective, that a bot can implement them. Subjectively enforcing poorly written rules is how we end up with selection bias and time wasting edit wars. --RAN (talk) 23:55, 7 March 2021 (UTC)

  No. Rowling and King were part of an overt discussion by the community where a consensus was reached it was determined that these were exceptions. This was mentioned above. There was no subjective enforcement. — billinghurst sDrewth 00:26, 8 March 2021 (UTC)

• Two authors out of 1,000 best selling authors, seems arbitrary, and is another example of selection bias. It is telling readers that only these two authors are important enough to be included, and that your favorite author is not important. The list would be objective if it was the best selling authors with no writing in the public domain. Maybe they fit that category, but why two and not ten? --RAN (talk) 07:03, 8 March 2021 (UTC)

  Says nothing of the sort. Stop getting your knickers in a twist about the community making decisions about issues that were discussed. The community is able to make their decisions based on the facts at that time. Go and read the discussion if you have that level of interest. — billinghurst sDrewth 11:54, 8 March 2021 (UTC)

The whole point of Wikisource:Scriptorium is to bring issues to the community without judgement. Being told I am getting my "knickers in a twist" is definitely not the way to handle a question, if you feel the need to insult the person asking a question, then let someone else answer it. Past practices should always be reexamined, especially when they create a selection bias. If the policy was clear and universally accepted, we wouldn't have half of the people responding to my inquiry come to opposite conclusion as you. You said the discussion was in "2010, 2011 or 2012", a decade ago, how is that representative of the thinking of the people contributing today? --RAN (talk) 13:25, 8 March 2021 (UTC)

The discussion actually appears to have been in 2017: Wikisource:Proposed_deletions/Archives/2017#Living_author_pages_with_no_works, and the "really famous author keep and mark {{copyvio author}}" concept was affirmed in 2020: Wikisource:Proposed_deletions/Archives/2020#Author:Richard_Dawkins.

Also, though it's not "Policy", but rather "help", the text Where a person is self promoting and/or no works are likely to be hosted at Wikisource. is listed as a contra-indication for creation of an author page at H:Author pages. So, though it's not clearly marked or laid out in policy or guidelines, the general discouragement is at least mentioned. I have expanded the help page to describe what I perceive to be "standard practice": Help:Author_pages#Authors_with_no_known_public_domain_or_free_texts. As it is help pages, it is not normative.

A better solution, IMO, will be, as I said in the Wikilinks discussion, to use something like {{wdl}} to allow us to anchor any incoming links to the immutable data item for a given modern author. In most cases, that will lead to Wikipedia. Then, one day, when an author page is created because a PD text finally surfaced, then the link will auto-switch to the author's WS page. Inductiveload—talk/contribs 14:58, 8 March 2021 (UTC)

• Thanks! I am not advocating for inclusion/deletion, I was just curious why we had an empty entry. Sometimes it takes someone seeing something for the first time to recognize it as not fitting in with the other entries. --RAN (talk) 23:36, 8 March 2021 (UTC)

• The documentation is at Template:Copyvio author about what and why, and Special:WhatLinksHere/Template:Copyvio author shows that we try to keep its usage to a minimum, so all other modern authors have typically been deleted. Reading the history of Author:Joanne Kathleen Rowling explains the process taken since 2007. — billinghurst sDrewth 11:42, 9 March 2021 (UTC)

Newspaper entries

I have been categorizing the various newspaper entries and I keep coming across ones that are not in any category. See for instance The Pittsburgh Press where it has a category for the city it is published in, and the state that it is published in. Can anyone think of way to find all the entries that are not categorized yet. We do not have a space for papers like "Periodical:The Pittsburgh Press" so the only way to find them is with an existing category. --RAN (talk) 00:34, 8 March 2021 (UTC)

Special:UncategorizedPages — billinghurst sDrewth 00:43, 8 March 2021 (UTC)

• That won't work, there are just too many uncategorized entries, I am at 5,000 I am still in the letter "A", and none were newspapers. --RAN (talk) 02:31, 8 March 2021 (UTC)

I will just have to be extra careful when I create them, to make sure I add a category, or they will get lost. --RAN (talk) 06:42, 8 March 2021 (UTC)

@Richard Arthur Norton (1958- ): ...think of way to find all the entries: do you mean articles or newspapers.

As long as you create articles under the newspaper's main page, they'll be easy enough to find. BTW, Wikisource:WikiProject Newspapers recommends Newspaper Name/YYYY/MM/DD/Article Name for the article page name, rather than Newspaper Name/YYYY/Article Name. Inductiveload—talk/contribs 09:35, 9 March 2021 (UTC)

• Yes, I am aware of Newspaper Name/YYYY/MM/DD/Article Name. I made my first contribution under Article Name and another editor migrated it to Newspaper Name/YYYY/Article Name, and I have been using that format before I read Portal:Newspaper. I recently contributed the paragraph describing all the permutations currently in use. --RAN (talk) 02:16, 10 March 2021 (UTC)

Turkish fairy tales and folk tales (1901).djvu

The following discussion is closed:

Source file updated, index adjusted.

The scan for Index:Turkish_fairy_tales_and_folk_tales_(1901).djvu is missing p186 and p187. I provided the two missing pages from a separate scan. Can you please integrate them. Afterwards, can you shift the text by +2 starting at Page:Turkish fairy tales and folk tales (1901).djvu/7 (the text for DJVU page 9 is on pg 7). Much Appreciated Languageseeker (talk) 02:17, 6 March 2021 (UTC)

@Languageseeker: Done --Xover (talk) 15:04, 29 March 2021 (UTC)

@Xover: Thanks! Languageseeker (talk) 16:56, 29 March 2021 (UTC)

This section was archived on a request by: Xover (talk) 07:11, 30 March 2021 (UTC)

Spaulding Guide 1894

The following discussion is closed:

Pages moved.

Can you move the text from Index:Spalding's official base ball guide, 1894.djvu to Index:Spalding's base ball guide, and official league book for ... - a complete hand book of the national game of base ball .. (IA spaldingsbasebal02chic).pdf. The offset should be +3. Thanks Languageseeker (talk) 20:19, 13 March 2021 (UTC)

@Languageseeker: The original pages where created by Match & Split, so just undo the "split" edits by phe-bot in mainspace and then redo the M&S to the new index. If the old index's file is broken in some way then please tag the index accordingly (or ask for deletion). --Xover (talk) 21:15, 13 March 2021 (UTC)

@Xover: Normally, I would but a user has proofread a number of the pages and I don't want to lose those revisions. Languageseeker (talk) 23:12, 13 March 2021 (UTC)

@Languageseeker: Done Migrated to Index:Spalding's Baseball Guide (1894).djvu and pages shifted. The PDF was of dubious quality, and that auto-generated name a real pain to work with, so I regenerated it from the source scans instead. As a nice side effect you've now got just north of 10x the resolution to play with. --Xover (talk) 10:21, 30 March 2021 (UTC)

@Xover: Thanks! Languageseeker (talk) 00:03, 3 April 2021 (UTC)

This section was archived on a request by: Xover (talk) 14:00, 6 April 2021 (UTC)

The Book of Evelyn

The following discussion is closed:

Index abandoned.

The generated DJVU on IA only has 372 pages while the JP2 archive has 379 images for Index:The book of Evelyn (1913).djvu. Is there anyone that can generate a new DJVU for this book and upload it? Languageseeker (talk) 15:48, 17 March 2021 (UTC)

@Languageseeker: the "missing" pages are usually non-content things like tissue paper plate protectors and the IA doesn't include them in the output formats like PDF and DJVU. Is there any missing content? Inductiveload—talk/contribs 16:06, 17 March 2021 (UTC)

There is no missing content. The difference is 2 plate protectors and 4 scan calibration pages (the zip contains 378 images, not 379). --Xover (talk) 16:22, 17 March 2021 (UTC)

@Xover, @Inductiveload: Thank you for checking. Yes, the book is unfortunately missing several pages. Languageseeker (talk) 16:38, 17 March 2021 (UTC)

This section was archived on a request by: Xover (talk) 10:25, 30 March 2021 (UTC)

The Aeneid of Virgil JOHN CONINGTON 1917 V2.pdf

The following discussion is closed:

Moved.

The file The Aeneid of Virgil JOHN CONINGTON 1917.pdf was deleted on Commons. So I uploaded The Aeneid of Virgil JOHN CONINGTON 1917 V2.pdf and moved the Index for The Aeneid of Virgil JOHN CONINGTON 1917.pdf to Index:The Aeneid of Virgil JOHN CONINGTON 1917 V2.pdf. Could you move the pages as well? Languageseeker (talk) 15:23, 23 March 2021 (UTC)

@Languageseeker: Done --Xover (talk) 11:13, 30 March 2021 (UTC)

@Xover: Thanks! Languageseeker (talk) 23:52, 2 April 2021 (UTC)

This section was archived on a request by: Xover (talk) 14:01, 6 April 2021 (UTC)

Ability to Add Formatting Guidelines or Important Information to the Index Page

I think that it would be extremely useful to have the ability to add some formatting help or general guidelines on the Index page itself. I know that we can add them to the Discussion section, but this is out of the view of users. This would save users having to look through the vast help section to just find a tiny bit of information. Instead, we could have a mini-help section on every Index page to deal with any bit of formatting necessary. This would be easy for users to find and make it possible for them to contribute. Languageseeker (talk) 20:39, 7 March 2021 (UTC)

This is done on the Index talk page. See Index talk:Manual of the New Zealand Flora.djvu for one of mine. Then look at the Index itself. Note the banner across the top pointing to the fact that there are formatting guidelines on the talk page. Beeswaxcandle (talk) 05:05, 9 March 2021 (UTC)

Yep, but that's less visible to users. My proposal is to move such guidelines from the discussion page to the main index page for greater visibility to reduce user error and confusion. Languageseeker (talk) 06:37, 9 March 2021 (UTC)

@Languageseeker: Putting all the formatting instructions on the Index page would make rather a mess of the page. Some (not many) works have a lot of notes (e.g. transliteration tables, etc), and are sometimes subject to in-line discussions. Also, the Index pages are "secretly" actually just a big ol' template MediaWiki:Proofreadpage index template with a magical form-based edit interface. Stuffing arbitrary formatting instructions (e.g. tables) into the form fields is going to throw up frustrating edge cases, even if you had a good place to dump them on the page.

I suggest looking at ways to make {{Index talk remarks}} clearer. It's possible (in theory) to add a field to the Index page template, and pass it along to {{Index talk remarks}}, but I'm not sure what we'd put in such a field. Inductiveload—talk/contribs 08:13, 9 March 2021 (UTC)

Less visible??? There is clear text that says look at the talk page for formatting. If they aren't seeing that then maybe they are not looking. The formatting on the Index: page really has a little bit of room at the top. There is a little scope to add something to the TOC field, as we did with DNB works in the early days though it was horrible for trying to add detailed instruction, hence why we moved to the talk page. — billinghurst sDrewth 11:10, 9 March 2021 (UTC)

I can see it, but even one additional click makes it less probable for a user to see it. Also, they don't always exist. It's easy to forget that some users may not know how to add bold text or italic. If a new users clicks on a transcription project, then they shouldn't have to dig through pages of documentation to find the answer. I'm basically asking for the ability to make a mini help page on every Index page. Languageseeker (talk) 12:43, 9 March 2021 (UTC)

@Languageseeker: but why is the index page the right place for this? If it's stuff like bold and italic, that's general formatting and the ">Help" menu at the top of the edit box on every page contains that. Index talk pages contain special formatting conventions for that index only. BTW, it's on my secret list to figure out how to make our edit box top bar more useful for what we actually do, since it's contains not much of use for WS special sauce. Inductiveload—talk/contribs 12:50, 9 March 2021 (UTC)

@Inductiveload, @Billinghurst:The index page is the appropriate space because it is the first thing that a user sees when they decide to participate in a transcription project. Based on the Index page, they will decide whether or not to contribute. Right now, the Index page is about as exciting as a library catalog record. I want to increase the motivation of users to participate in a transcription project. As an example, I wrote up what I would like to see on the Index Page here. Ask yourself the question, based on this information are you less likely or more likely to participate?

@Inductiveload: Your custom edit bar sounds awesome. I have some ideas if you want to hear them. Above all, I think that we should be able to customize the edit bar based on the specific book. Languageseeker (talk) 18:19, 9 March 2021 (UTC)

@Languageseeker: For most/many of our works there should be ZERO requirement for special formatting instructions. They typically have been used for more complex multi-volume works, or for where some PotM so that there is uniformity. Having someone increase participation is not through putting formatting instructions on an Index: page. If one click is going to deter someone, then 200 to 500 clicks later for a 200 page book is going to kill them. Do not try to make an Index: or an Index: talk page into a Wikisource:WikiProject page. For complex co-operative works, especially those in a series, a good project page is a better way to go, and from there you can add sections and transclude those sections into relevant Index: talk pages. I would much rather look to encourage project pages, and active project talk pages, and invest in linking Index: and project pages to work better. — billinghurst sDrewth 21:18, 9 March 2021 (UTC)

That said, I think that there is some scope for how we can utilise Template:Index talk remarks and do some simple/additional/introductory text on the talk page of the index that could be transcluded back into the Index: page. We don't want to make it too much or too busy. — billinghurst sDrewth 21:23, 9 March 2021 (UTC)

On that note, what we could do, fairly easily, is add a field to the Index page form and template for "relevant Wikiprojects". I'm not sure where would be best for them on the page (e.g. top right or the main table).

It occurs to me that we could also add a drop down for setting {{index transcluded}} parameters too, rather than stuffing it into the TOC field. Inductiveload—talk/contribs 21:25, 9 March 2021 (UTC)

@Inductiveload, @Billinghurst: Thank you for the detailed feedback that helped me clarify my thought. I definitely don't want to flood the index page with lots of extra content, but I do think a little help can be useful, especially for the tricky cases. This is the approach that video games take: a bit of guidance helps users to invest hundreds of hours in difficult tasks. I do think that we can provide a bit more information on indexes without cluttering it up.

I that the idea of transcluding from the Index Discussion to the Index Page is a great idea. Perhaps, we could have a set section called Transcribing Guidelines that we could transclude. This section would contain an introduction to the work, it's significance, and any special instructions. We would tell users to keep it short.

I disagree (my opinion only) that the Index: ns is the place to talk to them about the work; Index: and its talk space are solely a workspace and instructional, rather than contextual or informational; those aspects of a work belong in more obvious and in front-facing content namespaces, which for us have been main, WS:, Author:, or Portal:. I would encourage you rethink that approach, they are not a place for bloat. Help:Namespaces gives that direction. — billinghurst sDrewth 01:22, 10 March 2021 (UTC)

@Billinghurst: I'm not asking to add a wall of text. I'm merely asking to add one or two sentences to explain why a user should care about this book and a few special formatting instructions (if any). We can even make the section collapsible to reduce bloat. For more detail information, we can include a site matrix as we do on transcluded works. Look at the current Proofread of the Month Women of the West can you give me one reason that is obvious to the user as to why they should help to transcribe it? We can't just give users a book, tell them to transcribe it, and expect them to get involved. People need reasons to volunteer. Languageseeker (talk) 01:57, 10 March 2021 (UTC)

And I will say it again, the Index: namespace is not the place to do it, it is not joe-public front-facing namespace, you are already too deep. I have no issue with all the encouragement, excitation, interest, etc. and it belongs before you get to an Index: ns. Keep it simple. — billinghurst sDrewth 04:01, 10 March 2021 (UTC)

That seems like a fairly elitist attitude. Every user is a new user at some point. Where are they supposed to learn? Languageseeker (talk) 05:36, 10 March 2021 (UTC)

What?!? You are trying to do multiple things, and think that the Index: page is the panacea. I am disagreeing with your proposal.

Every new contributor has a welcome message that gives them all the basics, and should lead them to our general approach, and anything general. We have agreement and general practice that where this is specific formatting that may be specialist to the work that it belongs on the Index talk: page. However, all explainers about the work, its place in a corpus, etc. do not belong on the Index: page, they belong in our other namespaces where they are more visible and more organised, and I have explained why. So please stop the throwing of insults. We fix up the issues where they lie, not poke them all into an Index: page. As I mentioned previously, we have used the ToC field for some commentary about works, eg. DNB volumes. — billinghurst sDrewth 09:50, 10 March 2021 (UTC)

@Billinghurst: I’ve been feeling bad about my poor choice of words. You didn’t deserve them at all. Please, accept my apologies. Languageseeker (talk) 20:22, 12 March 2021 (UTC)

If we take a cue from video games, an "in-page tutorial" makes more sense to me. Like when websites flash a little "helpful" (I find them rather a speedbump, but they must be useful or they wouldn't be used) marker over "new" items. You are fundamentally right, on-boarding and new-user-support, as well as complete but accessible documentation, is hugely lacking, and we must improve it.

BTW, The main "portal" for basic instructions is Help:Beginner's guide to Wikisource. Inductiveload—talk/contribs 09:31, 10 March 2021 (UTC)

Absolutely agreed, PGDP include similar information on their project pages and it works. We cannot assume that users will know before they stumble on an index page. Even including a link to proofreading guidelines on index pages would be a huge help. We really need to make index pages less stark. I think there’s also fear of redundancy between headers and index page. Can’t we pull more information from the info on index pages rather than manual filling in headers? Languageseeker (talk) 20:00, 10 March 2021 (UTC)

Wikisource:Annotations => to official document

We have had the page sitting there and chanting it as an official document for so long, so I think that it is just time we take off the template at its top. Time to just move on it, or fix it and move on it— billinghurst sDrewth 05:43, 7 March 2021 (UTC)

•  Support --Jan Kameníček (talk) 13:30, 7 March 2021 (UTC)
•  Oppose I apologize for this being so long: The document as it is written now is contradictory concerning wikilinks within the body of entries, people are citing it and coming to opposite conclusions. This has created edit wars over adding/removing wikilinks, wasting time. I prefer wording that complies with the RFC located here: "Wikilinks are annotations and are allowed in Wikisource. The creation of wikilinks is optional, where created they should be based on context, the type of work and the likely reader." We now have contradictory policy pages Wikisource:Annotations and Wikisource:Wikilinks with each claiming they ban or allow wikilinks in the body of an entry. The same wording needs to appear in each, so they do not contradict each other. The policy should also address "specially labelled separate copies of works" that User:Xover keeps referring to, but has not been able to give an example of. It would also help if the policy page contained clear specific examples of proper annotations and improper annotations. When we only have words, and no examples, people interpret the words differently. Read the Wikilinks discussion on this very page, above, where multiple people are coming to opposite conclusions based on the wording of the very policy page under discussion as it written now. We also have to address when two policy pages contradict, which has supremacy. We also need a better policy on flagging errors of fact and flagging spelling errors. They should be addressed, but in a way that the annotation is distinguishable from the original source material. Just adding [sic] without telling the reader what the correct word should be, leaves the reader in the dark. For instance, the New York Times addresses errors in their online articles, we should have a system here, where it is clear that the correction is not part of the original article, and that it has been added by a Wikisource editor, perhaps at the bottom of the page below the license. Clearly we do not need these for fiction, but for newspaper articles. If we do not address errors, people will be using the errors as references in Wikipedia and Wikidata. If an obituary states that a person was born in 1880 and we have the birth certificate at Commons and we have that person in the 1900 US census, and those documents use 1878, we should address that at the bottom of the page, in a way that the reader recognizes it is an annotation by a Wikisource editor. --RAN (talk) 19:16, 7 March 2021 (UTC)
• I know that comparing texts in crude blocks is disallowed, but what about automatically comparing differences between editions? This is quite common in scholarship. Languageseeker (talk) 03:17, 9 March 2021 (UTC)

  Special:ComparePages? — billinghurst sDrewth 11:47, 9 March 2021 (UTC)

• @Languageseeker: Personally, I'm not philosophically opposed to the idea, but I have yet to see any really good example of that kind of thing. The problems are a pincer of the Mediawiki platform we have being a fairly poor fit for that kind of thing on the one hand, and on the other the fact that creating such a work is actually a pretty major undertaking compared to the value proposition and people generally burn out and abandon their project after the first chapter or so. Also, since we (rightly, IMO) don't allow interpretive annotations in any case, such a comparative work is extremely dry. I imagine a better WMF fit for such a thing might be at Wikibooks or Wikiversity, where a comparator can add extra commentary about the differences. Inductiveload—talk/contribs 12:18, 9 March 2021 (UTC)

• I have created what I think can be a model of an annotation pointing out an error-of-fact in an obituary. See Jersey Journal/1914/R. V. Schuyler. If we were to let it go unrecognized, we allow it to be used as a reference for Wikipedia with the error intact. I leave the error intact, mark it with "[sic]" and add a note below the license template, so no one would think it was part of the original text. Most people at Wikisource are transcribing fiction, so this would not be a part of their transcriptions. --RAN (talk) 04:14, 12 March 2021 (UTC)

Does anybody maintain the IA upload tool?

task T276222 task T276648 I've been having many issues with the IA upload tool and I've opened phab tickets for them. However, they seem to be just sitting idly. How long does it usually take to fix the tool? Languageseeker (talk) 01:47, 9 March 2021 (UTC)

It depends, sometimes months, sometimes years :-( --Jan Kameníček (talk) 12:18, 9 March 2021 (UTC)

Just like anything else in phabricator. Sometimes you just have to befriend a developer the right way. It is why there is a wishlist survey every year or so. — billinghurst sDrewth 21:27, 9 March 2021 (UTC)

Thanks for the replies. A bit frustrating, but that’s that how work. Time to make friends with a dev. Languageseeker (talk) 06:12, 12 March 2021 (UTC)

San Diego Union/1915/Cause of Taliaferro's Death Plunge Mystery to Airmen

I removed the "unlinked" tag from San Diego Union/1915/Cause of Taliaferro's Death Plunge Mystery to Airmen because it is now automatically linked at San Diego Union, however because I used the autolinking process that comes with the periodical header, it will not appear in "what links here". Are we requiring hand-made bulleted links, so articles do not get slapped with the "unlinked" tag? Or is the tag asking me to create an entry for the article in Wikidata? --RAN (talk) 06:06, 12 March 2021 (UTC)

@Richard Arthur Norton (1958- ): I think linking to it from San Diego Union is sufficient, since it can now be reached. However San Diego Union is itself functionally unlinked (the only links come from its own child pages). Probably it should be linked from Portal:Newspapers and Portal:California#San_Diego, as well at categorised to Category:Newspapers of California. Inductiveload—talk/contribs 11:59, 12 March 2021 (UTC)

Typically I just stick in previous = [[../../]] — billinghurst sDrewth 00:40, 13 March 2021 (UTC)

Deprecation of the Special:Book tool

I think now that the ebook exporter is working rather nicely, we should look at deprecation of the PediaPress Special:Book tool.

It has been functionally broken (the only think that works is the link to PediaPress for buying a book) for a very long time and, as far as I know, there is no effort being made to unbreak it. As it is, it just clutters the sidebar and de-emphasises the (working) WS-export links.

Furthermore, the Wikisource:Books is confusing, as all the entries are useless when the tool is broken. Inductiveload—talk/contribs 12:08, 12 March 2021 (UTC)

Disambiguation downloads

Is there any way to deactivate the big "Download" button on disambiguation pages, or do we want those to be downloadable? Visually, the button implies that the disambiguation page is a work in its own right. --EncycloPetey (talk) 01:16, 2 March 2021 (UTC)

It should be easy to code to deactivate, they have __DISAMBIG__wikicode and are automatically categorised. I don't think that people would need to download, though at the same time it is out of the way of the body so not causing issues. — billinghurst sDrewth 06:15, 2 March 2021 (UTC)

Tracked in PhabricatorTask T273708

This is phab:T273708 and the idea is that when done it will provide a magic word that we can add to templates like {{disambiguation}}. Inductiveload—talk/contribs 20:14, 3 March 2021 (UTC)

Addendum: The same issue applies to Versions pages. --EncycloPetey (talk) 21:10, 15 March 2021 (UTC)

This section was archived on a request by: Leave this with the phabricator ticket for further resolution. — billinghurst sDrewth 01:23, 10 April 2021 (UTC)

Wikilinks

When I asked previously about Wikilinks, I was told not to overlink, and no interpretive links. The example given was linking "Lewis Carroll best known book" to "Alice In Wonderland" because it is a subjective interpretation that can change over time. Why are my links to real objective people and real objective places being deleted? Again, this feels like tag-team harassment. Especially when I am probably going to be the only person on Earth using the entry for research over the next decade. --RAN (talk) 06:35, 2 March 2021 (UTC)

The links that I removed were all going off-site and not here on Wikisource. A Wikisource page should have minimal linking in the text, except to authors and other works. Links within text to Wikidata items are outside the intention of the Wikilinks policy (which was written before Wikidata existed). Links within text to Wikipedia are covered in the policy. "Myrtle Avenue" (to pick a random example) can reasonably be assumed by the intelligent reader to be a residential street based on the context. I intended my edit to be an exemplar of good practice, but I see you have just pasted your version over the top of my edit and removed the layout formatting I had done at the same time. Beeswaxcandle (talk) 07:55, 2 March 2021 (UTC)

Fiction and non-fiction should not be treated the same. No links may be good for fiction, but for non-fiction news articles, Coroner Rollins really is somebody, there is no speculation, it is not a roman à clef like Primary Colors where people endlessly speculate who corresponds to the fictional person portrayed. I can see where the London of Alice in Wonderland, may not be the real London, it may exist only as a fictional London. Non-fictional news-articles are different, and the reader loses, when the real historical Coroner Rollins is not identified and linked to his entry in Wikidata. I can search for all the entries for "Coroner Rollins" even if he is called "Aaron Burr Rollins" in another article or called "Sheriff Rollins" using insource. Fictional speculation should not be treated the same as identifying real people in news-items. The reader should get help and identify Moscow as "Moscow, Idaho" and not "Moscow, Russia" or "Moscow, Tennessee" or "Moscow, Kansas". People read fiction for entertainment, people read the news-articles because they are researching someone for a Wikipedia article or a Wikidata entry or a class paper, or a Ph.D. thesis. The links are part of being scholarly. No one is forced to click through to Wikidata, I have been reading from the Wiki Universe for about 20 years and I never felt compelled to click on a link I had no interest in, or was distracted by a word in blue instead of black. But I have come across places like "Myrtle Avenue" (to pick your random example) in news articles, and wondered if it was the same Myrtle Avenue from another news article I read earlier. --RAN (talk) 13:26, 2 March 2021 (UTC)

I have on my todo list trying to gain support for revisiting our linking policy. It is confusingly written and as written it excludes links to Wikipedia that were actually explicitly supported in the discussions leading up to it. I want to revise it to allow links to Wikipedia but to limit under what circumstances they can be used. It would seem sensible in such an effort to also address Wikidata and interlanguage links to other Wikisources. Possibly to explicitly outlaw them, but also possibly to explicitly allow them under certain constraints. For Wikidata, for example, we migh require the use of a template that has logic like "Link to Author: page if it exists, otherwise to Wikipedia article if one exists, otherwise add a tooltip with the Wikidata ID or a summary card of the Wikidata entity, and add the page to a tracking category" (randomly picked off the top of my head; community discussion would be needed to determine the details). My main goal would be to make the policy much clearer as to what is allowed, what is not allowed, and why. --Xover (talk) 13:47, 2 March 2021 (UTC)

I prefer linking to Wikidata because the links are more stable. In Wikipedia people are constantly being moved from name to name, and redirects are not always left behind to create a synonym. Especially where there might be a dozen people with the same name. "John Smith (politician)" might be moved to "John Smith (lawyer)" or "John Smith (New Jersey politician)" or "John Smith (mayor)" and a different person then occupies the previous "John Smith (politician)" or it may become a disambiguation page. Link rot is higher in Wikipedia. Usually the bare minimum information is just what I am looking for, not a full biography. Locations in Wikipedia are more stable. What do you think? Will the new rules allow a choice between Wikidata and Wikipedia by the person doing the transcribing, or will we insist on a Wikipedia link? --RAN (talk) 13:58, 2 March 2021 (UTC)

Without thinking in depth about it I suspect I would actually disagree with you on that, but that would in any case be up to the community to decide iff there was support for revising that policy. As it stands it certainly doesn't allow them. --Xover (talk) 14:24, 2 March 2021 (UTC)

Without wishing to get too bogged down in details, using the WD Q-id probably is the most reliable link target, and because we have JS and Lua access to Wikibase, it means that we can actually be much smarter about where we link to (i.e. fall back though Author/Portal, Wikipedia, then Wikidata, as well as show WS interwikis contextually—handy for foreign authors that we won't likely have here for a long old time). I would also say that, unless the user specifically asks for it (e.g. by clicking a Wikidata icon) delivering a user directly to Wikidata, as opposed to, say, Wikipedia, is a last resort, and that keeping them on Wikisource is preference when possible. Inductiveload—talk/contribs 15:10, 2 March 2021 (UTC)

Also, we should bear in mind that Wikidata hardly existed when most of this policy (such that it is) was written, and the ability to use Lua to deal with that data in a more sane way is even more recent. Inductiveload—talk/contribs 14:02, 2 March 2021 (UTC)

And I agree with the policy of not overlinking, no need to link to commons words, you usually only need to link to a person, or place name, the first time they are mentioned in a news-article. I agree with the policy of about being vigilant with speculative links within fictional works per my examples above. --RAN (talk) 14:22, 2 March 2021 (UTC)

• @Xover: Can you quote me exact phrase at Wikisource:Wikilinks that bans links to Wikidata. I see a ban on "interpretive links" to things like "favorite book", I can see where two people would interpret "favorite book" differently, or that it may change over time. I see a ban on links external to the Wikimedia Universe. That makes sense because of link rot being detrimental to long term stability. I can see the possibility for controversy over adding links within fictional works. --RAN (talk) 04:18, 4 March 2021 (UTC)

• @RAN: Sure. It's Links to Wikimedia-project pages are acceptable and considered to be annotations. [my emphasis]. The key here is that the linking policy says Wikimedia links are "acceptable" and in the same breath that they are "annotations". And annotations are, according to to the annotations policy, not allowed in works. In other words, the text that is phrased as if it is saying such links are ok is actually saying they're forbidden. This is incredibly confusing and nearly impossible to figure out without assistance (I had to have it explained to me after I had merrily added Wikipedia links to works for a good long while). --Xover (talk) 07:26, 4 March 2021 (UTC)

Then it is a good time to revisit policy pages, two policy pages I am constantly referred to, were marked "draft" and "essay". Eventually all lists of laws become self-contradictory when you have enough of them. --RAN (talk) 14:04, 4 March 2021 (UTC)

Also to make it even more of a mess, WS:ANN has never graduated to policy, it's still a proposal! Inductiveload—talk/contribs 14:18, 4 March 2021 (UTC)

• @Xover: The most recent ruling based on a consensus RFC was in 2013 and reads: "Wikilinks are annotations and are allowed in Wikisource. The creation of wikilinks is optional, where created they should be based on context, the type of work and the likely reader. There are a number of ways wikilinks could be miss-used (interpretative vs. non-interpretative) and a separate discussion will identify acceptable types of wikilinks." (2013) --RAN (talk) 14:42, 4 March 2021 (UTC)

This is why I hope people will support the interaction ban between myself and the person with admin rights who keeps enforcing their personal preferences as if it were !Wikilaw. There are plenty of other people to patrol new entries that can interact with me. --RAN (talk) 14:47, 4 March 2021 (UTC)

@RAN: As I wrote above, annotations are indeed permitted, but only in specially labelled separate copies of works and only when a complete unannotated version exists. The policy doesn't permit annotations (in the form of the links in question) in regular works. --Xover (talk) 18:14, 4 March 2021 (UTC)

You have said that multiple times, but the ruling makes no mention of the phrase "specially labelled separate copies of works" or anything that resembles that wording. --RAN (talk) 03:12, 7 March 2021 (UTC)

• Show us an example of where this is done. Showing how it should be done, rather than describing how it should be done, will make it clear to everyone. We are arguing over contradictory wording on multiple RFCs and essays and draft policy pages. One good example will make it clear. I am still not sure if the "complete unannotated version" you are referring to is the scan of the original document, or we are supposed to cut and paste the text twice in every entry, one with links and one without links. Are formatting changes also annotations? Is choosing the text size for a headline changing the urgency of a news article? Is it a type of annotation? Look at War! and War! and War!, the same word but the urgency is different. If we don't match the original exactly, is that annotation? Is not formatting a headline and just using the default ASCII text a type of annotation by changing the urgency of a headline? I am looking through all the news articles and many entries have not been formatted, they are just unformatted ASCI cut and pasted text. --RAN (talk) 23:45, 5 March 2021 (UTC)

One more point. If we ban wikilinking in the body of entries, there is really no point having the entry here. The only point of bringing a text into the WikiUniverse is to be able to link to Wiktionary, Commons, and Wikidata. Almost every text already exists elsewhere on the web via Internet Archive or Google Books. --RAN (talk) 22:49, 13 March 2021 (UTC)

UncategorizedPages on Wikisource

I was looking to see if wikisource had any sources in the Javanese language. A search on the term "Javanese" brought up several hundred pages, however when I tried to find out what category(ies) these pages belonged to I discovered many were not categorized. I then checked Special:UncategorizedPages and I see that the number of UncategorizedPages is quite large.

Just wondering if this is by design, or just lack of enough volunteers to tackle the problem? Thanks in advance, Ottawahitech (talk) 11:42, 10 March 2021 (UTC)

Several factors. Primarily, the special page is not designed for how Wikisource operates. Page: and Index: ns pages (our workspace) typically do not need categorisation; they are not seen as display pages, and generally only need categorisation with maintenance categories where it is required. We would normally only categorise subpages where they are different from the category of the work, so many subpages show. There is no means to have the special category display in main namespace works only nor to filter out subpages. Then often it is wrong as there are categories and it just doesn't sense it properly. Hence we are where we are. — billinghurst sDrewth 21:55, 10 March 2021 (UTC)

Then we possibly have fallen into laziness, and some of the fiction categorisation is not necessarily helpful. Displaying a listing of works in Category:Short stories is of limited value. — billinghurst sDrewth 21:59, 10 March 2021 (UTC)

Note that it would be of substantial value if the OPDS generator could present exportable works organised by category, like Feedbooks does (https://catalog.feedbooks.com/catalog/public_domain.atom). Inductiveload—talk/contribs 22:36, 10 March 2021 (UTC)

we also do not have a cadre hand creating and maintaining categorys. we could do some things, organize cleanup and search, but would need a quality circle to implement. Slowking4 亞 Farmbrough's revenge 15:58, 12 March 2021 (UTC)

Something that can be done right now is to watch New Texts as they are listed on the main page, then add suitable categorization for form, topic, etc. Quite often the only categorization these pages have when they are first listed are the automatic date and copyright categorization. --EncycloPetey (talk) 18:54, 13 March 2021 (UTC)

Category:Pages using duplicate arguments in template calls

Category:Pages using duplicate arguments in template calls has ballooned to over 700 pages, but looking at pages like Federal Reporter/First series/Volume 64 and Delaware Code/Title 4, I am not seeing any sort of duplicated parameters. This leads me to believe that false positives are being generated by templates on these pages, perhaps in the {{Header}} or {{Process header}} or {{Incomplete}} templates. BD2412 T 17:32, 12 March 2021 (UTC)

It will be something related to {{header}} or its modules, as that is where Inductiveload has been making the recent changes. — billinghurst sDrewth

I'm not sure that is true. AFAIK, it's not possible to call a template from Lua with duplicate arguments, because a Lua table can't have duplicate keys and tables are how you call frame:expandTemplate, so it must be an issue higher up than the modules.

Furthermore, very many items "in" this category are not actually in the category when you look at them (including your two example), so the category page itself appears to be/have been stale. The category itself seems to have shrunk again to ~75 entries, all of which, so far, have had valid issues (or transclude pages that do). I haven't changed anything, perhaps someone else has? Inductiveload—talk/contribs 13:26, 13 March 2021 (UTC)

@Inductiveload: Module talk:Pagetype/testcases gives every appearance of managing that feat from Lua. I couldn't be arsed to figure out just how it does it, since the thing needs to be taken out back in any case. But if you want to dig… --Xover (talk) 18:32, 13 March 2021 (UTC)

@Xover: OK, so you can do it with frame:preprocess, (in this case {'|page=Page talk:Example|page=custom text', 'custom text'},) but none of the modules I've messed with do that, because they call expandTemplate, which take a table, rather than taking manually-constructed wikicode but I haven't always been so tidy with constructing HTML tags. Inductiveload—talk/contribs 00:10, 14 March 2021 (UTC)

@BD2412: (CC Inductiveload since they were pinged) Most of this was just stale links tables, so touching the pages in the cat fixed it. It's possible the stale category association was caused by some of the header changes recently, but that's impossible to tell after the fact. In any case, the cat is down to ~75 entries now, and these appear to be legitimate instances of duplicate arguments. I've done some manual cleaning and these should be eminently fixable by hand. People may also want to watchlist this cat and fix any new entries as they are added. --Xover (talk) 13:27, 13 March 2021 (UTC)

Format help

Hi, can someone show me the correct format for inset text blocks like in the 3 un-validated pages of Index:Creole Sketches.djvu?

P.S. Why am I able to see whether edits are patrolled in Recent Changes? I'm just a User. It's a bit jarring to see the red exclamation mark next to all my edits. Ultimateria (talk) 07:02, 13 March 2021 (UTC)

@Ultimateria: I've set the poems on those three pages for you, but have not changed the page status. In re seeing the patrolled status on Recent Changes. This is a default setting for all signed-in users. I think you can turn it off in Preferences. Beeswaxcandle (talk) 08:07, 13 March 2021 (UTC)

"Gentlemen Prefer Blondes" status confusion

Looking at the index, it appears Gentlemen Prefer Blondes has been completely validated. Yet when you look at where the work is transcluded on site, it says "not proofread", and there is no indicator on site that the text has been completed. I don't understand what happened. (SurprisedMewtwoFace (talk) 13:32, 13 March 2021 (UTC))

@SurprisedMewtwoFace: The "badge" at Wikidata hadn't been updated. I have updated it, and that adds it to Category:Validated texts.

And before anyone asks, yes, in theory it is possible to generate a list (and or auto-fix) this case (Index proofread/validated but the mainspace page isn't in the correct category), but it's not a totally trivial task. Not least because not all indexes are connected to mainspace pages via Wikisource index page URL (P1957), so it would involve traversing the links at Wikisource, or making sure the Index:mainspace mapping is captured at Wikidata as a first step. Inductiveload—talk/contribs 13:52, 13 March 2021 (UTC)

@Inductiveload: Thanks so much! It's much clearer now. The only remaining issue is that it doesn't link directly to the title itself. It appears as "Gentlemen Prefer Blondes" (1926) and the link to the book still isn't active per se through Anita Loos' author page. I think this is because the edition was the 1926 one rather than the 1925 one. However, it should now be clear to everyone that the text is completely validated. (SurprisedMewtwoFace (talk) 14:01, 13 March 2021 (UTC))

@SurprisedMewtwoFace: It should be linked now from Author:Anita Loos. This should have been done right off the bat, but was evidently overlooked. I have also transcluded the copyright page to make it clearer why this is (1926) but the title page say 1925. Inductiveload—talk/contribs 14:10, 13 March 2021 (UTC)

@Inductiveload: Thanks! (SurprisedMewtwoFace (talk) 14:46, 13 March 2021 (UTC))

Funeral notice or a death notice

Would something as small as a 4 sentence funeral notice or a death notice be eligible to be stored here at Wikisource? I see where each entry for a dictionary has its own page, so size shouldn't be the reason for exclusion. A ruling at Wikimedia Commons agrees that they are ineligible for copyright even after 1964, since they consist of publicly available information and do not pass the threshold of originality, as opposed to an obituary which surpasses the threshold of originality. --RAN (talk) 06:51, 8 March 2021 (UTC)

Generally death notices and funeral notices are a compilation, and one would be an excerpt, so the way that I have been handling these at this time is transcribing onto the author talk page, and documenting with source. We can do data collection onto author talk pages and evidence curation, without the requirement to proofread, etc. I did do some differently years ago, and stopped as it wasn't a good practice. — billinghurst sDrewth 12:00, 8 March 2021 (UTC)

An isolated four-sentence dictionary entry without the rest of the dictionary entries and without the index page for the whole dictionary would also be considered an excerpt and would get deleted. So I also do not think that such isolated short notices are elligible for inclusion. --Jan Kameníček (talk) 00:27, 9 March 2021 (UTC)

• Is that because you consider funeral notices excerpts, or is it because of their brevity? I would assume an excerpt would be a single page of a book, or a single paragraph from page, something that cannot tell a complete story from start to finish. For instance, a newspaper issue is the entire publication, yet, each individual article can be read from start to finish, telling a complete story. See The New York Times for individual articles from the New York Times. If it is because of brevity, what is the minimum number of words that a Wikisource entry requires? The people who are transcribing the New York Times are transcribing very short advertisements. See The New York Times/1900/12/01/Advertisement—Pennsylvania Railroad. Do these need to be deleted? A funeral notice is a paid advertisement. --RAN (talk) 01:00, 9 March 2021 (UTC)

  No, because it is not an isolated excerpt, they are transcribing the whole newspaper.See my example with the dictionary above. --Jan Kameníček (talk) 01:24, 9 March 2021 (UTC)

So, what would make a funeral notice not an excerpt? A cluster of 5 funeral notices on a page? Or the entire newspaper? Why is a single article from a newspaper, not an excerpt from that issue, yet an advertisement from the same issue is? The New York Times ad I showed is one of 8 excerpts from a 13 page issue of the The New York Times from 1900. The rules here can be can be mind-numbingly opaque based on fuzzily worded policy pages describing "excerpts" and "annotations". Good intentioned people keep coming to diametrically opposed conclusions when interpreting them. --RAN (talk) 01:36, 9 March 2021 (UTC)

@Richard Arthur Norton (1958- ): This is a cousin discussion to Wikisource:Scriptorium#Inclusion_criteria_for_articles above.

My personal feeling is that a "self-contained article" within a larger work like a newspaper, magazine or journal is acceptable, and the entire issue should not need to be transcribed just for one article of interest.

I would (again personally) say that generally, a single death notices is a section of an article, and the whole "Births, Deaths & Marriages" section of the paper is the "unit" which we'd proofread and transclude at. Creating an entire mainspace page for every 4-sentence entry is too much overhead, IMO. Even language dictionaries often chunk into sections like Chambers's Twentieth Century Dictionary 1908/A Adhere and Dictionary of the Swatow dialect/ba rather than a page-per-entry. {{anchor}} can be used to link directly to a single entry within a page if needed.

You are right that the guidance on such items is badly lacking, but I'm hoping that at the every least the aforementioned discussion can result in something useful. Inductiveload—talk/contribs 08:45, 9 March 2021 (UTC)

A complete article from a newspaper is not considered an excerpt. There are still excerpts around, and I definitely created some in the early days, though as Inductiveload said they are problematic in nature, especially where someone starts to work on a whole newspaper. Transcribing and recording parts of compiled works is a laborious set of tasks to do properly.

Getting specific text to explain it and getting people to read it is always the challenge. Anyone who wants to build Help:Newspapers or Help:Transcribing newspapers with or without scans is always welcome here to transfer our snippets. Volunteers who can build good help pages has always been a shortfall here, we all prefer to work on the content. — billinghurst sDrewth 11:26, 9 March 2021 (UTC)

• Perhaps we need to distinguish between what is absolute policy (!Wikilaw concerning copyright), and what is considered proposed best practices. !Wikilaw can be weaponized to harass editors. Since I have been organizing the newspaper articles, I can see a half dozen ways that articles are named, and an equal number of ways that newspaper articles are formatted, and just as many ways that newspaper articles are aggregated into portals. I am sure that the researcher only cares about the information contained within the articles, the rest is just esthetics. I think editors here are still experimenting with the most useful and practical way to display information, so that it maximizes usefulness to the end user. --RAN (talk) 13:34, 9 March 2021 (UTC)

@Richard Arthur Norton (1958- ): Not allowing "excepts" is covered by policy under WS:WWI#Excerpts. However (big however) this policy is enormously unsatisfactory when it comes to things like newspapers, periodicals and collective works. Some attempt is being made to normalise it in Wikisource:Scriptorium#Inclusion_criteria_for_articles.

The problem is, that, as policy, the phrasing ...are generally not acceptable is problematic as it is subjective and not clarified why it is "generally" true, but not "absolutely" true. Furthermore When an entire work is available as [scan], works are considered in process not excerpts. is unclear as to whether this permits transcluding work to mainspace in any state, as long as there's a scan, or just means unfinished work can sit in Page namespace indefinitely.

I think editors here are still experimenting with the most useful and practical way to display information, so that it maximizes usefulness to the end user: I think this is absolutely the case. Inductiveload—talk/contribs 13:58, 9 March 2021 (UTC)

• Yes, I have seen poorly worded, subjective rules, weaponized to harass contributors. As I pointed out before, if you let people with admin rights delete what they do not like, you end up with selection bias. At one point every portal I created was deleted because an admin person said they were not notable enough for a portal "only famous people get portals" was the reason for deletion. --RAN (talk) 22:45, 13 March 2021 (UTC)

  @Richard Arthur Norton (1958- ): I am sick of this bullshit. You keep making all these false accusations. Show me where I said that? Show me where I did that? Special:DeletedContributions/Richard Arthur Norton (1958- ) shows ONE deleted portal, and that is because there is an author page. You have not been harassed, your edits are patrolled just as anybody else's edits are patrolled. — billinghurst sDrewth 14:29, 14 March 2021 (UTC)

  Just agree to an interaction ban and let someone else patrol my entries. There are plenty of other people patrolling, that do not have the negative interactions we have had. --RAN (talk) 16:26, 14 March 2021 (UTC)

  You will not be getting an interaction ban with Billinghurst so you might as well stop beating that horse. Legitimate complaints about their admin actions can be presented constructively, but the over the top accusations and assumptions of bad faith have to stop. They are staring to tip into harassment territory in their own right regardless of what merit your underlying complaints may have. --Xover (talk) 16:34, 14 March 2021 (UTC)

Box-Car Children (1924) now completely proofread

@Kathleen.wright5: I have noticed that you have been doing some validating and other work on https://en.wikisource.org/wiki/Index:TheBoxcarChildren1924.djvu This book is now completely proofread and all pages have been input. If you (or anyone else) is interested, the book could use validation for the parts of it that have not yet been validated. Thanks for all your help! (SurprisedMewtwoFace (talk) 21:57, 14 March 2021 (UTC))

Why not acceptable

Can someone other than Billighurst explain why this letter is not acceptable at Wikisource: letter concerning Louis Julius Freudenberg I (1894-1918) a letter held in a state archive.

See https://wwwnet-dos.state.nj.us/DOS_ArchivesDBPortal/WWICardDetails.aspx?CardID=809 --RAN (talk) 21:33, 1 March 2021 (UTC)

Well, I do not consider it suitable to exclude anybody from expressing an opinion on inclusion or exclusion of a work.

We already have Billighurst's opinion, he wrote "out of scope" when he made the move. I am looking for third party opinions. --RAN (talk) 23:25, 1 March 2021 (UTC)

As for the question itself: The letter was published (on the New Jersey State Archives website) and is verifiable as WS:What Wikisource includes demands. As for copyright licence, I would use {{PD-US-unpublished}}, but I am not an expert on US copyright law. The only reason why its inclusion might be doubted is its content. As it has no artistic value, only its documentary value can be considered, which is not very high, but there is imo some, considering that the person was not just a no-name soldier, but was an object of some local press articles at least. I tend to agree with its inclusion, but I am curious about other opinions too. --Jan Kameníček (talk) 22:28, 1 March 2021 (UTC)

"Artistic value" is subjective, we need objective reasons to include or not include, otherwise people's personal biases skew what we keep and what we discard. What is art, and what is not art is very personal, and museums and archives, are now confronting their past biases. Don't you think it would have been better for Billighurst to bring the move up for consensus before moving the letter from mainspace? Is Wikisource designed solely to contain art? Then why are we attempting to transcribe the entire New York Times up to 1925? You write that the letter's "[value] is not very high", again why aren't we taking the word of the archive that chose to preserve it, and display it at their website. We have too many subjective rules that are selectively enforced. That lets the rules be used to punish people that you don't like or don't get along with. I understand Billighurst has admin rights, but that shouldn't preclude him from gaining consensus first. He shouldn't be enforcing his personal preferences, or punishing people with audits for speaking out against him. --RAN (talk) 22:35, 1 March 2021 (UTC)

That a work is at an archive, does not in itself does not make it notable for reproduction at enWS. The criteria of WS:WWI is quite more specific than that.

Documentary sources are characterized by one of two criteria:

• They are official documents of the body producing them, or
• They are evidentiary in nature, and created in the course of events.

If we apply the presence of a document in an archive, then we can reproduce every probate document that has been registered, every land transaction, every police record, we can include every convict court case, and that becomes an unholy mess to manage and curate, and makes our scope extraordinarily large. WWI says we take published works, then expands into what are documents related to more issues of notability. We are not trying to be a family history website, we are not trying to be a local history website, we are not trying to be an archive office.

Verifiability relates to source, and only applies after scope has been confirmed. The copyright issue is another issue in itself, and not one that I was addressing. — billinghurst sDrewth 06:13, 2 March 2021 (UTC)

First of all: "They are official documents of the body producing them" and "They are evidentiary in nature, and created in the course of events." are both listed under Works created after 1925, I think you just randomly cut and pasted two phrases from the page. This work was produced in 1919. It is not a probate document, or a land transaction, or a police record, or a "convict court case", whatever that is. Again, why aren't you getting consensus, before moving documents, since this is not any of the forbidden documents that you just described? You again are using poorly worded, subjective interpretations, and selectively applying them, which feels like harassment. --RAN (talk) 06:26, 2 March 2021 (UTC)

•  Comment I think this letter is certainly in scope in my book. It's "evidentiary in nature, and created in the course of events", where the event in question is World War I. IMO, this is an interesting document in its own right. Not earth-shattering, but a certainly a small window into the past. Since WS:WWI is clear on pre-1926 works with Most written work (or transcript of original audio or visual content) published (or created but never published) prior to 1926 may be included in Wikisource, so long as it is verifiable. Valid sources include uploaded scans and printed paper sources., that's rather moot. Eloise Lindauer died in 1935, so there's no issue with the unpublished copyright side of things.
• As an aside, the work itself could really do with proper scan-backing to an index, and the wikilinks are a little OTT. But honestly, I don't mind the concept of wikilinks, and in fact I like the idea. Perhaps we can come to a different solution, for example making the default interwiki link colour in the text body closer to the surrounding text color? Or improve and promote Visibility.js, which allows the links to be unstyled entirely. Inductiveload—talk/contribs 08:38, 2 March 2021 (UTC)

So by your reasoning any document created prior to 1926 is acceptable? How many pre-1926 probate testaments would you like to see onsite? Do you think that is our scope? How about a receipt book from 1899. A store ledger from 1852 from the Victorian goldfields? What about the brand plate of a late 19th century piano? The most boring letter ever written by my great great grandfather of his trip to the UK from Australia in the late 19thC. I have those and never would have considered them in scope for WS. All written prior to 1926. They are documents or similar, not published works. The current modification to the policy came about re a discussion about notability, and was worded to address that we were not wanting non-notable documents per [10] The intent of that change was not to give pre 1923 and post 1923 documents that sort of differentiation that you are mentioning as year is clearly about PUBLISHED works.

Wikisource, as the free library that anyone can improve, exists to archive the free artistic and intellectual works created throughout history, and to present these publications in a faithful wiki version so that anyone may contribute added value to the collection.

— Wikisource:What Wikisource includes

If we are changing our scope to now have family history type information, and anything written anywhere prior to 1926. — billinghurst sDrewth 12:06, 2 March 2021 (UTC)

Noting the 2009 change to WS:WWI [11] — billinghurst sDrewth 12:11, 2 March 2021 (UTC)

WS:WWI is explicit, and it's the most policy-like policy we have. It's not RAN's fault if you don't agree with it. I might add that that what the letter of policies like "no-cross namespace redirects, even from Author: to Portal:" and "people with no extant works on site should be Portals" is also often used for unilateral action and complaints about people not adhering to the policies. Either what "policy" says goes, or it doesn't. It's not immutable, but, for now at least, it is there.

I'm not sure what that change in 2009 is relevant to. The salient part of the policy (emphasis mine) ("Any written work (or transcript of original audio or visual content) published (or created but never published) prior to 1923 may be included in Wikisource") has been there since 2007 (the referred discussion is Wikisource:Scriptorium/Archives/2007-09#Change_inclusion_policy), more than a year before your linked discussion. The change to the current "Most written" happened in in 2012 and the narrowing was apparently related to copyright, not notability. Furthermore, before that it said [Documentary sources] may range from constitutions and treaties to personal correspondence and diaries. We can argue whether or not this letter is truly under that definition, but there's certainly a case to be made that it is. I'd say, since WWI was kind of a big deal, it would be. It looks to me like this kind of thing has been explicitly allowed for over 15 years.

And, to be quite honest, all your examples sound fine to me, as long as they're scan-backed and formatted sensibly. I find it hard to get upset about too much material, only poor quality material. One's man's rubbish and all that. Ephemera is a whole, perfectly valid, historical field in itself. Inductiveload—talk/contribs 13:04, 2 March 2021 (UTC)

• Without digging too deep into every single argument presented in this discussion, I think I agree with Inductiveload's perspective on this. If an editor is putting up high-quality scan-backed and externally verifiable material, who am I to say that it doesn't count? There are literally hundreds of books with scans uploaded to WS that people arguably don't care about, so truly what is the difference between a published book and an archived letter for the purposes of establishing "notability"? As long as the allowance of ephemera doesn't make Wikisource un-navigable (which I don't think it would beyond its present state), I'm in support of the inclusion of any published work/archived work so long as there are editors to maintain it. -- Mathmitch7 (talk) 16:44, 3 March 2021 (UTC)

• Furthermore, "Published" is a very crude proxy for "notable/interesting/useful". Considering the immense quantity of published newspapers, periodicals of various degrees of specialism and general governmental/official effluent, I'd say that if something is interesting enough to someone that they'll spend the time and effort to present it (and well!) at WS, it's already proved that it's of more interest than, say, a classified ad offering a sewing machine for sale in Caspar, Wyoming. Inductiveload—talk/contribs 16:57, 3 March 2021 (UTC)

Oh, and noting Wikisource:For Wikipedians has a particular statement that there is a notability requirement for documentary sources, and this exists from page creation in 2011. — billinghurst sDrewth 06:11, 15 March 2021 (UTC)

• I'm going to be singularly unhelpful and say that I agree entirely with Billinghurst in this individual case, but that I don't think it is reasonably possible to read that out of WS:WWI (e/c, but otherwise essentially what Inductiveload said above). It's written as an unholy hybrid of a user help page and a policy, and as such it fails at being either. Unless you were intimately involved in the discussions around the time it was written it is impossible to figure out how it applies to the situation at hand. For example, coming at it without context, a straightforward reading of that page would lead one to believe that anything created prior to 1926 is in scope so long as it is 1) verifiable (anything held in an archive is verifiable for this purpose) and 2) public domain. That is of course ludicrous as an inclusion criteria, but it is the plain reading of that page.
  On that basis I'm tempted to be unconstructive and say that until we actually fix our inclusion policy we really can't make any blanket statement and have to have a discussion for every single work individually. Very tempted. But as that would achieve little beyond annoying almost everyone I'm going to settle for my standard plea that we give some priority towards developing proper policies for the core issues, scope and inclusion—exclusion criteria foremost among them. I'm happy to help drive that, but only so long as the community actually wants a real policy framework for this. Based on every indication I have so far the community actively does not want more or better defined policy, and certainly has no appetite for the effort involved. I live in eternal hope that one day I will turn out to be wrong about that. --Xover (talk) 13:37, 2 March 2021 (UTC)

  While I agree that WS:WWI is far from perfect, it is the only guidance we have here at the moment and so until the community decides to change it we cannot blame anybody for following it. Of course that the community vote can overrule it in this (or any other) specific case, but it was very unfortunate that the letter was removed from the main namespace without a proper discussion at Wikisource:Proposed deletions, which is exactly what we have that page for. Had it been done so, we could be spared from a lot of bitterness now.

  So my suggestion is to return the work back into the main namespace for now, and delete it only if the community decides to overrule WS:WWI in this particular case.

  At the same time I absolutely agree with the need to refine WS:WWI, and although I do not think that the points raised by Billinghurst here apply to this specific case, I understand them generally and I will definitely take part in discussions trying to implement them into the rule in a sensible way. --Jan Kameníček (talk) 21:00, 2 March 2021 (UTC)

• First, I think it problematic that wasn't taken to Proposed Deletion instead of summary deletion. Secondly, WS:WWI doesn't support the deletion here, and I don't think it was intended to. I was told, I don't recall by whom, that the then-1923 line was good enough, because the time alone would be enough to keep people from posting vanity junk. I can't say I'm always stunned by the value of what other people chose to work on, and I lack any evidence that there's enough problem with people posting pre-1926 works problematically to justify arguing about which vintage works are acceptable.--Prosfilaes (talk) 08:16, 3 March 2021 (UTC)

  @Prosfilaes: The created work at enWS was never and still is unlabelled as being from an archive. I will note that the actual images are so labelled, though it was not images that are immediately evident, and Commons inclusion criteria is different from ours. So the action that I took has to be seen in that perspective. — billinghurst sDrewth 10:43, 3 March 2021 (UTC)

I still do not see a rationale based on !Wikilaw for you finding the document ineligible for Wikisource. So far you have only used the slippery slope fallacy, that if this letter is eligible, it will allow "every probate document that has been registered, every land transaction, every police record, … [and] every convict court case." I still get the vibe it has more to do with personal animus toward me after I complained that you were using your admin rights to enforce your personal tastes. --RAN (talk) 15:00, 4 March 2021 (UTC)

Interaction ban proposal

• Can I also suggest an interaction ban with Billinghurst, someone else should be patrolling my entries, if they need patrolling. Someone that does not have the record of enforcing their personal preferences as !Wikilaw. As I reported above he was imposing his THIS IS A DRAFT!!! as !Wikilaw and instead of acknowledging it, his response was to remove the THIS IS A DRAFT!!! title, instead of having a proper consensus !vote. As I also pointed out there is a half-dozen ways that newspaper entries are being presented, and there seems to be no rush to harmonize them. When there is a rush to change my entries I get the vibe that there is some personal animus involved, hence the inquiry about an interaction ban. I feel like I am getting a punitive audit from Billinghurst for bringing up the previous THIS IS A DRAFT!!! interaction. --RAN (talk) 22:16, 1 March 2021 (UTC)

• can we trout both of you?
• "So by your reasoning any document created prior to 1926 is acceptable? " i love the parade of horribles; could you please not do enforcement with RAN, he is mostly harmless, if obstreperous. how are you going to decide which manuscripts to include? you understand the Smithsonian is transcribing Freedman's Bureau log books, and LOC is transcribing Women's Party meeting minutes.
• but "!Wikilaw...THIS IS A DRAFT!!!" you are well aware this is how admins behave. you are lucky he did not make a filter just for you, as some admins do. if you could organize a task flow, with a newspaper (i.e. [12]) rather than clippings, or pitch in at POTM, you would have more sympathy. Slowking4 亞 Rama's revenge 23:40, 2 March 2021 (UTC)

• @Slowking4: I think I mentioned earlier on this page, there are half a dozen ways newspapers are being organized, and an equal number of ways clipped articles are being named and formatted. I have seen the New York Times portal, but even that is a hybrid mixture of formats and clipped articles using various naming schemes. What do you consider the best formatted newspaper collection at Wikisource? --RAN (talk) 23:59, 2 March 2021 (UTC)

• the fact that newspapers are a mess, with raw clippings from 2012, is not a good argument to spread the mess. i would like to see a migration from non-scan back clippings, to scan back works using chronicling america. https://chroniclingamerica.loc.gov/ a portal sweep up of clippings is a start, but the scans need to be incorporated. Slowking4 亞 Rama's revenge 12:32, 3 March 2021 (UTC)

• @Slowking4: when has a filter ever been used to target non-vandals/spammers at enWS? Inductiveload—talk/contribs 09:58, 3 March 2021 (UTC)

• filters are routinely used to block youtube, stopping work when government documents use them as references. but i guess the "admin may i" works for you. it is not a leap, to imagine RAN being summarily labeled as a "spammer". Slowking4 亞 Rama's revenge 12:10, 3 March 2021 (UTC)
  • That filter was actually turned off, because, recently, Youtube links are becoming more common in WS works than in spam (and we have more granular abuse filters now, partly because a fixed blacklist is a blunt tool, so we can target actual spam more effectively). This adversaralism is ugly. Just because you've received bans elsewhere doesn't mean all admins here are power-tripping maniacs. And no one is calling anyone a "spammer", and the suggestion that they might is, IMO, entirely unjustified in this context. Inductiveload—talk/contribs 12:46, 3 March 2021 (UTC)

    FWIW technically Mediawiki:Spam-blacklist and Special:AbuseFilter are different things. So trying to add some clarity and difference, not stir the pot. — billinghurst sDrewth 10:49, 4 March 2021 (UTC)

    • filters and blacklist are opaque and summary processes. veteran editors are astonished at their appearance. you would need to provide a feedback page for each one, and some oversight and reporting. for instance, where and when did you decide to change youtube blacklist? you gave no indication this was possible in the past. this "adversarialism" started with abusive admins bearing false witness. RAN has been abused elsewhere; therefore, it is entirely possible he would be abused here as well. Slowking4 亞 Rama's revenge 05:09, 5 March 2021 (UTC)
• if i might suggest a way forward: put on a maintenance category - "newspaper clippings not scan backed"; organize clippings by work portal, listed by date; find and upload scans of newspapers to commons; create indexes that include newspaper clipping, "newspaper clippings scan backed"; strongly encourage editors to follow this process. this would be better than exclusion criteria, by pivoting editors to doing better work. Slowking4 亞 Rama's revenge 18:27, 5 March 2021 (UTC)

Visibility now a gadget - long s and external links

The "Visibility" gadget that controls the appearance of {{ls}} in mainspace has been made into a proper gadget. It has a new function: external wikilinks can be toggled on and off as well. This is to try to take some of the heat out the Wikilinking discussion by allowing readers to choose to see the light blue external links (i.e. Wikipedia, Wiktionary, Wikidata, etc) or not.

The gadget is not enabled by default at present. Documentation is here: Help:Gadget-Visibility. Inductiveload—talk/contribs 20:26, 8 March 2021 (UTC)

I do like the tool, but I would also like to point out that in fact it is useful only for users contributing to Wikimedia projects (who usually sign in and so can switch the gadget on) and is of no use to common readers. The tool could be useful to readers only if it were possible to turn it on no matter if they have their wiki account or not. --Jan Kameníček (talk) 00:18, 9 March 2021 (UTC)

Second, there should be a toggle on pages with long s to turn it on and off regardless of whether a user is logged in. However, this may be more challenging to implement in practice. Languageseeker (talk) 03:15, 9 March 2021 (UTC)

Making this available to all users, logged in or not, by default is as simple as adding default, but I didn't want to do that without any consultation. Inductiveload—talk/contribs 06:53, 9 March 2021 (UTC)

• In 15 years of reading in the Wikipedia Universe, I have never felt compelled to click on a blue link, unless I had an interest in learning more. I don't think an unregistered user would feel any more, or less, compelled than me. I think the people most interested in not seeing blue links are long term registered users. --RAN (talk) 13:26, 15 March 2021 (UTC)

Tech News: 2021-11

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Recent changes

• Wikis that are part of the desktop improvements project can now use a new search function. The desktop improvements and the new search will come to more wikis later. You can also test it early.
• Editors who put up banners or change site-wide JavaScript code should use the client error graph to see that their changes has not caused problems. You can read more. [13]

Problems

• Due to database issues the Wikimedia Beta Cluster was read-only for over a day.

Changes later this week

• The new version of MediaWiki will be on test wikis and MediaWiki.org from 16 March. It will be on non-Wikipedia wikis and some Wikipedias from 17 March. It will be on all wikis from 18 March (calendar).

Future changes

• You can add a newline or carriage return character to a custom signature if you use a template. There is a proposal to not allow them in the future. This is because they can cause formatting problems. [14][15]
• You will be able to read but not edit 12 wikis for a short period of time on 23 March at 06:00 (UTC). This could take 30 minutes but will probably be much faster.
• You can use Quarry for SQL queries to the Wiki Replicas. Cross-database JOINS will no longer work from 23 March. There will be a new field to specify the database to connect to. If you think this affects you and you need help you can post on Phabricator or on Wikitech. PAWS and other ways to do SQL queries to the Wiki Replicas will be affected later. [16]

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

23:22, 15 March 2021 (UTC)

Nearing 500k: how do we advertise/celebrate this?

Since we are presently at 498,752 texts in English, that means we are close to a big round anniversary-style number. Do we have any idea how we can use that success to promote Wikisource? A one-off logo? A blog post at Diff? —Justin (koavf)❤T☮C☺M☯ 18:20, 15 March 2021 (UTC)

Both, I think the more publicity the better. Also, could we borrow the idea of banner showing how many pages were proofread this month from the French Wikisource? It always makes me feel so impressed. Languageseeker (talk) 18:49, 15 March 2021 (UTC)

Actually, we don't have that anywhere near that many texts because the count includes sub-pages, most of which are chapters of a text. I haven't been able to think of a way to accurately count how many texts we hold because it needs to be all the root pages that aren't disambiguation pages and some subpages where the work is an anthology that reprints works that were separately published. Then there's the problem of how to count journal and newspaper issues articles. Beeswaxcandle (talk) 06:08, 16 March 2021 (UTC)

Can we download from HathiTrust?

I've previously downloaded books from HathiTrust, but they've changed their setup, so I don't know how to work around their system anymore. Is anyone able to download full books from HathiTrust anymore?--Prosfilaes (talk) 23:50, 15 March 2021 (UTC)

If you do an Inspect element, go to the network tab, and then find the image, you can download it image by image using a batch downloader. The image files are sequentially numbered. Languageseeker (talk) 02:16, 16 March 2021 (UTC)

You sure? I've downloaded them image by image before, but now I'm only getting blobs.--Prosfilaes (talk) 02:20, 16 March 2021 (UTC)

What's the book? Languageseeker (talk) 02:30, 16 March 2021 (UTC)

@Prosfilaes: The web view is indeed now using some kind of "blob" mechanism, so right click and copy image address doesn't work any more. However, the image still downloads as an image. The URL scheme is https://babel.hathitrust.org/cgi/imgsrv/image?id={ID_HERE};seq={N};size=10000;rotation=0, where you set size to something absurd like 10000 to make it give you the biggest size it has.

The Data API also still works (this is my preferred way as it's easier to work out how many images you need and an easier set of endpoints for things like page-number manifests. But you need to register an "UoM Friend" account to get your API keys (free and not very hard).

I have updated Help:Image_extraction#Hathi_Trust since it previously recommended right-click/copy-address for single images. Inductiveload—talk/contribs 07:39, 16 March 2021 (UTC)

Not sure if I'm missing something as to why it hasn't been mentioned, although the HathiDownloadHelper tool still works for me? https://sourceforge.net/projects/hathidownloadhelper/ Nickw25 (talk) 10:40, 17 March 2021 (UTC)

Upping the ante: All additions to Template:New texts to have been entered into WD

I think that it is time that the community upped the ante on new texts. I think that we should expect that they all have a Wikidata item, and that we look to have a means to record the WD item into the template. At this time, I don't expect the addition to do anything, though it gives us the scope to work on it, and maybe we can have a means to just say {{new texts/Qnnnnnnn}} and let it just pull the data. I think that we need to be setting up WS to allow better queries and interactions with WD, and this seems a good place to start. Especially if it helps us to link additions up to our twitter feed.

Now to think about whether there is value or ability to migrate our validation dates into a field in WD. — billinghurst sDrewth 00:35, 18 March 2021 (UTC)

Template:Potus-eo should be aligned to Template:Header

We have been actively updating and modernising our header template< and this spin off needs to be undergoing similar updates, either to be converted to be a subsidiary template of "Header" or it needs to have the components utilised in header added to it. It definitely needs to have Module:Plain sister aspects added to it so that its interwiki links can display automatically. It should have its download aspects compliant. — billinghurst sDrewth 01:34, 18 March 2021 (UTC)

Index page numbering for unnumbered pages

User:Billinghurst and I have a disagreement on his talk page about what to do with how the front matter pages at Bobbie, General Manager (a work that I proofread in full) should be labelled at Index:Bobbie, General Manager (1913).djvu. He calls the labels I give in my index pages, such as "copyr", "dedic", "title", etc., as "manufactured labels" and "butt-ugly" and other things like that. He wants every page that's not numbered in some way by the book itself to be labelled as "—" instead of having a label that tells what it actually is.

I absolutely disagree with that. I'm not sure how you can consider such things as looking atrocious as he does, since I consider them to look extremely nice on an Index page and in transcluded pages. But I suppose aesthetics and beauty in the end are matters of subjective opinion, which is why I decided to bring it up here. He also brought up that User:ShakespeareFan00 is a user who shares my opinion.

If he wants to continue to assert that Index page labels should not be a matter of personal preference, I propose that we instead come to a consensus on how policy (not guidelines pages or "advice") should deal with Index page labels. I'm not sure what policy page would be sufficient for asserting this as a rule though. So, would you guys vote Support to allow Index pages to label copyrights as "copyr", frontispiece pages as "frontis", covers as "cover", or title pages as "title", for example? Or would you vote Oppose, that we should disallow this, and assert that every page that's not numbered should be labelled as "—"? PseudoSkull (talk) 16:06, 18 March 2021 (UTC)

This matter is covered by the "general recommendations" in the yellow box on Help:Index pages#Parameters, which have been in place since 2013. Changes to these conventions would need to be the subject of a formal RFC as they have been in place for so long. Beeswaxcandle (talk) 16:44, 18 March 2021 (UTC)

As per Help:Index pages, pages with text on them should not be labeled "-". I'm not terribly stressed about how they should be labeled, but I find it very important that we can tell blank pages from pages with content on them at a glance of the index page.--Prosfilaes (talk) 01:40, 19 March 2021 (UTC)

I think that it's better to have standard label for such content than just -. They're not blank pages, but crucial pieces of information. The more specificity the better in my book. If we are to reproduce books exactly, then we need to include all the material. Copyright pages have a very specific and necessary purpose. Languageseeker (talk) 02:20, 19 March 2021 (UTC)

My approach recently , has been to consider pages before numbered pages (1,2,3) etc. to be numbered using lowercase roman (even when not actually done so in the book), based on the convention adopted by countless books that do numbers these pages directly. The exceptions to these are:

• "Title" which is identified as such in an imported pagelist from an external source and image-plates, which often sit outside the numbered page range. (I also unless there is a half-title or other numbering to go on consider Title and i to be the same page in a work, the pages following the title to be ii , and so on.
• "Adv" for advertisement material, which in many works is not numbered. (It can also be considered to be numbered with lower-case roman numerals in some works.)

ShakespeareFan00 (talk) 08:15, 19 March 2021 (UTC)

I am also looking through the pagelist's I contributed a while back as part of cleanup effort, with a view to updating them to any new conventions, that are put in policy. However, I would like a consensus on ONE set of guidelines so I am not changing things multiple times as different contributors express different views. ShakespeareFan00 (talk) 10:17, 19 March 2021 (UTC)

@ShakespeareFan00: I don't think you need to be concerned about your past indexes. I for one am grateful you did them at all, and I couldn't care less (at least for my uploads that you generously pagelisted) if the half-title is "iii" or "Half-title". I too will follow whatever rules are put in place, but for now at least, you seem to be well within accepted (and documented) practice. If a review of files is needed, it's not only your files!

Obligatory plug: I have also suggested an addition to the <pagelist/> tag to allow an auxiliary label for this kind of thing: phab:T274740. Comments welcome on it being a terrible idea and why, preferred syntax, etc, etc. Inductiveload—talk/contribs 10:42, 19 March 2021 (UTC)

• The timing isn't best for digging into this IMO (there are a few too many other unsettled matters), but I think it's an issue we should devote some attention to at some point. We're using these for multiple purposes, and some of those purposes are in direct conflict. In addition to a solid helping of personal preference, which of these aspects you find most important will determine what approach makes the most sense to you. Mapping from printed page number to index-into-file numbers may be the original purpose, but labelling pages for easy navigation by proofreaders and providing link targets on transclusion are not inherently invalid concerns. All of this can be solved through technical measures, but we need to figure out which aspects are worth solving and what that solution looks like. Nothing is free and some things may be prohibitively expensive. --Xover (talk) 12:31, 19 March 2021 (UTC)

If Billinghurst wants to start an RFC campaign he is more than welcome to. But since this has already been established practice since 2013 (and I didn't realize that when I made this Scriptorium topic), I refuse to start any RFC discussion myself, as I think making any change at all would be a waste of time. Besides just me, anyway, it looks pretty likely that community consensus won't be on his side on this one. PseudoSkull (talk) 13:21, 19 March 2021 (UTC)

Call for review, comment and discuss my PhD thesis on Wikimedia movement

Hello,

Just a short message to call people interested to review, comment and discuss my PhD thesis on Wikimedia movement. All the best, Lionel Scheepmans (talk) 19:35, 19 March 2021 (UTC)

Which dynamic layouts are actually used?

As part of the ongoing cleanup of the JS that provides Dynamic Layouts, I'd like to canvass opinions on which of these people actually use:

• Layout 1: default (full width with narrow gutter for page numbers)
• Layout 2: 36em column (similar to itWS and frWS)
• Layout 3: wide right gutter, header on right (but seems this isn't working: the header is off the screen for me). A little similar to deWS, but they have a much more data-heavy, vertical, header format (e.g. Ein Friedensstörer)
  • Only 5 pages set "Layout 3" as default
• Layout 4: identical to Layout 2, but with a width 540px (this is not ideal, as it encodes a fixed concept of pixels:font size,. At the common default of ~16px=1em, it's about 35em, so functionally the same as Layout 2). Notably, this will look very constrained if a visually-impaired user has set a higher default font size.
• Proposed Layout: Appears similar to Layout 3, but with a full-width, working header. Confusing name, since it's been "proposed" for nearly a decade. No-one appears to set this as default.

My personal feeling is that we should:

• Replace Layout 3's CSS with the Proposed Layouts, then scrap the Proposed Layout.
• Scrap Layout 4 as redundant to Layout 2

Inductiveload—talk/contribs 10:42, 1 March 2021 (UTC)

Layout 2 was the one I've used most often if I have set one.

I also note various layouts set up directly using <div class="prose">..<div class="pagetext"> and others from earlier periods of Wikisources development, are there plans to deprecate those so they can be removed?

Additional layouts might eventually be needed for things like playscripts.. (see {{stagescript/s}} for example), The intent with that template was to eventually make it so the script formatting could be changed by user preference, like with dynamic layouts. ShakespeareFan00 (talk) 11:20, 1 March 2021 (UTC)

@ShakespeareFan00: <div class="prose">..<div class="pagetext"> are separate, but allowing dynamic layouts to work on non-scan-backed pages is also on my hit-list, but it will have to come later. At that point those classes would be fully obsolete.

Additional layouts are fine (by me, at least), but we can come to that later. For now, I just want to know if we can trim the existing layouts. Inductiveload—talk/contribs 11:29, 1 March 2021 (UTC)

I passionately loathe all except No. 1 to the point that I don't work on texts with any of the others forced on me. I don't believe that we should be dictating to the end user a constrained width of what's displayed. By all means, have a couple available for readers to use as they choose. Every work that we make available should work and behave under Layout 1 (including playscripts). It doesn't help that I find the others to be ugly and stale in their design. Also, sidenotes are bad enough in Layout 1, they're dreadful in the others. Beeswaxcandle (talk) 17:37, 1 March 2021 (UTC)

@Beeswaxcandle: You are absolutely right that all works should render as well as possible in all layouts. Default layouts are one thing, but if a work is actively broken in any layout, that's a problem. To bang on my one of my many favourite drums, making sure that things work in both Layouts 1 and 2 is the bulk of the work of ensuring a work can export (because layout 2 is not far from the size of the content on a mobile/e-reader screen.

BTW, you can disable the ability of pages to use {{default layout}} to override with "Allow pages to override my dynamic layout preference on a case-by-case basis" in your gadget prefs. Making this a user toggle next to the layout selector is on my list. Inductiveload—talk/contribs 08:19, 2 March 2021 (UTC)

I use Layout 2 frequently for poetic and dramatic works, where (a) a margin constraint is necessary to keep the text alignment, and (b) serifs are almost necessary to be able to distinguish I, l, and 1 throughout the text (in various combinations as words and abbreviations, e.g., III. vs. Ill. or Roman numeral 3 versus the abbreviation for Illiad). But otherwise, I do not apply layouts at all. --EncycloPetey (talk) 01:20, 2 March 2021 (UTC)

OK, so before this gets archived, any objections to the proposal above, viz.:

• Replace Layout 3's CSS with the Proposed Layouts, then scrap the name Proposed Layout.
• Scrap Layout 4 as redundant to Layout 2

This leaves us with the following three globally-enabled choices:

• Layout 1/2 as they are
• Layout 3 will change to get a full-width header, the body is the same as it is now

Inductiveload—talk/contribs 10:47, 22 March 2021 (UTC)

It is a start. This might be moved to the proposal section. CYGNIS INSIGNIS 11:37, 22 March 2021 (UTC)

Vertically aligned text

I have been working on Page:EB1911 - Volume 28.djvu/409 which contains a table with text on its side. I thought about using Template:Vrl but it places the text sideways with the start of the text orientated towards the top. The EB1911 table has text orientated with the start towards the bottom. I have a work around in the table (diff) using style="writing-mode: sideways-lr;", but has anyone written a template to do this?

This is a vrl example

This is a"sideways-lr" example

-- PBS (talk) 15:01, 16 March 2021 (UTC)

I would just make it horizontal ... CYGNIS INSIGNIS 16:51, 16 March 2021 (UTC)

@PBS: So far as I know we don't have any existing template for this (I had a similar case recently and went looking). We should probably have one that allows arbitrary rotation in degrees (i.e. using CSS transform: rotate()), both for flexibility and because the CSS writing-direction is primarily designed for non-Latin scripts rather than visual formatting (which is what we usually deal with on enWS). --Xover (talk) 18:30, 16 March 2021 (UTC)

There is {{Rotate}}, but it has problems. I don't recommend reproducing most rotated text, rather reserve it for when the special effect is required. We're not aiming to identically reproduce all the nuances of a restricted-size printed page onto a webpage that needs to flow and cope with all sizes of displays. Beeswaxcandle (talk) 04:49, 17 March 2021 (UTC)

yeah, i would go horizontal also. all rotated text does is make the reader crane their necks. Slowking4 亞 Farmbrough's revenge 15:32, 22 March 2021 (UTC)

File:Hans Holbein the younger (Volume 1).pdf

Would it possible for this to be replaced with a djvu scan to match the volume 2 uploaded? 88.97.96.89 09:29, 22 March 2021 (UTC)

here you go c:Help:Converting PDF to DjVu --Slowking4 亞 Farmbrough's revenge 15:35, 22 March 2021 (UTC)

Tech News: 2021-12

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Recent changes

• There is a Wikipedia app for KaiOS phones. They don't have a touch screen so readers navigate with the phone keys. There is now a simulator so you can see what it looks like.
• The reply tool and new discussion tool are now available as the "Discussion tools" beta feature in almost all wikis except German Wikipedia.

Problems

• You will be able to read but not edit twelve wikis for a short period of time on 23 March at 06:00 (UTC). This can also affect password changes, logging in to new wikis, global renames and changing or confirming emails. This could take 30 minutes but will probably be much faster.

Changes later this week

• The new version of MediaWiki will be on test wikis and MediaWiki.org from 23 March. It will be on non-Wikipedia wikis and some Wikipedias from 24 March. It will be on all wikis from 25 March (calendar).
• Syntax highlighting colours will change to be easier to read. This will soon come to the first wikis. [17]

Future changes

• Flagged revisions will no longer have multiple tags like "tone" or "depth". It will also only have one tier. This was changed because very few wikis used these features and they make the tool difficult to maintain. [18][19]
• Gadgets and user scripts can access variables about the current page in JavaScript. In 2015 this was moved from wg* to mw.config. wg* will soon no longer work. [20]

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

16:53, 22 March 2021 (UTC)

Question about completed texts appearing under "New Texts"

What factor allows for texts to appear under "New Texts" once they have been completely proofread? I noticed that https://en.wikisource.org/wiki/The_Box-Car_Children didn't appear under New Texts once we completed proofreading of it. Also, https://en.wikisource.org/wiki/Gentlemen_Prefer_Blondes_(1926) didn't appear yet, even though it's been completely validated. Just curious whether New Texts is automated or requires some kind of additional input. (SurprisedMewtwoFace (talk) 19:40, 22 March 2021 (UTC))

While the Inex of "The Box-Car Children" shows that all of the individual pages have been proofread, the book has not been assembled from those pages. If you look at the Main namesapce page of The Box-Car Children, all of the chapters links are red links. The pages have been proofread, but the book has not been assembled. --EncycloPetey (talk) 20:09, 22 March 2021 (UTC)

That explains "The Box-Car Children", but "Gentlemen Prefer Blondes" is clearly assembled, validated, and has working chapter links. (SurprisedMewtwoFace (talk) 20:14, 22 March 2021 (UTC))

There is no appointed person taking care of all new works to feature them under New Texts. Once somebody has proofread a work, transcluded it into the main namespace and provided it with an appropriate licence, they can add it to Template:New texts. If they do not add it there, nobody else is likely to do so. As Wikipedians like to say: Be bold :-) --Jan Kameníček (talk) 21:09, 22 March 2021 (UTC)

Okay, I just added Gentlemen Prefer Blondes to the New Texts template. Hope this helps! (SurprisedMewtwoFace (talk) 22:04, 22 March 2021 (UTC))

@SurprisedMewtwoFace: Great! Only it should be added to the top, but I have already corrected it. You can also use the "display" parameter to show the exact title of the book if it differs from the name of the page. --Jan Kameníček (talk) 22:34, 22 March 2021 (UTC)

What to name the old version?

I moved the previous contents of The Barbarism of Slavery to HERE. I want to place it back in the main namespace but can't come up with a suitable title. The text is identical, but the old version focused on wiki linking the article to Wikipedia sources, and had no italics or any text modifications we use in proofreading. Suggestion are welcome. — Ineuw (talk) 21:21, 22 March 2021 (UTC)

The scan backed version is good and faithful to the original, I do not think we need the other one which is not scan backed and includes features not present in the original version. --Jan Kameníček (talk) 21:27, 22 March 2021 (UTC)

I don't want to destroy someone else's work.— Ineuw (talk) 21:40, 22 March 2021 (UTC)

Replacing a text by a scan backed version is not destroying, but improving. The old version comes from 2005/2006, and I am afraid this is the final fate of all contributions from that time. --Jan Kameníček (talk) 22:27, 22 March 2021 (UTC)

Thanks for the clarification. In that case, I will delete my sandbox copy.— Ineuw (talk) 23:03, 22 March 2021 (UTC)

Is it possible to add a Download button next to "transcription project" when the text has been Proofread or Validated?

Would it be possible for the small scan link template to generate a Download button once the pages have been transcluded? Right now, it almost appears as if no transcription projects are finished. Perhaps, the button could also be colored to indicate the status of the text: Red if not all the pages have been proofread, Yellow if all the pages have been proofread, Green if all the Pages would be validated. Also, once entire text has been validated would it make sense to change the text from "Transcription Project" to "Transcription Completed"? In this way, users would be able to directly download from an Author's page. Languageseeker (talk) 12:52, 24 March 2021 (UTC)

{{small scan link}} should be removed once the work is proofread, so I'm not sure this exact proposal would add much value.

However, if we were better about setting the proofread (Q20748092) and validated (Q20748093) badges, we could theoretically get this information for the mainspace pages (to a max of 400 links per page, but less than that if the page loads other WD items).

Adding a download button on demand is pending phab:T275003 (though you can fake it with a link directly to the files like {{export}} does, you don't get the pretty new export dialog). Inductiveload—talk/contribs 13:18, 24 March 2021 (UTC)

I've seem plenty of examples of small scan link buttons not removed. Instead of asking users to manually remove them which creates more work for users, wouldn't it make more sense to automatically transform them into a download button? Languageseeker (talk) 13:28, 24 March 2021 (UTC)

No, because a work can have had its index be fully proofread and even validated, but that does not mean that the text has been assembled in the Main namespace for download. The proofreading of the pages from the index and the preparation of a complete text for download are two separate processes. And Inductiveload is correct, the small scan link is meant to be temporary and should be removed once the work has been fully transcribed. --EncycloPetey (talk) 20:31, 24 March 2021 (UTC)

Good points, what about changing the small scan link to download link that would generate the same Download button as on the transcluded page. In that way, it would be possible to directly download a work from the Author's book instead of having to go to the a different page just to get a download link. Languageseeker (talk) 22:29, 24 March 2021 (UTC)

That would only work in some situations. Sometimes the listed work is a poem inside another larger work. Sometimes the link takes the reader to a versions page where more than one version of the work is available. Sometimes, a work is moved and the link on the Author page is a redirect. Each of those situations is common and can generate bad results. Having the Download link on the page for the work itself makes the most sense. --EncycloPetey (talk) 00:02, 25 March 2021 (UTC)

Royal Naval Biography Marshall funny missing pages

There are some funny pages in Royal Naval Biography Marshall most of them seem to be dealt with Ok for example:

• Page:Royal Naval Biography Marshall v1p1.djvu/67 is to a page 40* which follows page 40

This page appears in the index page:

• Index:Royal Naval Biography Marshall v1p1.djvu

However just before page 40 are two pages which do not appear in the index page:

• Page:Royal Naval Biography Marshall v1p1.djvu/64 39*
• Page:Royal Naval Biography Marshall v1p1.djvu/65 39**

and consequently the article Royal Naval Biography/Montagu, George includes all the text, but it has strange pagination, It goes:

• 39, 40 , 41, 40 , 40* , 40** 41...

It is mostly correct—apart from the first instances of 40 and 41 which should be 39* and 39**. How can this be fixed so that the two missing pages appear in the index and the pagination is correct in the article? -- PBS (talk) 12:47, 27 March 2021 (UTC)

DoneFixed by adjusting the pagelist. Beeswaxcandle (talk) 17:31, 27 March 2021 (UTC)

Transcluding Index Comprised of Images

How would I transclude an Index comprised of individual page images? I can't seem to get something like <pages index="Brazilian_short_stories" from="5" to="16"/> to work. Languageseeker (talk) 04:10, 27 March 2021 (UTC)

Normally, an Index is created from a single file with all the pages together. I have never seen this approach done successfully for any work with more than 2 or 3 pages, nor have I seen it attempted where the Index page name has no file type extension, nor when the contained pages have file extensions (and these extensions differ from page to page). Normally, an Index is created from a single file with all the pages together. "Due to the extra complexity and other drawbacks of this process, this is not recommended for anything other than very short works: such as single pages or works of just 2-3 pages in length". as per Help:Index_pages vide "Using individual image files". --EncycloPetey (talk) 05:12, 27 March 2021 (UTC)

yeah, i would go take the individual images, and create a multipage pdf, using the publishing program of your choice, and then upload to commons. my library has ms publisher installed. but the process and prework is not documented well. Slowking4 亞 Farmbrough's revenge 23:12, 28 March 2021 (UTC)

I figured it out you need to use {{page|Image Name|num=Page Name} or {{page|Image Name}} to transclude. Wikisource is quite slow when handling large PDF, Chunked Uploader can't handle file over 2gb, and importing from the IA often results in 503 errors. You can also use the syntax <pages index="Index" from="First Page Image" to="End Page Image" /> Languageseeker (talk) 23:37, 28 March 2021 (UTC)

While using {{page}} is a solution, it is deprecated in most uses. Beeswaxcandle (talk) 08:24, 29 March 2021 (UTC)

Is there a category for specific typography?

I want to categorize articles containing manicules. See Category:Works with manicules, what larger category should that category be in? --RAN (talk) 03:02, 29 March 2021 (UTC)

This concept is not something that fits into our four-fold categorisation scheme (type of work; genre/subject; when published; licence), so before advising need to know how such a category would be used. [Maybe, even, why it would be useful here at Wikisource.] Beeswaxcandle (talk) 08:36, 29 March 2021 (UTC)

Tech News: 2021-13

Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.

Recent changes

• Some very old web browsers don’t work well with the Wikimedia wikis. Some old code for browsers that used to be supported is being removed. This could cause issues in those browsers. [21]
• IRC recent changes feeds have been moved to a new server. Make sure all tools automatically reconnect to irc.wikimedia.org and not to the name of any specific server. Users should also consider switching to the more modern EventStreams. [22]

Problems

• When you move a page that many editors have on their watchlist the history can be split. It might also not be possible to move it again for a while. This is because of a job queue problem. [23]
• Some translatable pages on Meta could not be edited. This was because of a bug in the translation tool. The new MediaWiki version was delayed because of problems like this. [24][25]

Changes later this week

• The new version of MediaWiki will be on test wikis and MediaWiki.org from 30 March. It will be on non-Wikipedia wikis and some Wikipedias from 31 March. It will be on all wikis from 1 April (calendar).

Tech news prepared by Tech News writers and posted by bot • Contribute • Translate • Get help • Give feedback • Subscribe or unsubscribe.

17:30, 29 March 2021 (UTC)

Has there been a Header function change

Am I remembering incorrectly, or did the {{Header}} used to add a navigation bar at the bottom of each page, so that readers would not have to scroll back to the top of the page in order to proceed to the next chapter / section of a text? --EncycloPetey (talk) 04:15, 30 March 2021 (UTC)

@EncycloPetey: Yes, it did. The footers are generated automatically by MediaWiki:Gadget-DisplayFooter.js (loaded by default as part of the Site Gadget) based on the data in {{header}}. It got broken during recent code cleanup here because it depends on the header having at least one next or previous link with the HTML IDs #headerprevious and #headernext. These IDs were lost in the cleanup so the footer script was just bailing out instead of doing anything. It's fixed now, but you may have to purge or null-edit a page before it becomes visible. --Xover (talk) 07:09, 30 March 2021 (UTC)

Adding portal categories for the Ottoman Empire?

I noticed Corps de droit ottoman/Preface needed to be linked to some pages.

From Portal:Index#Class_K_-_Law I see various portals, and Portal:Law/Subclasses has a lengthy list. However I'm not sure what country code to assign to the Ottoman Empire.

Note the work was written by a British man and published in the UK, but it concerns the Ottoman Empire. Only the English-language preface is on the English Wikisource because most of the work is in French. This is public domain under US copyright, but because of its UK copyright the French language content can't yet be uploaded to the French Wikisource (as that project considers British copyright law) WhisperToMe (talk) 05:33, 30 March 2021 (UTC)

You can find classification codes listed in Library of Congress Classification. Because the system was initially developing prior to World War I, and this is an update to that system, some listings will be grouped oddly. "Turkey" is the correct place for Ottoman Law, and the code is KKX. --EncycloPetey (talk) 22:11, 31 March 2021 (UTC)

scan quality in pdf

I notice the scan quality of the images in this pdf in a good deal worse than the images at the source. I initially thought it was an overly compressed djvu, where a 'c' character might be substituted for an 'e'; an example is the correction I made here which does show 'which' quite clearly at the source. CYGNIS INSIGNIS 05:03, 31 March 2021 (UTC)

They are indeed. But the original scan quality isn't great either. The scan is relatively low resolution, partially out of focus, and has been aggressively compressed. IA has then downsampled those images and recompressed them even more aggressively (4.4MB of scan images have been crunched down to a 672kB PDF; for reference a single scanned page alone should be somewhere between 600kB and 1MB). On top of that MediaWiki will extract the page using ghostscript and then reencode it into a 1024px wide "thumbnail" that is displayed in Proofread Page. At that point it's been recompressed three times, each time with generational loss. Frankly, it's amazing the results are legible at all.

If you need it for anything I can make you a DjVu that preserves as much quality as possible of the original scans, but as mentioned they aren't great to begin with. --Xover (talk) 05:49, 31 March 2021 (UTC)

It appeared on new texts, and I corrected a couple of scannos while reading it. The transcript is close to perfect so not sure it is worth the bother, although having a better djvu in situ may help with validating the work. CYGNIS INSIGNIS 09:15, 31 March 2021 (UTC)

FYI, this script might help, since, as well as adding handy links to pages, it adds an ability to load high-res images directly from the IA or Hathi into the Proofreadpage image pane: User:Inductiveload/jump to file. Inductiveload—talk/contribs 09:47, 31 March 2021 (UTC)

Petition of John Wilkes to the House of Lords

Do we have any experts in 18th Century cursive, who would be willing to check the short, single-page Petition of John Wilkes to the House of Lords, 1768, please? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:04, 31 March 2021 (UTC)

Done --EncycloPetey (talk) 02:58, 1 April 2021 (UTC)

New toys in 1.36.0-37 - work-specific CSS and fixed DjVu paragraphs

There are a few new features in ProofreadPage in today's deployment of MediaWiki 1.36.0-wmf37:

• Index page styles (work-specific styles): you can now make a CSS page at Index:Foo.djvu/styles.css (or .pdf) and it will be applied to all Page namespace pages and any transclusion using the <pages/> tag. There are some more details at Help:Page styles.
  • For example: Page:Sandbox.djvu/19 uses a style from Index:Sandbox.djvu/styles.css
  • A patch to add a link to the Index page to the associated stylesheet did not make it into this deployment, it'll land next week.
  • There is also a gadget that can be enabled to add some helpers to the the toolbar in the CSS edit mode: Help:Gadget-WikiEditorTools.

• DJVU paragraph control characters are now finally parsed properly, so you should see paragraph breaks in the text layers of new DJVUs. I don't think it will apply to existing DJVUs until someone runs an updater script on the server.

Inductiveload—talk/contribs 21:55, 31 March 2021 (UTC)

None ASCII characters in article titles

the Dictionary of National Biography uses date ranges in disambiguation extensions to article titles. There was pressure by some Wikipedia editors to change from dash to ndash to fit in with the Wikipedia policy on this issue. Wikipedia handles the issue by having ndash in the article title and a redirect with a dash.

The use of dashes was justified for Wikisource because it simplifies the URL and that no redirects were necessary.

I am currently working on Wikipedia linking article to Wikisource articles in the Royal Naval Biography I have just come across these two articles:

• Royal Naval Biography/M‘Farland, James
• Royal Naval Biography/O’Neill, Terence

The are no ASCII redirects

• Royal Naval Biography/M'Farland, James
• Royal Naval Biography/O'Neill, Terence

So should the articles to be moved, or should there be redirects, or are the names fine as they are and there is no need for redirects? -- PBS (talk) 14:59, 25 March 2021 (UTC)

@PBS: They should be moved. For punctuation like that (and dashes, quote marks, etc.) we use plain ascii in page names (nb. in page names: article titles are a different story). --Xover (talk) 15:24, 25 March 2021 (UTC)

I am using the term "article titles" to be the part of the URL that is used to make the URL unique. I think you are using the term page name to mean the same thing. -- PBS (talk) 15:29, 25 March 2021 (UTC)

Oh, yes, sorry; I should have been clearer. I just meant it as an aside in case there was confusion, but assumed you knew that so didn't want to over-explain. I should have just left it out. But, yes, the page name, which is what ends up in the URL, uses ascii punctuation. I'm just hedging around "article" since mainspace wikipages on Wikisource can contain zero or more "articles" (with or without non-ascii titles)—from a newspaper, for example—unlike Wikipedia where a mainspace wikipage by definition is "an article". --Xover (talk) 15:56, 25 March 2021 (UTC)

<-- There could be a lot of these how do I go about requesting a bot to run through the titles? -- PBS (talk) 12:24, 27 March 2021 (UTC)

Make a request at WS:BOTR. Beeswaxcandle (talk) 17:32, 27 March 2021 (UTC)

Small note: the turned comma in M‘Farland represents a superscript c (M^cFarland), not an apostrophe. So "McFarland" would be the better ASCII equivalent, perhaps with "M'Farland" as another redirect. (And M^c is just an abbreviation for Mac, as G^eo is for George, but that's a separate discussion.) Pelagic (talk) 23:04, 17 April 2021 (UTC)

Tweak archive settings for the Scriptorium

Currently the configuration for automatic archiving of the Scriptorium is set to archive threads in which there has been no new comments for 30 days, and to archive threads which are explicitly marked as resolved 31 days after the date they are marked as resolved. This means that in practical effect nothing ever gets archived by being marked as resolved.

In order to have the ability to clean out this sometimes a bit overwhelmingly long page I propose we change the interval for resolved sections to something more reasonable like a week, or possibly even 3 days. We rarely explicitly close threads here, and when we do it's the "Hey, how do I do this? / Here's how. / Ok, thanks."-type threads. Conversely, the threads that really need long-term visibility are either marked with the "do not archive until" tag (which lets you set an arbitrary future date before which the thread is ineligible for archiving) or, for things like the RFCs and proposals, once the discussion in the "Proposals" section is closed they should be posted as an announcement in the "Announcement" section where they will stay for an additional minimum of 30 days (the ordinary auto-archive interval).

Absent indications to the contrary my expectation is that this proposal is uncontroversial, so if there are no comments on this proposal I will take that as tacit approval. If anyone has any concerns with it I would appreciate comments to that effect, or even just "Wait, I have to think about it first". --Xover (talk) 09:17, 29 March 2021 (UTC)

And since nobody objected or yelled "Wait!", I've now tweaked the settings accordingly. I'll leave this thread open for a bit for stragglers, and after that anyone that wants to tweak the settings further can just open a new thread. --Xover (talk) 19:27, 9 April 2021 (UTC)

@Xover: Meh. This proposal gets lost in the morass of other components at the top of the page. Personally think that we need to rethink the scope of how the page works. As it is not working efficiently. I think that if we are going to have so many active proposals that they need to be subpages of this page and transcluded in, or linked out with some better means of notifications through Mediawiki:Watchlist-announcements (as we used to do. Numbers of the proposals that have come forward are blue sky thinking and more like WS:RfCs for long discussions, rather than simple proposals. — billinghurst sDrewth 00:40, 9 May 2021 (UTC)

I would happily see users who turned up yesterday (apparently) censored from kludging up the page, and this small community's time, with revived proposals. The page would also be a good deal shorter if nearly every thread was spared the often confounding and adversarial commentary of our resident bigot. With that aired, I think the scriptorium provides a lot of information to silent readers wishing to improve their contributions, buried within the commentary are solutions that are difficult to find in the help pages. Three days to a week is a very short time frame for this site, and often discussions need more airing and thoughtful input. I think the proposal addresses some of these concerns, but hope that urgency does not override the development of guidelines through broad and considered opinions. CYGNIS INSIGNIS 11:07, 9 May 2021 (UTC)

@Cygnis insignis: My proposal in this thread (which has been implemented) was only in regards threads that are manually closed (by adding {{section resolved}}), which we almost never do. All other threads are archived 30 days after the last comment. That's why I considered it appropriate to make the change so quickly and without a real !vote: it makes very little practical difference.

Billinghurst's subsequent comment (which echoes my own thoughts on some of the problems with this page) is directed at the broader issue of this page getting unwieldy, and will need both a broader set of changes to address, and broader discussion before any changes can be decided on. Xover (talk) 12:34, 9 May 2021 (UTC)

@Xover: Excuse the moaning then, I haven't seen any archiving here that was objectionable and what you've implemented sounds reasonable. CYGNIS INSIGNIS 12:41, 9 May 2021 (UTC)

This section was archived on a request by: — billinghurst sDrewth 08:24, 28 May 2021 (UTC)