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

Filter preventing good-will contributions?

The following discussion is closed:

No actionable points at Wikisource; complaints about the effects of global filters not felt at Wikisource may go to meta:Babel and about any other project's filters to the project in question

Absolutely any filter which prevents the good-will contribution of any potential contributor is counter the wikipedia ethos. It will be highly indicative if this comment is censored. 49.189.202.201 23:48, 17 December 2021 (UTC)

given how opaque filters are, you need to provide a screenshot or filter number to understand what you are talking about. we have global filters of proxy ip's and tor nodes, that are not under the control of admins here. i agree this is a major pain point where functionaries are betraying "anyone can edit" --Slowking4Farmbrough's revenge 00:45, 19 December 2021 (UTC)
Given the context of the original location of the comment, I assume "filter" here is meant in a barriers not blocking sense. (i.e. hurdles or difficulties that make it hard for people to contribute, even if relatively small, "filter out" people who are interested in making good-will contributions). MarkLSteadman (talk) 02:55, 19 December 2021 (UTC)
My only comment is that given the scope and long-term vision many people desire, expanding the size of the contributing community here would really help. I am not sure what analysis has been done either by looking at historical user behavior or via usability testing and to what extent changes there might help get more visitors to become contributors. MarkLSteadman (talk) 03:03, 19 December 2021 (UTC)
filters are the backend used by functionaries to fight abuse, and prevent a save. normally seen on some proxy ip's, or exceeding rate limit with quick statements in backround. normally accompanied by a big red screen, for example File:Abuse filter warnings, removal of information template, OTRS by non-OTRS member and unconstructive edit.png. --Slowking4Farmbrough's revenge 00:33, 22 December 2021 (UTC)
If you're hitting any filters at Wikidata or Commons, something is working as intended because you are not supposed to be editing there, since you are indef blocked. Inductiveloadtalk/contribs 07:17, 22 December 2021 (UTC)
funny, i have never hit a filter at commons or wikidata, so i guess they are not working. --Slowking4Farmbrough's revenge 00:18, 23 December 2021 (UTC)

  Comment Seems a spurious complaint, and approaches trolling with its kicker. This IP address has not had abuse filters prevent them from editing nor spam abuselog, nor any deleted edits. Reviewing the DISALLOW edit filter log shows nothing that we would say is an unreasonable action in recent times. We deal with reasonable complaints, especially where our system inhibits good-will editing. That said there needs to be systems in place to prevent spam and bad-will editing, and perfect differentiation is non-existent. — billinghurst sDrewth 08:40, 19 December 2021 (UTC)

i agree it is too vague to be actionable. but veteran editors continue to get caught up, and are astonished. (be it filtering Amtrak, or African country phone networks) we need an improved filter standard of practice to simplify explaination and feedback. the unaccountability is not a good look, and you should expect periodic venting until filter use is improved. --Slowking4Farmbrough's revenge 02:52, 20 December 2021 (UTC)
We are not having veteran editors continue to get caught in abusefilters, and when we have a false positive, they are typically updated. — billinghurst sDrewth 12:40, 20 December 2021 (UTC)
i'm sure it is just a misunderstanding: When you try to get something done on @wikidata --Slowking4Farmbrough's revenge 00:39, 22 December 2021 (UTC)
There are no rate limited filters at Wikisource. What's your point? Inductiveloadtalk/contribs 06:42, 22 December 2021 (UTC)
I won't speak for WD abuse filters, well not here, so please limit local commentary to what we can address locally.

And MM is a big boy and knows wikis well enough to know the best way to resolve abusefilter issues. I would argue that whining about it on twitter while maybe comforting is not the best way to get an AF issue addressed. — billinghurst sDrewth 12:09, 22 December 2021 (UTC)

well, apparantly when you are running quick statements in the backround, and try to edit, you get this filter. or if you are editing at an editathon with lots of new editors (which i saw). MM dunking on wikidata is all very amusing, and clearly communication channels are strained. it would be very difficult to trip a rate limit here, but it is theoretically possible. thankfully, we do not have filter problems here, but the principle remains, that veterans trip them up from time to time, and work stops - the permission work arounds are so obtuse as to be non-existent. i would not say spurious, but rather "did not stick around long enough to find out" --Slowking4Farmbrough's revenge 00:16, 23 December 2021 (UTC)
This section was archived on a request by: Inductiveloadtalk/contribs 12:29, 23 December 2021 (UTC)

November Monthly Challenge

 

Hear ye! Hear ye! This November, the Monthly Challenge has smashed though all previous records! 6079 pages proofread, validated, or marked "no text", completely eclipsing the target of 2000 by a factor of three, and claiming a new monthly record by a margin of over two thousand pages! This represents over 25% of all the ~22000 pages processed in all of English Wikisource in November. More excitingly, the number of pages validated, 2275, accounts for over 50% of all validation undertaken this month, meaning that the Monthly Challenge is roughly twice as effective at validations-per-page than the site average. We also had over 100 pages processed every day of the month, except the 29th, which fell short by only 2 pages. We also broke the record for most pages proofread and processed in a day (286 and 416 respectively, on the 1st) and pages validated in a day (205, on the 2nd). In fact, we saw so much work being done that a server performance hot-spot was discovered (don't worry, it was fixed).

This amazing result appears to support the idea that a broad range of works in the challenge attracts more interest throughout the month. Thank you to everyone who has contributed, and special thanks to Languageseeker for again marshalling the challenge as works were completed.

Works proofread in November include:

Validated works:

Several works, previously completed but never transcluded, were also transcluded and rounded out:

Only one work expired without being completed:

Coming up in December are some very interesting new works, as well as continuations of series from last month:

 
An early precursor to the familiar maps fantasty worlds

...as well as many more. There's something for everyone, so please come on in and help us smash more records! You can see the works included here. Nothing can chase away the wintry blues and ghosts of Christmas-almost-there like a nice, mince-pie-fuelled proofreading session!

Finally, as we are now a month away from Public Domain Day 2022 (a.k.a. New Year's day to those strange "normal" people), we are starting to collect works that will fall into the public domain for a special-edition challenge in January. Please submit ideas for 1926 works (and any other works entering the PD) on the nominations page.

Ho ho ho and merry proofreading, one and all! Inductiveloadtalk/contribs 14:54, 1 December 2021 (UTC)

How to name images without captions?

About half of the 120+ images in Index:The_Count_of_Monte-Cristo_(1887_Volume_1).djvu are without captions and are waiting to be uploaded. In my previous commons uploads, the text surrounding a captionless image provided context for the title and the caption.

These, with few exceptions, depict possibilities of situations from the illustrator's imagination. They are so irrelevant that they should be considered page decorations. I want to name and categorize these as such. e.g: COMC V1 D125 Page decoration.jpg.— Ineuw (talk) 05:41, 7 December 2021 (UTC)

@Ineuw: I would name these File:The Count of Monte-Cristo (1887 Volume 1), p. 49 illustration.jpg. That is, 1) use the same base name as the DjVu with which they are associated, 2) human readable without parsing "COMC", and 3) descriptive of what they are. On that last point, they are illustrations not "decoration" (decorations would be flourishes and fleurons and so forth) and spelling out distinguishes it from a possible image representing the whole scanned page (which may well be added on a collaborative / multi-project media repository). This naming schema can be extended to unnumbered plates as well:File:The Count of Monte-Cristo (1887 Volume 1), facing p. 49.jpg. Alternately, you can use each illustration's title from the list of illustrations. Xover (talk) 06:41, 7 December 2021 (UTC)
@Xover: I cannot accommodate your suggestion for several reasons. The first is that I don't want to create 100+ character titles. Especially, in an acronym driven universe such as Wikimedia and the web. Because this wasn't my upload and only focused on the images, I missed the list of illustrations. Much thanks for bringing it to my attention and will add them.
My use of book titles as acronyms goes back to my first uploads. At that time, I was warned on the commons that sufficient info is required in the title and the description, so I combined the two.
Every work I upload is properly categorized on the commons. This is the main category and this is the subcategory. I always use DjVu numbers because page numbers are not accurate. Every so often, they are duplicated or omitted. Also, Publishers/printers use divergent page numbering rules when it comes to blank pages and images.— Ineuw (talk) 22:43, 7 December 2021 (UTC)
crop tool automatically renames copies of image pages, https://commons.wikimedia.org/wiki/Commons:CropTool commons tends not to engage in rename drama of the semi-automatic. --Slowking4Farmbrough's revenge 23:31, 7 December 2021 (UTC)
User:Ineuw I name mine the same number as the ia file has (or hathi, with the addition of some zeros), something like Count of Monte-Cristo 1887 Vol. 1-0049.jpg if there are multiple images on one page, then Count of Monte-Cristo 1887 Vol. 1-0049-3.jpg and Count of Monte-Cristo 1887 Vol. 1-0049-4.jpg. It is only awkward in that the scan number often does not match the djvu page, but, it keeps them sequential in the Category at commons, and it is easy to determine if one is missing and where it is at and, it is pretty easy to script a mass name change, if that is something that is needed. Also, at commons, consider using Category:The_Count_of_Monte-Cristo_1887_(books)/Volume 1, etc.--RaboKarbakian (talk) 23:50, 7 December 2021 (UTC)
As an aside and a plug: https://ws-image-uploader.toolforge.org/ will attempt to construct some kind of useful title based on the work title, the page number if any and the image type (e.g. illustration, initial, etc), as well as a token effort at categorisation based on the image type. The user can then tweak the auto-generated name and cats.
It doesn't use the caption for anything other than dumping into the info template, but One Fine Day, it will add the right SDC fields at least. Inductiveloadtalk/contribs 23:59, 7 December 2021 (UTC)

21:59, 6 December 2021 (UTC)

WS changes

As far as I know, there are no major Wikisource-relevant changes due this week on top of these those listed last week #Relevant_changes_to_Wikisource_due_in_wmf.11 that never got deployed.

I know of no further comments that have been received about the OSD viewer and thoughts on improving the UI beyond its current "initial" implementation. The patch demo here is still a place you may try the proposed interface options. As always, Beta Wikisource represents the current "master" branch of the software at a given moment. Everything you see there is normally expected to be deployed within a week or so (changed merged on a Tuesday or Wednesday will normally be delayed to the next week since they miss the cut-off early on Tues mornings UTC). Inductiveloadtalk/contribs 22:47, 6 December 2021 (UTC)

Since yesterday (8 December, in my AM), while editing in Index:Some file.djvu/page#, the djvu page shows itself briefly and then disappears. After it disappears, the enclosing area remains and is not clickable. The new and the old zoom do not work and the area is pretty much void of any response to any interaction. Toggling the edit tools "on" brings everything back. I have not tried Beta yet.--RaboKarbakian (talk) 13:41, 9 December 2021 (UTC)

Marking what's wrong with files

I've been using a PETSCAN query - https://petscan.wmflabs.org/?psid=20861011 to disperse the category of "File to Fix" into more specfic subcats, and re-evaluating files placed in that category.

Much appreciated if other experienced contributors could aid this effort, especially as @Inductiveload: wrote a tool for generating pagelists recently. ShakespeareFan00 (talk) 16:17, 11 December 2021 (UTC)

Barbados becoming a republic

How about collecting Constitution (Amendment) (No. 2) Act (Barbados) here? – Kaihsu (talk) 20:53, 7 December 2021 (UTC)

it seems they are asserting all rights reserved http://www.governmentprintery.gov.bb/terms_conditions --Slowking4Farmbrough's revenge 15:27, 12 December 2021 (UTC)

New features in Page namespace editing?

Is there a way to turn off the coloring of wiki codes?— Ineuw (talk) 12:56, 11 December 2021 (UTC)

@Ineuw: -- I don't get any such coloring. You may have activated the relevant gadget. See under editing tools of gadgets menu of your preferences. Hrishikes (talk) 13:47, 11 December 2021 (UTC)
@Hrishikes: I looked before posting. There is no mention of highlighting the code on the edit, gadget of beta. Besides I use very few features, most are the default or disabled.— Ineuw (talk) 14:01, 11 December 2021 (UTC)
@Ineuw: You have most likely clicked the   button in the toolbar, which turns on syntax highlighting for wikicode using the CodeMirror extension. Just click it again to toggle it off (it doesn't work all that well). Xover (talk) 14:31, 11 December 2021 (UTC)
@Xover: Many thanks, another supernatural event resolved. Now, if I can only get back the page image scrolling, but then, that is up to the gods on Mount Phabricator. — Ineuw (talk) 01:28, 12 December 2021 (UTC)
yeah, i agree the image zoom, rather than scrolling up and down, is a major UX unimprovement. clearly no testing done there. --Slowking4Farmbrough's revenge 22:29, 12 December 2021 (UTC)

22:27, 13 December 2021 (UTC)

Public Domain Day 2022

here are some works to plan on:

  • 13th edition of Encyclopædia Britannica
  • Agatha Christie, The Murder of Roger Ackroyd
  • Arthur Conan Doyle, The Land of Mist
  • Ernest Hemingway, The Sun Also Rises
  • Franz Kafka, The Castle
  • T. E. Lawrence, Seven Pillars of Wisdom
  • A. A. Milne, Winnie-the-Pooh
  • Dorothy L. Sayers, Clouds of Witness
  • Ruth Plumly Thompson, The Hungry Tiger of Oz [4]
  • Gorton James et al., Profit Sharing and Stock Ownership for Employees,
  • Willa Cather, My Mortal Enemy
  • Christopher Vecsey, Adventurous Religion and Other Essays
  • Seán O'Casey, The Plough and the Stars
  • Katharine Brush, Glitter
  • Isabel Cotton Smith, Blue Book of Cookery and Manual of House Management
  • Felix Salten, Bambi
  • Florence Armstrong Grondal, The Music of the Spheres: A Nature Lover's Astronomy --Slowking4Farmbrough's revenge 03:07, 1 December 2021 (UTC)
There is already a page for coordination at Wikisource:Requested texts/1926 which lists some of these. --EncycloPetey (talk) 20:47, 1 December 2021 (UTC)
So much to do, so little time. Can these lists be merged into one? — Ineuw (talk) 01:42, 12 December 2021 (UTC)
You can certainly add to that but these lists are pretty arbitrary because not all works are going to end up Wikisource any time soon. This should really be a list of what we anticipate someone will actually work on adding. —Justin (koavf)TCM 23:28, 15 December 2021 (UTC)

Help: Half of edit screen is no longer accessible

I use over/under page editing and that is broken. How can I fix it?— Ineuw (talk) 09:41, 14 December 2021 (UTC)

i'm afraid i have gone to side by side with timeless skin for small screen, as the image zoom, rather than scroll up / down, has deprecated over / under for me. (i suspect you have a zoom issue) since UX is not done, we are left to do trial and error, for the combination of skins and toolbars that work best. --Slowking4Farmbrough's revenge 15:18, 14 December 2021 (UTC)
I find it interesting that everyone involved with this update is silent. Is this going to be the permanent layout, My current options are for over/under editing in Vector:
  • CSS settings to improve my view no longer work.
  • Half screen editing with an empty half right screen.
  • No scrolling of the image.
  • Overhead preview is the width of the screen, as opposed to the page width, which was controlled by CSS.
  • Previously existing editing controls in Preferences are gone.
These changes are not acceptable to me.Ineuw (talk) 01:39, 16 December 2021 (UTC)
@Ineuw: Are you using the Timeless skin? I'm not familiar with it but you can maybe reach out to who maintains it for help. —Justin (koavf)TCM 02:30, 16 December 2021 (UTC)
@Koavf: I always used Vector. Then I tried Timeless today, which made no difference. So, I returned to Vector.

Title Page Not Exported

In The Works of H. G. Wells (Atlantic Edition)/Anticipations, the Title Page is not exported because it is not part of the TOC. What is the best way to allow for the export of the Title Page? Languageseeker (talk) 02:43, 15 December 2021 (UTC)

There are three main main ways I can think of:
1. Just transclude it on that page like a "normal" book, above the TOC
2. Use an AuxTOC (if the main TOC sets ws-summary, e.g. if it uses {{TOC begin}}, you also need to wrap everything in an {{export TOC}}, or only the ws-summary element will be used.

Example: Separate Aux TOC

{{export TOC|
{{AuxTOC|title=|
* [[/Title page/]]
}}
...The usual TOC
}}
 (not listed in original)

...The usual TOC

3. Use the class "wst-toc-aux" with the {{TOC begin}} templates. This does not need {{export TOC}} as it's all in one table. Clearly not popular as when I tried it I got reverted because "it wasn't in the original". For things like Prefaces and so on, it seems much neater (to me) than tacking on AuxTOCs.

Example: {{TOC begin}} and wst-toc-aux

{{TOC begin|width=35em}}
|+ CONTENTS
{{TOC row 2-1|class=wst-toc-aux|Title {{Smaller|(not in original TOC)}}}}
{{TOC row 2-1|Chapter 1 | 1}}
{{TOC row 2-1|...}}
{{TOC end}}
}}
CONTENTS
Title (not in original TOC)
Chapter 1 1
...
Inductiveloadtalk/contribs 06:56, 15 December 2021 (UTC)
@Inductiveload Thanks for the help. I'm wondering if it's possible to add a hidden parameter so that it will export the Title Page, but will not actually show up. Languageseeker (talk) 01:59, 16 December 2021 (UTC)

Last Chapter of Southern Historical Society Papers, Volume 3

First of all, sorry for my grammar. So, basically at Index:Southern Historical Society Papers volume 03.djvu there is the last chapter. The pages for the last chapter itself at the Index is declared as problematic and without text, already transcluded on the other scan. My question is if I want to transclude it, what should I do? Thanks. Mnafisalmukhdi1 (talk) 12:02, 17 December 2021 (UTC)

@Mnafisalmukhdi1 The roster was already transcluded at A Roster of General Officers. I updated the TOC to reflect this. Languageseeker (talk) 12:42, 17 December 2021 (UTC)
Then I don't have to transclude it again for the last chapter? Thanks. Mnafisalmukhdi1 (talk) 12:50, 17 December 2021 (UTC)

Missing page placeholders

"The following pages in this file are missing: 5,6.

"Placeholders should be inserted so that when the pages are inserted, existing content will not become misaligned."

How is that done, please? The index in question is Index:Aerial travel for Business or Pleasure - Thos Cook & Son - 1919.pdf. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 23:14, 17 December 2021 (UTC)

@Pigsonthewing the short answer is using some kind of PDF editor (I think PdfSam is supposed to be good, I use a command line tool called stapler) and inserting some kind of placeholder page, for example File:Generic placeholder page.pdf at the appropriate point. Inductiveloadtalk/contribs 23:28, 17 December 2021 (UTC)

In Windows you can also use the djvm command line tools from DjVu for replacing a page using the -d (delete) and -i (insert) switches.

@echo off
: location of the app
: C:\Program Files (x86)\DjVuLibre\djvm.exe

: if you wish to add the location to the path
: path=%path%;C:\Program Files (x86)\DjVuLibre

echo on
:"C:\Program Files (x86)\DjVuLibre\djvm" -d "I:\0_djvu\A_Chapter_on_Slavery.djvu" 166
: check result
"C:\Program Files (x86)\DjVuLibre\djvm.exe" -i "I:\0_djvu\A_Chapter_on_Slavery.djvu" "Replace_it_with_page152.djvu" 166

:see any error messages
pause

:exit

Ineuw (talk) 07:15, 18 December 2021 (UTC)

22:05, 20 December 2021 (UTC)

Help with columns spanning page breaks

For the text Index:Axiochus (Spenser, 1592).pdf, I have implemented two columns for the purpose of presenting both the original and a modernized version (without archaic spellings) of the source text. In the Transcluded text however, Axiochus (Spenser)/Axiochus, page breaks are causing new lines between pages. How would I go about fixing this, such that transitions are seamless between pages. On Help:Page breaks I saw templates relating to lists and tables, so might it be advised to use the "Column list" template as opposed to "columns" to fix this? Oryang7 (talk) 05:09, 17 December 2021 (UTC)

I hate to say this, but I personally don't think you should try to do any such thing. Wikisource is about presenting (relatively) faithful digitized versions of the source document, even when that source uses archaic spellings or may be otherwise difficult to read/comprehend by modern readers. I can see you've put a fair bit of effort into this, so I don't say this lightly, but I think you should just be presenting the text as written, rather than trying to modernize it. — Dcsohl (talk)
(contribs)
18:43, 17 December 2021 (UTC)
Do you know if it's possible to someway preserve the modernized version on wikisource? If I removed the modernized version from the transcluded text might it be possible to create another page, unsourced, of the modernized version? Or, if it needs to be sourced, may uploading a pdf of the modernized version, which I assume could be done as the original is out of copyright, and transcluding it to a seperate page suffice? Or, and I don't know if this is already present on wikisource, is there a way to implement a feature allowing readers to toggle between the original and modernized versions on the single transcluded page? Oryang7 (talk) 23:49, 17 December 2021 (UTC)
I'm not entirely sure whether to point you to Wikisource:Annotations or Wikisource:Translations, but I would think reviewing those two pages might be helpful to you. — Dcsohl (talk)
(contribs)
19:51, 22 December 2021 (UTC)

Template:Lang

The documentation to the template {{Lang}} says that the code for the language being displayed should be provided using ISO 639. However, there many languages have different codes in ISO 639-1, ISO 639-2 and ISO 639-3. Which of these should be used? --Jan Kameníček (talk) 18:57, 22 December 2021 (UTC)

@Jan.Kamenicek as far as I know, they HTML lang attribute can use either two or three-letter language codes (so ISO 639-1 or -2), as well as a two-letter country code suffix. According to https://www.w3.org/2005/05/font-size-test/starhtml-test.html, this is RFC 3066, and you can also have other IANA "subtags" for "variants", the example given is "en-scouse", but there are lots.
Generally, just the two letter one will do in most cases unless you really need to drill down into the language variant. Inductiveloadtalk/contribs 19:11, 22 December 2021 (UTC)
According to this document from W3C, the standard is IETF BCP 47, which is something of a catchall that "combine[s] subtags from other standards such as ISO 639, ISO 15924, ISO 3166-1 and UN M.49." — Dcsohl (talk)
(contribs)
19:12, 23 December 2021 (UTC)

Naming convention

I have a question that is not yet governed by Wikisource:Naming conventions & Help:Disambiguation.

A work has two different editions published in the same year. How should we name these editions, such as

  1. "aaa (2021a)" and "aaa (2021b)",
  2. "aaa (1st edition)" and "aaa (2nd edition)",
or otherwise?

Thank you.

--Miwako Sato (talk) 11:41, 23 December 2021 (UTC)

@Miwako Sato: What actually distinguishes the editions? The usual go-to for disambiguation are publisher and place of publication. If neither is sufficient we have to get more creative. "1st edition" and "2nd edition" may be reasonable options for some cases. For laws and similar it may make sense to use both month and year, or even a full date. It's a bit of an edge case, so we don't really have firm guidance or practice on it, and the details of your particular case will be necessary to suggest anything sensible. Xover (talk) 11:50, 23 December 2021 (UTC)
The editions in question were published in the same year, by the same publisher, and at the same place. The only thing that distinguishes them is their contents (the subsequent edition contains revised & additional contents). So, in this case, I think "1st edition" and "2nd edition" would be appropriate (because it's a book). Thank you so much, and special kudos for a very quick response! --Miwako Sato (talk) 12:04, 23 December 2021 (UTC)

Page image vanishes after loading

Whenever I try to edit in Page: namespace, the page image loads, and then immediately vanishes. I am proofreading against an empty box. I have blanked my .js and .css files, tried another djvu file, tried changing skins, tried changing browsers. Am I the only person experiencing this? This is a total blocker for me -- I might as well quit Wikisource if I can't get this fixed. Hesperian 23:24, 16 December 2021 (UTC)

After noticing RaboKarbakian's comment above, I discovered that the problem goes away if I "Enable the editing toolbar" in my preferences. I'm not exactly happy at being forced to turn on a toolbar that I don't want, but at least there is a workaround. Hesperian 23:31, 16 December 2021 (UTC)
Yeah, I tried it without the toolbar today and nothing has changed. Even with the toolbar, today I had to reload every other page to get the djimage to load. To the best of my knowledge, the only complaints have been about the changes in zoom and not being able to change the orientation and I wonder if people actually use the tools or just have them enabled....--RaboKarbakian (talk) 23:38, 16 December 2021 (UTC)
yeah, i have the new OCR button enabled but do not use it. no improvement to require the learning curve. this is visual editor redux, where the devs make their "improvements" without UX. it was hard enough getting the devs to put their button on the toolbar, rather than hovering over the image in the way. --Slowking4Farmbrough's revenge 01:14, 17 December 2021 (UTC)
This was fixed within hours over a week ago, but wasn't merged until today: https://gerrit.wikimedia.org/r/c/mediawiki/extensions/ProofreadPage/+/745460. Since today is Friday, there are now no backport windows until Monday. I'm personally not available today at the normal window times of 1200, 1900 and 0000 UTC anyway; if anyone wants to request an emergency backport, they'll have to do it themselves or find someone who can. I have, however, done a backport patch for that person to use: https://gerrit.wikimedia.org/r/c/mediawiki/extensions/ProofreadPage/+/748095.
The whole torture of actually getting OSD deployed properly (there have been a lot of compounding issues, including the main developer going AWOL, the usual lack of review capacity/interes, a near total lack of engagement on outstanding UX questions and, unrelated to WS, missed MW releases for weeks at a time) is why Wikisource needs a proper Code Steward to take point and make sure important fixes don't just rot on Gerrit and, when fixed, get backported promptly. Inductiveloadtalk/contribs 11:28, 17 December 2021 (UTC)
huzzah, and beware, you may have self-nominated. i would support a individual grant, https://meta.wikimedia.org/wiki/Grants:Programs/Wikimedia_Community_Fund or set up the go fund me / patreon. --Slowking4Farmbrough's revenge 03:49, 18 December 2021 (UTC)

So, that magical Monday has come and gone, and the problem still persists.--RaboKarbakian (talk) 15:39, 24 December 2021 (UTC)

There are actually no Release Engineering backports (or deploys) this week or next because of the holidays. I didn't realise there wasn't even a window on Monday when I wrote that. This will not, therefore, be resolved until the 5th Jan at the earliest. Inductiveloadtalk/contribs 16:01, 24 December 2021 (UTC)
Please excuse me for whining on a holiday. Thank you for the information, and eh, Seasons Greetings!--RaboKarbakian (talk) 17:02, 24 December 2021 (UTC)

Wikipedia @ 20 (published by MIT Press)

I am starting to translate individual essays from Wikipedia @ 20 (published by MIT Press) to Croatian and I wonder if it would be good to import the whole book (licence), as well as to advance with translations here on Wikisource? --Zblace (talk) 15:27, 26 December 2021 (UTC)

Upcoming Call for Feedback about the Board of Trustees elections

You can find this message translated into additional languages on Meta-wiki.

The Board of Trustees is preparing a call for feedback about the upcoming Board Elections, from January 7 - February 10, 2022.

While details will be finalized the week before the call, we have confirmed at least two questions that will be asked during this call for feedback:

  • What is the best way to ensure fair representation of emerging communities among the Board?
  • What involvement should candidates have during the election?

While additional questions may be added, the Movement Strategy and Governance team wants to provide time for community members and affiliates to consider and prepare ideas on the confirmed questions before the call opens. We apologize for not having a complete list of questions at this time. The list of questions should only grow by one or two questions. The intention is to not overwhelm the community with requests, but provide notice and welcome feedback on these important questions.

Do you want to help organize local conversation during this Call?

Contact the Movement Strategy and Governance team on Meta, on Telegram, or via email at msg wikimedia.org.

Reach out if you have any questions or concerns. The Movement Strategy and Governance team will be minimally staffed until January 3. Please excuse any delayed response during this time. We also recognize some community members and affiliates are offline during the December holidays. We apologize if our message has reached you while you are on holiday.

Best,

Movement Strategy and Governance --Civvi (WMF) (talk) 10:57, 27 December 2021 (UTC)

Proposal to delete The Count of Monte Cristo 1844 publication

This The Count of Monte Cristo is a problematic version. It mentions no source except Wikipedia, and it has no affiliated source file. The title is also problematic. French language publications hyphenated Monte-Cristo. Only two English publications did so, while every other Latin alphabet publication did not. I propose to delete it.— Ineuw (talk) 06:11, 7 December 2021 (UTC)

@Ineuw: Use WS:Proposed deletions to, uhm, propose something for deletion. :) Xover (talk) 06:42, 7 December 2021 (UTC)
@Xover: It is posted there as well. The problem is that I had to usurp the disambiguation page for the 1887 version. Ineuw (talk) 17:59, 5 January 2022 (UTC)
This section was archived on a request by: Ineuw (talk) 17:59, 5 January 2022 (UTC)

Template:Highlights - change to rotating highlights

Hi all, new here but wanted to pick up what has been said on the Template:Highlights talk page and further back in the main discussion archives.

Currently, the highlights section on the main page is a static depiction of what this project has to offer with some US-centric elements. It would be better to change to a rolling display of validated articles that are of interest to the wider community. Whether this is done randomly throughout the high-quality sources or leans on nominated/ voted for sources, I am sure there is a way to modify this to better reflect both the diversity of content across a global setting. Jamzze (talk) 06:03, 22 December 2021 (UTC)

  Support doing something different. I'm not sure if rotating or just a bigger/wider list of portals is better. We have a lot more chunkier portals now than we did when that list was created (but still lacking).
I do not support a voting system, that's just going to rot and go stale. If we wanted a way to highlight excellent portals, we can do that better (e.g. include in Featured Texts, have a Featured Portal, or whatever). However, we don't really have enough Portals that good. Inductiveloadtalk/contribs 08:36, 22 December 2021 (UTC)
Yeah I think a new list of highlighted areas would be good! Who could action this, though? Jamzze (talk) 21:48, 27 December 2021 (UTC)
Two out of six of the lines are US-centric. For an English-language project, that's entirely reasonable and not something that needs to be "remedied". {{highlights}} is Used on the Main Page to display highlighted collections and areas on Wikisource. and is paired with the "Explore Wikisource" box just below it. Together they provide visitors with a broad overview and multiple ways to start navigating Wikisource.
That they should point to high-quality content is a requirement, but is not their main purpose: these are not "featured" areas of Wikisource (nor do or should all the links point to portals). We could have featured portals etc. too, but that's a separate issue: our featured content is up in the top left monthly featured text box. Any further "featured something"should take that as a starting point.
In other words, {{highlights}} can almost certainly be improved, but that would be "further improve" and not "fix a big problem with". And Wikisource not being Wikipedia, we don't really need to worry about a "global perspective" except to the extent that English is a world language for which we have content from literally every continent (including Asia). If we can find better places to direct visitors than the "US history" and "US laws" collections then by all means, but I'd need to see some concrete proposals for that. Xover (talk) 09:34, 22 December 2021 (UTC)
  •   Comment I would agree that a static highlight section is not helpful for us with all our content. Whether a dynamic list is the best, or a curated list based on our input, I am not sure. We are not brilliant at curating portals for display. — billinghurst sDrewth 12:13, 22 December 2021 (UTC)
  •   Comment Just adding to this as I haven't seen any changes come about because of the above discussions. Just wondering who would be able to mix-up the highlights section to include other things? Jamzze (talk) 13:30, 7 January 2022 (UTC)
  •   Comment The last time that I made changes to the Highlights, I had the sense that the listings were selected because they are larger and/or general listings by topics and genre. Yes, the History listings are US-centric, and if someone could identify a better set of Portals in good shape, I'd certainly favor changing that line. But I'm not sure that having a rotating list would help anyone, as that means that the section is randomized for readers, which I don't think would be popular with visitors who return only to discover they cannot find the listing they followed previously. The "Highlights" are essentially the topical listings, as opposed to the Author / Era listings which appear below that section. IMO, the best way to improve this section is to work to improve key Portals on Wikisource. This can take a lot of work, but it can be done. --EncycloPetey (talk) 01:14, 11 January 2022 (UTC)

InternetArchiveBot

I would like to propose bringing InternetArchiveBot to Wikisource to fix dead links on the wiki. Although individual pages in the page namespace on Wikisource tend to not have external links, a recent database query found that there are over 1.3 million external links on English Wikisource. Given that any website can go offline at any time, that is a potentially large number of broken links on Wikisource. InternetArchiveBot is very flexible and customizable, designed to run on small and large wikis alike. The InternetArchiveBot development team is happy to work with the Wikisource community to make sure the bot recognizes all the local reference, web archive, and dead link templates. We are happy to answer any questions you have. Harej (talk) 23:47, 20 December 2021 (UTC)

What? No, absolutely not. For one thing, and despite your fondness for throwing around large numbers, enWS generally does not use external links except those to… the Internet Archive and other archives that can be presumed to be stable and persistent. That other external links occasionally occur and occasionally break barely rates as an annoyance. But more importantly, no external entity with goals and priorities different from the Wikimedia movement—but who seems happy to spend significant chunks of change paying people to edit the projects, act as their advocates, and developing and running bots—should be allowed to run a bot designed to add the largest possible number of links from Wikimedia projects to their own service. And it certainly doesn't help that whenever this paid team has met with community pushback and criticism they've refused to engage with, much less address the concerns of, the community; opting instead to rely on proxy champions extolling the virtues of the bot and its big huge number of edits ("Look look, lots of edits! Millions of links!").
I've told you (the collective you) before: if you want to contribute to the Wikimedia movement projects rather than just inflate the number of high-value incoming links to the WayBack Machine, put some effort into your bibliographic metadata story. We very much rely on the IA book scans, and the state of their bibliographic data is just atrocious. Competing for the same volunteers through OpenLibrary instead of joining forces is just dumb. Building a pretty (user friendly) frontend to crowd-source and manage that data on Wikidata (or something else Wikibase-based) and integrating with WikiCite for the Wikipedias and integrating tightly with Proofread Page for the Wikisourcen would have great synergies for both communities; and it would put the volunteer effort towards increasing quality rather than mere number of links. The Wikisourcen spend a significant amount of volunteer effort researching, documenting, and managing bibliographic metadata connected to book scans sourced from the IA; and all that effort currently stays locked up in unstructured form, instead of directly and immediately contributing value to the global bibliographic web including IA. Xover (talk) 08:00, 21 December 2021 (UTC)
Yes please. Excellent idea. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 10:44, 21 December 2021 (UTC)
Weak support: the vast, vast majority of external links are to the IA itself, Hathi, Google or some other text repository. However, we do have a non-zero number of links, especially in, Author talk spaces which represent valuable contributor research and can indeed sometimes go dead.
But I'll agree with Xover: this is not the best way that the IA can assist Wikisource. Critically, archiving a handful (and on the scale of the WBM, it's a very small handful) of enWS links is also not the best way that Wikisource can assist the IA: the biggest moan about the IA, that almost literally everyone Wikisourcerer who has been to the IA has, is not that the metadata is very bad, that's just life in the library, but that it cannot be changed. If nothing else: Internet Archive ID (P724) provided perfect access to Wikidata metadata which is much better, when it exists! Inductiveloadtalk/contribs 10:57, 21 December 2021 (UTC)
meh effort better spent at wikidata, the notion of IA archiving its own website, seems a little circular, but there is also google books, and hathi trust. and yeah, a metadata cleanup drive with some mechanism for crowdsourcing IA metadata is more a priority for this community. but thanks for thinking of us. cheers, racer x.--Slowking4Farmbrough's revenge 23:36, 21 December 2021 (UTC)
and thanks for the linking Introducing Trusted Book Providers, maybe some round tripping of metadata would improve trust. --Slowking4Farmbrough's revenge 15:54, 22 December 2021 (UTC)

External links by namespace

I have pulled a report of external links on English Wikisource, organized by domain and by namespace. A link to the full dataset, with all the namespaces, is here. However, InternetArchiveBot only runs on content namespaces, so for my analysis I am focusing on the main, Page, Author, and Index namespaces.

For each namespace, I came up with a list of the top linked domains in each namespace. With this approach, counts for subdomains are reported separately from the primary domain. Because of this, I then proceeded to make aggregate counts across all relevant domains for the Internet Archive, Hathi Trust, Project Gutenberg, Google Books, Google Scholar, and Wikimedia sites (including wmflabs.org and toolforge.org). These collectively are the "focus sites." Here are my high level findings:

  • In the main namespace, there are over 654,000 links to 4,575 domains. Links to the focus sites add up to about 2.6% of those links. In fact, over 93% of links are to a single domain, openjurist.org, which has over 100 times the links as the second most linked domain, archive.org.
  • The page namespace only has around 6,400 total links, linking to fewer than 2,200 domains. Over 10% of the links are back to Wikimedia and over 6% to the Internet Archive. In total, 17.7% of links are to focus sites.
  • The author namespace has over 373,000 links to 1,314 domains. In total, 23.1% of links are to focus sites.
  • The index namespace has over 135,000 links to only 124 domains. It is worth noting that 98.5% of these links are back to Toolforge and Wikimedia Commons.

If we exclude links associated with the "focus sites," then in each namespace, there are 637,654 external links in the main namespace, 5,277 external links in the page namespace, 287,024 external links in the author namespace, and 1,662 external links in the index namespace. This analysis did not look into how many of these links are dead, since that would require visiting each URL individually, which would take a very long time for hundreds of thousands of URLs. However, there is no guarantee that any website on the Internet stays online, and any website can go offline at any time. Harej (talk) 23:37, 21 January 2022 (UTC)

FYI, your figures for the author namespace are dominated by the links provided by the {{authority control}} template, which provides the links at the bottom of author pages. Not that those links cannot be usefully archived, they surely can be and should be. However, a systematic drive to archive authority control resources should probably operate using Wikidata's data directly, as Wikisource does not always have a page for every possible author, so you'd miss millions of records by approaching that from the Wikisource side. Inductiveloadtalk/contribs 23:53, 21 January 2022 (UTC)