Floating images to the right mid-paragraph with captions
Latest comment: 10 months ago4 comments3 people in discussion
Here is a sentence. The image {{FreedImg| file = Quinby and Son (1925) frontispiece.png
| float = right
| width = 100px
| caption = A caption.
}} is in the middle of the paragraph.
Here's the next paragraph.
I've been trying to float an image (along with a caption) to the right in the Sandbox today (see recent edits), but the image needs to float mid-paragraph, while allowing the rest of the paragraph to sync together despite that. Anyone know how to do this? PseudoSkull (talk) 16:02, 24 December 2023 (UTC)
Latest comment: 10 months ago4 comments3 people in discussion
The Wikipedia article on the film The Singing Fool (1928) claims that it is based on a short story called "The Singing Fool" by Leslie Barrows, a pseudonym of Charles Graham Baker. The film itself also confirms the fact that the story was based on a work (but it doesn't specify short story) by Barrows. Can anyone find where this short story originated? Which periodical was it, and what year was it first published? PseudoSkull (talk) 00:50, 3 January 2024 (UTC)
This seems spurious to me: sources today call this a "short story" or sometimes a "play" and I can find several books from the 1970s that call it a "play", but from The National Board of Review Magazine, 1926, p. 8, they list it as "original screen story", which seems the exact opposite of being an adaptation. —Justin (koavf)❤T☮C☺M☯01:57, 3 January 2024 (UTC)
It sounds like it was just the screenplay written for the film. Those things never end up getting released, so listing it as a separate work here doesn't really make sense after all. Thanks for confirming. PseudoSkull (talk) 03:24, 3 January 2024 (UTC)
It works now. If you have very large overfloat images, then it sometimes doesn't work as well. If you want to tinker with some of the widths and spacing, you can maybe make it a little prettier, but again, this is serviceable. If you need help that, let me know. —Justin (koavf)❤T☮C☺M☯05:42, 6 January 2024 (UTC)
Replacing a djvu file cancelled post
Latest comment: 10 months ago6 comments3 people in discussion
What do you mean when you say that you "installed" that copy? Where? What did you do, and have you done anything other than that? --EncycloPetey (talk) 22:23, 1 January 2024 (UTC)
What about the text layer? Now that you have created all the 700 pages, you will not get the new one (which is why has been stated several times that is is not a good practice to create red pages in bulk). Or you created the pages because you wanted the old one? Mpaa (talk) 23:30, 1 January 2024 (UTC)
Please delete the Index:Jesuit education, its history and principles viewed in the light of modern educational problems.djvu and the related pages, including the book on the commons, if possible. Both books have scribbles and notations, but this copy is cleaner. I can reconstruct the text from either scans. I keep copies of both in text, jp2 and djvu. I have copies of the proofread text. Please don't be concerned about them.
Pages creation really helps to get my bearings. It is created in my own fashion with a Python macro on my desktop or laptop. Without it I am lost. — ineuw (talk) 22:24, 2 January 2024 (UTC)
There are hard technical limitations to what is allowed for page names, but you can pipe in any display form [[PAGENAME|DISPLAYNAME]]. The character | is sometimes called a "pipe". --EncycloPetey (talk) 18:10, 4 January 2024 (UTC)
I was reading H:REF, but it wasn't exactly clear what I should do. Do I copy in the references from the separate volumes with <ref> tags? Do I transcribe volumes five and six separately and then somehow put in appropriate links/transclusions in volumes one through four? (In which case, I suppose I should do the notes volumes first.)
Please advise.
(I imagine the index is a whole other can of worms I can ask about when I get there.)
Also, if I need to do the endnote volumes on their own first, what's the appropriate format to use? A table? Or is there a template for this? - Dave314159 (talk) 20:00, 11 January 2024 (UTC)
@Dave314159: Tackling that work as your first project here and on your own is utter insanity, and I strongly urge you to find some easier works to start with before you go to the magnum opus. Your odds of getting any significant way on that one without burning out are slim if one looks at the history of such efforts here.But if you cannot be dissuaded, you are probably looking for {{authority reference}}. I've never had to use it myself, so I can't vouch for how well it works in practice, but it does seem to be made for this use case. Xover (talk) 20:08, 11 January 2024 (UTC)
@Xover, My plan is to get it started (and get some of the hard parts figured out, if not out of the way completely), then recruit some folks to help slog through the page-by-page editing of the main text. Is that still utter insanity? I'm shooting for "attainably ambitious".
As your first work I'd say it's still insane. You need some experience before you start seeing how it's smart to approach certain issues, and before that you'll be prone to make things more difficult for yourself and your would-be helpers than is necessary, to get pushback from the community on the ways you're doing some things, create a future cleanup need, etc. I am also curious where you plan to recruit your helpers from. It is rare to find people interested in contributing to Wikisource that do not already have their hands full with their own pet texts. It'd be great if you pulled it off, but if your plan depends on others I'd suggest recruiting them first.It'd be awesome if you pull it off, and it has been done before, but large text + complex text + contributor's first project equals an unsustainable combination in my experience. I'd like people to stay around as contributors for years and years, not burn out and quit in frustration on their first project. Xover (talk) 08:59, 13 January 2024 (UTC)
OCR shifted by one page
Latest comment: 9 months ago10 comments3 people in discussion
A few volumes of the text I'm working on (volume seven, for example) have the OCR text shifted by one page. How should I deal with this? Is there some fix? Did I do something wrong when I used ia-upload? (And would re-doing it properly fix it?)
I just tried that on a few random pages, and it didn't seem to make any kind of difference. Good to know it exists, though. - Dave314159 (talk) 19:11, 11 January 2024 (UTC)
I cleared my cache, tried multiple browsers on two different devices, and It's still shifted by one every time. I even cleared the cache on my router and also checked it on my phone with mobile data.
I suppose the solution may be to just re-OCR the pages.
Citations in the Scanned Encyclopaedia Britannica Pages
Latest comment: 9 months ago3 comments3 people in discussion
Does the encyclopaedia britannica have any citations? I was looking at the scanned page for Spes, and neither the book or the scan has any citations. Does it have any or are they just somewhere else? I'm not an editor or anything, I was just trying to find the origin of Spes as a nature goddess, and every website seems to be copy & pasting the 1911 Encyclopaedia Britannica 644 SPERMACETI—SPEUSIPPUS page, and I was wondering if the encyclopaedia cites anything. Thank you! 162.217.74.2922:17, 13 January 2024 (UTC)
Correct. The encyclopedia article you're probably referring to, located at 1911 Encyclopædia Britannica/Spes, did not contain any citations in the form of footnotes. It does (loosely) cite a German book called Religion und Kultus der Römer (1902) for the statement at the end of it, though. The scan of the page, used as evidence of what's in the text, is located at Page:EB1911 - Volume 25.djvu/666.
Wikisource only aims to provide the material as it was written by the authors. If the original source didn't contain citations, we're obliged not to add any ourselves, to stay true to the original text of the work (even if the information in it is inaccurate, outdated, biased, or poorly substantiated).
Latest comment: 9 months ago4 comments3 people in discussion
Hi. I'v been werking on The Constitution of India, and I need help whith two desianing mesars. The one is crieting a ref tag that uses * instet of nombers (usefol in page 68 for exsempal) and the other is creating two laires of texst thet stay togaver in a formate similer to a mathe eqution (usefol in page 67 by 159). Thenks, איש עיטי (talk) 19:36, 14 January 2024 (UTC)
What is the process for moving an author page to the portal namespace?
Latest comment: 9 months ago4 comments3 people in discussion
I created Author:Edward Fitzgerald (1763-1798) to list works about Fitzgerald only to realize that he has not authored works available for inclusion in Wikisource. (I haven't found his archived letters published before 1929). The page Wikisource:Portal guidelines mentions moving pages from author to portal namespace but doesn't say how to do that.
People who have not authored any works themselves can be represented by a portal. Example: Portal:Joe Hooper. […] These pages may move between the Portal and Author namespaces as more information becomes available.
Are only certain users able to move pages between namespaces?
(In theory, his letters could be included on Wikisource, especially since they're all PD-old. So eventually this can become an author page.) PseudoSkull (talk) 20:23, 16 January 2024 (UTC)
Categorisation of authors
Latest comment: 9 months ago3 comments2 people in discussion
I am trying to wrap my head around how authors are categorised. In the Authors category, there is a subcategory called Authors by occupation. In this category, numerous occupation categories can be found following the naming scheme [occupation] as authors. This seems logical to me, although I am not sure why the categories are not simply named by the occupation, forgoing the as authors part.
There then exists a sister category called Authors by type. This seems to contain a mix of what may be considered a type (novelists, podcasters) but also what may be considered professions (governors, professors).
To complicate matters further, there is a third sister category called Occupations. This category seems to contain occupation subcategories but without the as authors naming scheme. The subcategories are empty of pages and contain a template declaring them to be meta categories. I have not found any good explanation as to why they exist and what they offer that the categorisation mentioned in the first paragraph does not. Stefán Örvar Sigmundsson (talk) 01:40, 16 January 2024 (UTC)
Your comment has helped me realise that the Occupations category is being used for works, as you mentioned, but also authors, as I mentioned. After all it is a child of the categories Works by subject and Authors. Then each subcategory within Occupations will have its subcategories, Biographies of [occupation] (for works) and [Occupation] as authors (for authors). It seems a bit circular but at least I am beginning to wrap my head around the logic of it. Stefán Örvar Sigmundsson (talk) 19:37, 17 January 2024 (UTC)
Help is asked with my an image overlay
Latest comment: 9 months ago1 comment1 person in discussion
Latest comment: 9 months ago2 comments1 person in discussion
Hello (again). I'm trying to transclude the poem The journey of life by Author:William Cullen Bryant, which is spread across pages 225 and 226 of the source. I'm using the code <pages index="Poetical works of William Cullen Bryant (IA poeticalworksof00brya).pdf" from=225 fromsection=The Journey of Life to=226 />, but it doesn't work — it just transcludes p. 226. I have added the necessary label to p. 225 and triple-checked the spelling on both ends to no avail. What am I doing wrong?
Thanks, Cremastra (talk) 22:40, 19 January 2024 (UTC)
Nevermind, I think I figured out the glitch: there was a {{centre}} template applied to the whole text, so that the section was inside the template... anyway, I believe I fixed it. Cremastra (talk) 23:01, 19 January 2024 (UTC)
Sheet music in film
Latest comment: 9 months ago4 comments3 people in discussion
Score automatically centre unless you wrap them in a block alignment template. An initial repeat sign requires a lot of arcane fiddling to make it happen and still look okay, some of which is not permitted on MediaWiki sites because it requires Guile statements. Thus I haven't attempted to reproduce that. Beeswaxcandle (talk) 06:58, 20 January 2024 (UTC)
Okay. Because I would never use an initial bar line before the first note (it's bad music typographic practice) when setting scores away from here, this didn't occur to me. Beeswaxcandle (talk) 21:59, 20 January 2024 (UTC)
Adding links to Wikipedia/Wiktionary
Latest comment: 9 months ago12 comments5 people in discussion
I recently added links to WP and Wiktionary to provide context for obscure words, then realized this seems to be a pretty controversial issue. Should I a) undo these additions, b) leave them as is, or c) do something else? Here are the links I added:
@Cremastra: I belong to those who believe such links should be used only very exceptionally, if at all. But if you have already added them, do not bother about it too much, you can leave them and if somebody will find them disturbing, they will remove them. --Jan Kameníček (talk) 23:17, 18 January 2024 (UTC)
It's nonsense. I am using them for over 10 years. I regularly link words to Wiktionary and Wikipedia. The links exist for the benefit of the readers. — ineuw (talk) 06:26, 20 January 2024 (UTC)
There are technical components (e.g. how links display on various devices causing undue emphasis) but that is not really the main issue, which has more to do people's beliefs about their "value" and "cost" as well as faithfulness to the text. The particular example mentioned here is mentioned as acceptable in the policy: Wikisource:Wikilinks#Context-appropriate_links but that does not mean people find it desirable. MarkLSteadman (talk) 04:49, 22 January 2024 (UTC)
@Ineuw: That you've done something for ten years does not make it policy, it just means nobody has called you on it yet. Neither version of our links policy permitted linking words inside a poem to a Wikipedia article. Wiktionary links are permitted for cases such as the one asked about here, but only very sparingly. Xover (talk) 06:22, 22 January 2024 (UTC)
Specically: "With poetry or fiction, little or no wikilinking may be more appropriate. However, archaic or obscure words may be wikilinked to their definitions on Wiktionary to aid the reader" There is the last passage about older works and cultural references, however ... MarkLSteadman (talk) 14:43, 22 January 2024 (UTC)
@EncycloPetey: {{normal}} only resets italics, not small-caps or other formatting. For small-caps you want {{font-variant normal}}. You can also do custom formatting for a cell through Index Styles by adding a unique class to a table cell or row (choice depends on which approach is least tedious to do).PS. There are some naming conventions for CSS classes at Help:Page styles#Class naming conventions. You don't want to use "work_toc" as a class because it might clash with a class from a template, extension, skin, etc. Use _toc (yes, just an underscore at the beginning) or, if the underscore bugs you, use the actual work's name Sunset_Gun_toc. If you need a throwaway class name for a single thing use two underscores (__).@ShakespeareFan00: Think first, modify stuff later, ok? :) Xover (talk) 20:37, 21 January 2024 (UTC)
I updated {{normal}} stylesheet to do the smallcaps escaping as well. It also meant I could migrate a vast number of uses of {{normalcaps}} which should be deprecated at as duplicate of {{{1}}}. It would be nice if someone could look at all of these and others like {{nobold}} and {{noitalic}}/{{upright}} with a view to having ONE consistent naming approach.
For reasons {{normal}} doesn't cancel bold formating because of some specfic use cases in EB1911. If specfic uses cases can be resolved, I've got no objections to Template:Tlnormal being a standard format canceller/"escape" template.ShakespeareFan00 (talk) 21:08, 21 January 2024 (UTC)
@Xover: Perhaps you can explain why Templatestyles DO NOT work for span based elements?
@ShakespeareFan00: {{normal}} is documented to only cancel italic formatting. That it is possible to do something does not necessarily make it a good idea to actually do it.I also do not understand what problem it is you're seeing in the linked page, so some explanation would be nice. Xover (talk) 06:46, 22 January 2024 (UTC)
@ShakespeareFan00: TemplateStyles do not work when the first instance of it is inside a wikilink. If there is at least one invocation of the template outside a wikilink and before the one inside the wikilink, it works fine. It's a known issue with no easy solution (but an half-ugly workaround: put a dummy call to the template in the header). Xover (talk) 10:48, 22 January 2024 (UTC)
What I was attempting to do was set-up 'escaping' rules in the style-sheets (see {{Italic block}}'s style for one use case, where thanks to template Styles it's possible to use '' '' as escapes (meaning a lot of Template:Tlnormal calls can be eleminated. It would be nice if someone sat down and has a look at {{boldblock}}/{{nobold}} as well. ShakespeareFan00 (talk) 10:56, 22 January 2024 (UTC)
Further trouble with section transclusion
Latest comment: 9 months ago3 comments2 people in discussion
I'm trying to transclude the poem The Irish Guards from Index:Rudyard Kipling's verse - Inclusive Edition 1885-1918.djvu, which is split across three pages, starting halfway through one and ending halfway through another. 242, 243, and 244. Once again, the entire third page (244) is getting transcluded, instead of stopping at the label. I'm using the code <pages index="Rudyard Kipling's verse - Inclusive Edition 1885-1918.djvu" from=242 fromsection="The Irish Guards" to=244 tosection="Pharaoh and the Sergeant"/>. Please, what is going wrong? Thanks, Cremastra (talk) 19:07, 21 January 2024 (UTC)
It doesn't stop at the label, it transcludes the section between to the section begin and section end tags on the "to" page. So that would include only the new poem at the bottom of the page, and not the the portion above it. MarkLSteadman (talk) 19:17, 21 January 2024 (UTC)
@CalendulaAsteraceae I think you have a problem here: --> local formattedStatements = mw.wikibase.formatValues(statements)
another way could be: local formattedStatements = item:formatStatements(args.prop, {mw.wikibase.entity.claimRanks.RANK_PREFERRED, mw.wikibase.entity.claimRanks.RANK_NORMAL}), skipping the getBestStatements before.
::BTW, in case you are not aware, this is very convenient for debug purposes , so you do not have to save the page every time. Mpaa (talk) 22:07, 21 January 2024 (UTC) OK, you mentioned it above :-)
Index page strange alignment
Latest comment: 9 months ago4 comments3 people in discussion
@EncycloPetey: Yes, I’m using it for experimentation. Feel free to blank the Index Styles page if you want to work on it and the weirdness is getting in the way. Xover (talk) 17:47, 23 January 2024 (UTC)
As long as it's a deliberate thing for experimentation purposes, then I totally understand. My concern was that there might be some unintended cause at play. --EncycloPetey (talk) 18:16, 23 January 2024 (UTC)
Latest comment: 9 months ago7 comments4 people in discussion
What happened to progress bars that showed the number of pages validated, proofread, not proofread etc? Are they coming back? I found them very useful Joannaszulcm (talk) 19:04, 16 January 2024 (UTC)
The underlying issue in Proofread Page should hit just about now if all goes according to plan. I haven't checked the state of our local workarounds so it may need further action here after that's done. Xover (talk) 17:39, 17 January 2024 (UTC)
Latest comment: 9 months ago3 comments2 people in discussion
In Module:Citation, one of the first things I do with the arguments is this:
-- lowercase args and remove hyphensfork,vinpairs(args)dolocalnewk=string.gsub(string.lower(k),'-','')mw.logObject(k..' to '..newk)args[newk]=args[newk]orvend
Latest comment: 9 months ago11 comments2 people in discussion
I am working on an inactive Wikisource project and need to figure out how the information that appears on index pages (Title, Author, Year etc) is bound to the form that appears in the edit mode. I am attempting to update the messages (MediaWiki namespace) and modules that begin with Proofreadpage, such as MediaWiki:Proofreadpage index template and Module:Proofreadpage index template, to be in line with the latest revision of this project (except localised) but it all sort of looks like magic to me.
As I understand it, MediaWiki:Proofreadpage index template is the backend for index pages, but how it all comes together I can not say. Our current version of that message has some localised parameters but I can not figure out where they are coming from and how they bind to the information that appears on index pages and the form that appears when index pages are being edited. I attempted to copy-paste the English versions into our project and then localise them but that just broke index pages completely and the binding between form fields and index information. I also got seemingly nonsensical Lua errors about Module:Proofreadpage index template not existing, which it very much does. Stefán Örvar Sigmundsson (talk) 13:16, 23 January 2024 (UTC)
@Xover: I have faithfully copy-pasted the English messages, templates and modules that begin with Proofread in their names to my local project. I have only localised MediaWiki:Proofreadpage index data config.json. Some of the messages, such as the ones beginning with Proofreadpage quality need not be created as they appear to exist on translatewiki.net.
Before I updated MediaWiki:Proofreadpage index template, I could notice that on index pages, the localisation was taking place, but the binding between form fields in edit mode and index properties (or values or whatever they are called) was severed. I hoped that once I updated MediaWiki:Proofreadpage index template, the binding would be restored and everything would work, but alas, when I view index pages I get: "Lua error in mw.text.lua, line 25: bad argument #1 to 'match' (string expected, got nil)." I will share some links with you to my local equivalent of the pages we have discussed.
@Stefán Örvar Sigmundsson: I'm looking into it. I'm pretty sure I know what's going on, but there're some tricky issues related to this being an existing Wikisource that I need to figure out before pinpointing a fix. In the mean time it's best not to edit any pages in the Index: (Frumrit:) namespace. Xover (talk) 13:24, 24 January 2024 (UTC)
@Stefán Örvar Sigmundsson: Ok, I think I've got a handle on this.What's going on is… When you edit a page in the Index: namespace, Proofread Page reads the configuration in s:is:MediaWiki:Proofreadpage index data config.json and constructs an edit form with fields based on that config. It populates the fields with the values it reads from the Index:. Before the field configuration was moved to a JSON file Proofread Page used s:is:MediaWiki:Proofreadpage index attributes for this configuration. Once you're done changing the edit form and hit the "Publish" button, Proofread Page grabs the current values of the fields, wraps them in a call to s:is:MediaWiki:Proofreadpage index template, and writes that to the database. When MediaWiki goes to display a wikipage in the Index: namespace it (grossly simplified) finds a more-or-less normal wiki page, calls the template with the arguments and values provided, and returns the results to the web browser.The existing Index: pages on isWS were created before you imported s:is:MediaWiki:Proofreadpage index data config.json, so Proofread Page used s:is:MediaWiki:Proofreadpage index attributes when constructing the edit form, and then saved those key names as parameters. The reason it's blowing up when displaying the Index: now is that s:is:Module:Proofreadpage index template isn't getting any of the parameters it's expecting (i.e. those with the same names as the JSON config), because what's actually saved are the Icelandic parameter names. When you try to edit an Index:, Proofread Page uses the JSON config, not the template, and hence it doesn't blow up, but since there is no data saved in those parameters it just displays an empty edit form. If you save an empty form you'll get an Index: with the new parameter names but with no data in any field.The quickest fix is probably to just delete the JSON config and revert the template to the previous version. If the Index: pages look ok then, then it is probably also safe to edit them.The more involved fix is to have a bot convert all the old parameter names to equivalent parameter names that exist in the JSON config (and that the module knows to look for). I can run a bot through for you, but you'll need to tell me the correct mapping for the following parameters (I can guess some of them, but better if you provide the mapping):
Afnotaleyfi
Ár
Athugasemdir
Bindayfirlit
Bindi
Frumrit
Höfundur
Mynd
Ritstjóri
Röðun
Síður
Staða
Staður
Titill
Útgefandi
Þýðandi
All of these have been present at some point and exist in the existing Index: pages, but not all of them have been used and not all of the ones that have been used are sensible to preserve. e.g. one of them is a general comment field (that sort of thing should go on the talk page), and one of them is "Original source" (that should go on the associated File: description page).Once the existing Index: pages have been converted you'll have… Index: pages showing with English-language labels, because I haven't made the module internationalized yet. It should (I hope) be fairly straightforward to make the visible strings configurable, so that you can localize them in the /config, but it might be a while until I have time to sit down and get it done. Whether you want to go the simple route or the more involved route is up to you. Now is not the greatest time to be importing the enWS module because it's in the middle of a migration so it contains a lot of mostly duplicate code, and some stuff that's just cobbled together temporarily. On the other hand, you are certainly going to want to move to a JSON config and a module-backed index template eventually, so it might be just as well to take the pain now while you're getting the project up and running again rather than later on down the road. I'm happy to help either way, but I can't guarantee my response time. Xover (talk) 21:57, 24 January 2024 (UTC)
@Xover: On the Icelandic Wikisource there is an agreement to simply copy-paste English-language templates and modules and only make the necessary changes to localise what the reader sees, not what the editor sees (such as template names and parameters). We reached this conclusion late last year due to the low activity in recent years and the added work of translating everything. I have deleted nearly all the native templates and replaced them with the current revisions of the enws ones. This way we will be able to update things as they are improved here and adopt new features with the least amount of re-engineering.
I have posted a notice on the local noticeboard about the index namespace being currently out of order as things are updated. I think we should carry on with the more involved route. Better to future-proof the project once and for all.
As I understand it, the list of parameters you gave me are supposed to map onto values in the JSON file. The problem is that I can not find corresponding values for all of them. Here is my translation with notes:
Afnotaleyfi = License
This means license, or more literally, usage license. There does not appear to be any such value in the JSON file.
Ár = Year
Athugasemdir = Remarks
The JSON file contains both Remarks, which this most likely maps onto, but also Notes. The word can mean either really.
Bindayfirlit = Volumes
Since volume in Icelandic is uncountable, this is probably supposed to correspond to Volumes in the JSON file. It literally means volume overview.
Bindi = Volume
Frumrit = Source
This can mean original (noun) and I suppose source too. For some reason the original editor of the project also chose to use this word for the index namespace. Maybe index sounded too generic to them for source material.
Höfundur = Author
Mynd = Image
Ritstjóri = Editor
Röðun = Sorting
This is most likely the key value in the JSON file which is used for sorting.
Síður = Pages
Staða = Progress
This literally means status but I can see why they would use it for progress.
Staður = Address
This literally means place. I think it is the Address value in the JSON file.
Titill = Title
Útgefandi = Publisher
Þýðandi = Translator
I notice that some of the values in the JSON file are not there. They include: Type, Language, Illustrator, School, ISBN, OCLC, LCCN, BNF_ARK, ARC, DOI, Transclusion, Validation_date, Width, Header, Footer, Notes. Maybe they were added to enws in a later revision, after the translation took place on isws. Stefán Örvar Sigmundsson (talk) 04:23, 25 January 2024 (UTC)
@Stefán Örvar Sigmundsson: Ok, I've migrated the existing Index: pages so they use English parameter names under the hood, and it looks like it's working now. Index: pages should be safe to edit again now, and should now use the parameters from the JSON file. When you edit an existing Index: now you may see "phantom" changes in the diff. This is because the JSON config provides default values for some parameters that are not present in the old Index: pages. This should be entirely harmless.I didn't preserve the |Afnotaleyfi= parameter, mainly because that would require editing the JSON config which in turn requires +sysop. I also strongly suggest you do not put this information on the Index: page, in favour of using licensing templates on the mainspace where it is transcluded. If you do want to have this field you can just edit the JSON config to add it (let me know if you need help with this).Also, as mentioned, all the user-visible field names in Index: pages are in English now (they'll be in Icelandic when you edit it). I plan to make these strings configurable so you can localize them, but I make no promises about when I'll get around to it. At the same time I'll look into making the various automatically added category names configurable too, but those might be harder to do.Other than that you should be good to go. Feel free to grab me if you find something broken (or if there's anything else I can help with). If you drop a note on my user talk page on isWS I'll get a notification, which might be more reliable than a ping. Xover (talk) 21:43, 25 January 2024 (UTC)
It's the "just" that's the problem. If you edit the module directly it'll get overwritten the next time you import from upstream (here). So the plan is to internationalize the module here, so that you (and any other project that wants to import it) can localize it using configuration that stays unchanged even if the code changes (i.e. either in /config or in a separate /strings, or some other suitable way). Internationalizing the code is a middling large job, complicated by the fact the code is in the middle of a migration here (lots of duplicate code etc.), which is why I'm saying I can't commit to any particular timeline for it. I realize it's annoying to have English labels for the fields when the Index is displayed, but there's no quick fix for that.There's no on-wiki way to view the raw markup for Index: pages that I'm aware of. You have to do it through the API (which was why I had to use a bot to migrate the indexes on isWS even though there are just 19 of them). Xover (talk) 10:51, 26 January 2024 (UTC)
@Xover: I see what you mean. It would be very nice and thoughtful of enws to structure its templates and modules in such a way as to make them easily importable. I created a template to be included on /doc subpages of templates and modules explaining that the page in question was imported and should be changed as little as possible during localisation. In most cases, nearly nothing has to be done other than change strings which can then be copy-pasted back into later revisions when an update is brought in from enws, with the help of diff. This works in most cases and only a few things such as Module:Author need to be modified as the code for processing birth and death dates for categorisation is very English-specific. The template I created states that any further modification behond string localisation should be documented in the /doc subpage of the page in question. I think that this approach will have to do for now. I just figured out that I can use AutoWikiBrowser to view the row content of Index: pages. Thank you so much for all your help. Stefán Örvar Sigmundsson (talk) 11:50, 26 January 2024 (UTC)
Lua equivalents of #time and #dateformat?
Latest comment: 9 months ago4 comments3 people in discussion
Working on Module:Citation, one issue I've run into is that {{citation}} is doing some things with parser functions that I'm not sure how to replicate in Lua. Specifically, I'd like to accomplish the following without resorting to frame:callParserFunction:
But… Why are you faffing about with a citation template that 1) we generally shouldn't use (there are very few legitimate use cases for citation templates), and 2) we certainly shouldn't base on {{citation}} if we were to use it, since it was obsoleted by the CS1/2 modules on enWP years ago? Xover (talk) 22:06, 25 January 2024 (UTC)
@Xover: Thank you! TBH this was mostly a Lua-learning exercise for me. I wouldn't have wanted to sink too much time into it, but a couple days was fine and I actually think I learned a lot. —CalendulaAsteraceae (talk • contribs) 23:07, 25 January 2024 (UTC)
Download/print sidebar
Latest comment: 9 months ago1 comment1 person in discussion
I am trying to find where the source of the sidebar "Download/print" is located. I know that MediaWiki:Sidebar stores the source for the navigation section above it. I need to make a modification to the sidebar in a sister project. Stefán Örvar Sigmundsson (talk) 20:39, 26 January 2024 (UTC)
Blank ppoem throws a stray DIV.
Latest comment: 9 months ago5 comments4 people in discussion
The blank call was found in reference to code in {{Tpp}} where a blank ppoem call had arisen in the absence of a quoted portion. {{tpp}} should be made substable and removed. ShakespeareFan00 (talk) 23:55, 25 January 2024 (UTC)
Latest comment: 9 months ago4 comments4 people in discussion
The {{WD author}} does not work in the third row of the list in Rapunzel, with the message "Lua error in Module:WD_version at line 41: attempt to index field 'datavalue' (a nil value)." Any ideas what is wrong? -- Jan Kameníček (talk) 19:20, 2 February 2024 (UTC)
The documentation for {{WD author}} states it's for Author pages. Rapunzel is a work page, so there may be something odd going on there. [Doesn't mean I agree with it's use, but that's a different debate.] Beeswaxcandle (talk) 21:34, 2 February 2024 (UTC)
@TE(æ)A,ea., @Xover: Thanks. Looks like phab:T357242 is a dupe, but I saw that task first (because I have the bad habit of going from bottom to top in my emails!). It was an issue with the temp directory filling up. It's cleared now and the exports should be working again, and it will hopfully not happen again (have put a fix in place). SamWilson09:41, 12 February 2024 (UTC)
error messages
Latest comment: 8 months ago3 comments3 people in discussion
@Ltbdl: Thanks for the report. It should be fixed now.@CalendulaAsteraceae: It was an undeclared variable tripping up 'no globals'. Regression test added here. Note also that this shows up an issue with handling "c." (it gets doubled). Xover (talk) 12:52, 12 February 2024 (UTC)
For example, the footnotes in Page:Holy Bible Berean Standard Bible.pdf/119. In the page, the numbers shown only refer to the verse number (e.g. 12, 18, 22, 28). But if I try to transclude the page to the main namespace, for example, to Bible (Berean Standard)/Exodus, ideally the footnotes should include the chapter number (e.g: 34:12, 34:18, 34:22, 34:28) for ease of reference. The template would be like:
@Dziego~enwiki: That wikipage is 75 printed A4-sized pages. Instead of getting fancy with templates and making custom navigation that doesn't exist in the text as published it would be better to rethink the structure. If you put each chapter on its own (sub)page the need for navigation and dynamic verse numbers goes away. Generally, and as a rule of thumb, whenever you go looking for a template that changes output based on namespace you should step back and rethink your approach. There are valid uses for such templates, but in most cases the need for one is a red flag that something in your approach needs a critical assessment. Xover (talk) 15:48, 14 February 2024 (UTC)
From the first (bold line) of the documentation, scan page link uses an offset for the second parameter. So, for indices where the offset never changes, like in your example, you would use {{spl|1|22}}, {{spl|13|22}}, {{spl|35|22}} etc., as the djvu number is always offset from the page number by 22. If the offset is not constant (depending on how blank pages/images are dealt with), or if you find this confusing, you can use {{scan page link 2}}, which has the parameters flipped around, but is perhaps more intuitive, i.e. {{spl2|23|1}}, would put a link to djvu/23 and display a "1" in the table of contents. Hope that makes sense. TeysaKarlov (talk) 03:44, 14 February 2024 (UTC)
CSS doesn't do text along a path. Wikimedia doesn't (yet) have a post-script module, so LaTeX/pstricks is not available here either. You can use an image for this, but a non-silent majority of the editors here would prefer just simple {{xxx-larger}} text.--RaboKarbakian (talk) 15:44, 16 February 2024 (UTC)
For instance, I scanned and cleaned up the factotum initial to the right with the letter T, for a text I'm transcribing. I put the T in the SVG and used the whole thing like an ordinary drop cap image.
No, there's no way to do this with {{di}}. It's also excessively complicated. Just use a regular image for these. If MediaWiki ever allows that level of dynamic content for SVG it might be possible in a sane way, but for the foreseeable future the answer is no. Xover (talk) 08:20, 24 February 2024 (UTC)
Search help
Latest comment: 8 months ago4 comments3 people in discussion
What do I need to enter in the search box that will only find specific values in the header template author field, in my case John Keats? Thanks Chrisguise (talk) 01:35, 27 February 2024 (UTC)
@CalendulaAsteraceae: Be very very careful when using insource:// searches. They are extremely expensive and only one can run on the site at a time, so unless you know what you're doing the odds are you'll end up with a query that times out and in the meanwhile blocking others from such searches. It's not something I would recommend casually to most normal users. Xover (talk) 05:30, 27 February 2024 (UTC)
@XoverThank you. I'd been trying to figure out how to do something with the 'insource' approach based on the search link from disambiguation pages, but I'll keep away from that. Chrisguise (talk) 07:32, 27 February 2024 (UTC)
Drop initials sit too high?
Latest comment: 8 months ago8 comments6 people in discussion
I'm trying to create a drop initial that is only *slightly* larger than the surrounding text (here). However, when I set the font-size of the drop initial to "larger", it also shifts it up higher than the surrounding text, instead of dropping below. Does anyone know why this is? —Beleg ÂltBT (talk) 15:26, 21 February 2024 (UTC)
Ok, I did some experimenting at User:Beleg Tâl/Sandbox#Slight dropinitial and found the source of the problem: the template style sheet Template:Dropinitial/styles.css hard-codes the line-height of the dropinitial at 1em, which is smaller than the default line-height of 1.6em in the Vector skin.
Drop initials are a horrible hack. You can do one drop initial with every aspect fixed fairly reliably, but generalising it is impossible (one major reason is that the "correct" value for several aspects is relative to the computed font size and line height of the surroundings, which is not exposed to CSS). So {{di}} is and will always be for "rough approximation" that will fall down when things around it change.The only way to get really reliable drop-initials is for browsers to implement them natively. There's a draft spec and several of the big browsers support it, but Firefox has zero support and the support in other browsers comes with several caveats. We're a lot closer now than e used to be, but there's at least several years left before we can do these properly.In the mean time, my strong recommendation would be to futz as little as possible with {{di}}. Setting the size of the initial is fine, as is using an image. But all further tuning you do is essentially going to turn into a backlog at some point, potentially one that needs to be dealt with manually, and the more tweaking you do the harder it will be to migrate once we get a real solution for it. Xover (talk) 08:37, 24 February 2024 (UTC)
@CalendulaAsteraceae: Apologies for not uploading earlier (but did consider it). Here is the image Subsequent images have the same issue. I will try it with other browsers because of possible issues with Firefox 123 and my traditional longstanding modifications for editing on the web. — ineuw (talk) 03:22, 5 March 2024 (UTC)
Latest comment: 8 months ago3 comments3 people in discussion
I see these two "contributors" (which I think are actually some sort of OCR correction tool rather than a human contributor), referenced in a lot of our non-scan-backed works from IA. An example: Talk:Scribner's Monthly/Volume 3/Number 1/The Mullenville Mystery. So what exactly is this technology? I've looked across the web and Wikisource many times and can't find what this refers to. I'd like to figure out exactly what "rag" is because I'd like to see if I can use bits and pieces of this technology for my own work. SnowyCinema (talk) 17:19, 4 March 2024 (UTC)
Akme is one of our most long-standing, prolific, and skilled proofreaders (it's a crying shame that they have an aversion to using Proofread Page, but their transcriptions are excellent all the same). The "ragpicker" stuff is just them having a bit of fun in the textinfo template. Xover (talk) 18:57, 9 March 2024 (UTC)
Help uploading: The Land of Fetish
Latest comment: 8 months ago4 comments3 people in discussion
How to download the Wikisource-markup contents of a book?
Latest comment: 7 months ago3 comments2 people in discussion
Is there a way to download the Wikisource-markup contents of a book? i.e., including all template tags and other Wikisource markup, in order to do a global search for use of certain templates/etc in a book, rather than searching the contents one page at a time (in page Edit view). Harris7 (talk) 19:52, 12 March 2024 (UTC)
I was just wondering if wikisource had classes that were not located in template css. I thought about that because some templates used custom classes but did not have the templatestyles mention. After looking closer, it looks like even those had a css subpage. I was wondering if there was a page containing wikisource-specific classes for templates that were not in a styles.css subpage of a template. Alien333 (what I did and why I did it wrong) 12:04, 15 March 2024 (UTC)
There are some global styles that used to live in MediaWiki:Common.css, but which have now been moved to default Gadgets. These are being actively reduced wherever possible, and are generally not used to style template output. Legacy templates use hard-coded inline style attributes, but we're migrating as many as possible to use TemplateStyles (there are issues with using TemplateStyles inside wikilink markup that prevents us moving some templates). Proofread Page also supports per-work styles in subpages of the Index: namespace. These and some naming conventions for classes are documented in Help:Page styles.As a rule of thumb, for new code, wst-* is a class added by a template; wsg-* by a Gadget; ws-* is a global style. Per-work styles use _* class names, or __* for once-off styling (things you might otherwise put in inline style attributes). On the Main Page we also use a enws-* prefix but that's somewhat of a special case. Also note that some templates may add classes for which they do not specify styling. There are several reasons this is so, one of them being to provide something for user or per-work styles selectors to target. Xover (talk) 12:40, 15 March 2024 (UTC)
Latest comment: 7 months ago3 comments2 people in discussion
Hello, I'm from Wikisource Indonesia. I'm courious, for {{missing pages}} it says, "Placeholder should be inserted....". If the proofread has been done, how to create a placeholder page? Thanks. Mnam23 (talk) 01:12, 25 March 2024 (UTC)
Two steps are generally required: 1. The actual insertion of placeholder images into the scan file and then reuploading to Commons / Wikisource. 2. The movement of the pages to reflect the updated source file. The first depends on the file format, for example djvus are typically easy while PDFs are trickier. Help:DjVu_files#Inserting_a_new_pages_(e.g._a_placeholder) is an example of how to do that. The movement of large numbers of pages can be automated by requesting a bot / admin to bulk move as not to lose history. For help you can ask in the Wikisource:Scan_Lab. MarkLSteadman (talk) 01:21, 25 March 2024 (UTC)
Just making sure, if the placeholder page is already inserted, the template should stay on the Index page, just changing the parameter placeholder=yes, right? Mnam23 (talk) 07:59, 26 March 2024 (UTC)
Copyright question
Latest comment: 7 months ago7 comments4 people in discussion
My personal opinion is that this is in the public domain in the US, since it does not contain a copyright notice and it was published before 1977. I think academic papers generally count as published, but I'm not 100% sure. @TE(æ)A,ea.: What do you think? SnowyCinema (talk) 12:45, 25 March 2024 (UTC)
It's also Q119716034 for some additional info. On The Caltech page, it says, "Thesis Availability: Public (worldwide access)", which I don't believe means the copyright status. It also says "Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided." Which sounds more like a copyright status, but I'm not sure. Nobody (talk) 13:45, 25 March 2024 (UTC)
Nobody, SnowyCinema: We had a lengthy discussion about this in the past, see here. The result of that discussion, as stated, was that there was no precedent, but that the thesis in that case was not copyrighted. I do not see how the decision is not precedent, as the arguments in that discussion were general and not relating to that thesis in particular. My opinion, and the sources of that opinion, are stated in more detail in that discussion. In sum, the publication is voluntary, and it is “general” because anyone can access the copies which are available in the university library’s collection. Because there was a general publication, and there was no copyright notice, it is in the public domain. TE(æ)A,ea. (talk) 17:36, 26 March 2024 (UTC)
Latest comment: 7 months ago2 comments2 people in discussion
Still very new to Wikisource, so sorry if my question seems dumb. After creating NOAA Storm Events Database – 2023 Matador tornado, I tried following the Index directions to create Index:NOAA Storm Events Database – 2023 Matador tornado, but the “Index” tab still doesn’t seem to show up on the text source. I’ve been working with another editor to add works of NOAA (U.S. government) (Category:PD-USGov-NOAA), and I’ve been struggling with figuring out how to do the Index pages with the few texts we have up. Can someone give some Index page assistance for the Matador tornado’s text as well as with the couple of other pages in that category. I’m not sure what I’m doing wrong exact. Thanks. WeatherWriter (talk) 16:11, 26 March 2024 (UTC)
An index works with a file. It is mostly a handy way to organize pages of that file.
The source for this appears to be a website. There is no point having an index without a file.
@WeatherWriter Since it's a PDF (as opposed to an HTML webpage), it's best to upload it to Commons and use the Index: namespace. I'll upload it to commons on the understanding that it is a U.S. federal government work. Cremastra (talk) 16:16, 1 April 2024 (UTC)
Latest comment: 7 months ago4 comments2 people in discussion
This is a bit of an odd case and I have no idea how to fix it. Page 155 (153 as numbered in the book) of File:The Salticidae (Spiders) of Panama.djvu appears to be blank. However, in the actual book it has text and in the file there is a text layer on that page that has the correct text correctly laid out. But for some reason the image for that page doesn't show the text. I was able to track down the physical book and scan the page in question, which is now at File:The Salticidae (Spiders) of Panama page 153.jpg. Is there any way to either integrate that file into the Index or repair the faulty DjVu file with it (preferably without losing the existing text later)? I have not been able to find any other scan of the book other than the one at the Biodiversity Heritage Library, which is the source for our scan. Nosferattus (talk) 22:08, 3 April 2024 (UTC)
Latest comment: 7 months ago16 comments5 people in discussion
I've been working on Ginzburg's Legends of the Jews as my first Wikisource project. (And yes, I understand that it's a massive, probably foolhardy endeavor for a first project. But here I am.)
I've just finished with the first chapter of volume 1 and the corresponding endnotes in volume 5.
I'd really appreciate one or more experienced editors looking over what I've done so far and giving me constructive feedback. (And it seems reasonable to validate what I've done along the way.)
Is my error rate acceptable for a "proofread" text? Am I using the {{Authority reference}} template correctly? Am I inadvertently making more work for myself or others down the line? Heck, is there anything I'm doing straight-up wrong? Any feedback, positive or negative, is much appreciated. If I'm going to invest the time and effort to complete this project, I want it to be good, not just good enough. And I'd like to hew as best I can to Wikisource best-practices.
A few notes on what I've already done:
The endnotes are chock-full of abbreviations, and I found that preserving the distinction between word spaces and sentence spaces there improved readability a fair bit. I've kept the sentence spaces to 0.5 em, so hopefully that won't bother ardent single-spacers too much.
I have not done the same in the main text, as it didn't seem necessary. But if that consistency is important, I can certainly do the wider spaces in the main text as well.
There is a fair bit of German, Hebrew, Latin, and Greek in the endnotes. I've studied the first three languages formally and have confidence in my ability to edit/proofread the snippets of those languages. Hopefully, relatively few errors have slipped through. I am completely self-taught in Greek, however, and would certainly appreciate someone who has formally studied Ancient Greek to scrutinize those bits.
Is it normal for endnotes to take quite a bit more effort than the main text?
A few things (though I haven't been around for that long and am not that experienced) :
I saw while looking at a random page that you used {{SIC}} for an outdated spelling (marvellous). It is not for that, but only for typos. Outdated spelling and other things that seem wrong but intentional should be kept as they are.
Section titles should not be using ='s and the like, because they make formatting as it is made on-wiki and that was not in the original text. For example, on Page:Ginzburg - The Legends of the Jews - Volume 1.djvu/49, it should only be centered and sc'd, as ='s put in in bold and that was not in the scan.
Rather than raw wikitable formatting, you have {{TOC row r}} to take care of the titles in TOCs (the formatting you did also did not display after the second |+).
A maximum width is sometimes put in TOCs with |width=, although it is in no way an obligation, to make it more easy on large screens (It can become hard navigating a max-width ToC when the line number is far to the right). Example taken with random page : Tom Swift Among the Diamond Makers
>I strongly recommend against using templates for TOCs. [No I don't. It's the dot-leaders that are actually bad. Table wikimarkup over TOC templates is just a weak general recommendation. But I'm an idiot that keeps dashing off messages when I don't have time to actually check what I'm writing and end up saying dumb stuff. --Xover (talk) 22:12, 4 April 2024 (UTC)] I generally recommend table wikimarkup over TOC templates. None of the ones we have are technically sound, and using raw wikimarkup for table gives you both more flexibility and more robust results. The main downside is the learning curve, but when you figure it out it can be reused across all works. But if you insist on using ToC templates, at least stay away from the ones that try to fake dot leaders. They're oh so tempting, but they are also oh so broken and will create problems.Also, width is set on transclusion with dynamic layouts ("Layout 2" is a common constrained-width layout). Setting the width of the table directly is a bad practice and should be avoided, unless there is a really pressing need for it. Xover (talk) 20:52, 4 April 2024 (UTC)
Web standards do not support dot-leaders natively (there's a proposal that's been floating around for ages, but no forward motion). That means we have to fake them using other mechanisms, and all those mechanisms end up being directly contrary to how the standards are supposed to work. The result is that the implementations we have are very ugly, hard to support, have many problems even when they seemingly work, and are very prone to break when things around them change or they are used in a different context.For example, since there is no way to generate a dynamical-length line of dots one implementation spits out something like a hundred period characters and spaces and uses styling tricks to hide the overflow. That breaks on very wide screens (add more dots and spaces, and everyone pays the cost on every page load). The hiding uses a white background color, which breaks in dark mode and in ePub exports (can somewhat be worked around with ever more special-case styling). Hacking around web standards also leads to incredibly complex markup, which combines with stuff like the 100+ dot characters, and leads to those implementations outputting huge amounts of data that we pay for every single time those features are used. I posted a particularly egregious example in WS:S#Orley Farm Contents+Illustrations Lists.Note that my harping on about this is in the vein of advocating that these templates be deprecated. They aren't actually deprecated, so nobody is going to chide you if you do use them and this all is strictly speaking just one contributor's opinion. But I intend to keep bringing this up every chance I get in the hope that the community will eventually get tired enough of it to deprecate them just to make me shut up already. Xover (talk) 09:25, 5 April 2024 (UTC)
I feel a but stupid asking this and there is probably an obvious answer, but how does the table markup you gave on that page (<tr> <td style="text-align:right;">I.</td> <td style="font-variant:small-caps;">—the commencement of the great orley farm case</td> <td style="text-align:right;">1 </td></tr> ) do dot-leaders? To me it seems it only does the alignment. Or is it intentional, are you saying we should just omit dot leaders? — Alien333 (what I did & why I did it wrong) 10:29, 5 April 2024 (UTC)
The dots are drawn first, over the whole area; maybe the whole page. I know this because I tried to use it in the {{AuxTOC}} and changed the background color to Transparent, which exposed them. And then, the table elements have a background color to hide those dots.
The complicated templates, those using {{TOC begin}} are not as "bad" as the easier to use like {{Dotted TOC line}} and kin. The latter make a whole separate and different table for each entry (row). There is a category here, filled with indexes that use those (truly, much muchmuch easier to use) templates where several of the pages just simply will not draw due to the heavy requirements upon the computers here. Unfortunately, my links to that discussion and category are elsewhere, so you will just have to take my word for it. They are worse than uncrunched pngs for processing time and *terrible* for small dedicated devices (like ereaders and maybe phones).
Being an avid and devout lover of the dotted tocs, I have been trying (and succeeding) to avoid them for Table of Contents use, but, the occasional small table found within works -- well, I indulge, feeling guilty like eating French Silk Pie for breakfast (which is technically scrambled eggs). But really, don't use the simple templates ever. And if you do -- Inductiveload has a script/tool which will convert the simple ones into complicated ones. My link for that is somewhere else also -- perhaps Mr. IDL will show up and drop it here.
The advice about manually making the tables is very good and the sooner started to learn the sooner learned. TOCs, while many are simple, several have a fineness and uniqueness of character which will improve anybody's skill level.--RaboKarbakian (talk) 11:01, 5 April 2024 (UTC)
(I'd never heard of {{dotted TOC line}} before this discussion)
And then what about the {{TOC row}}'s? They are less problematic, but how problematic are they? I have been using more or less only them since coming here.
The best guidance we have right now is slightly camouflaged inside Help:Page styles. The most flexible, powerful, and robust way we have of doing tables right now is table wikimarkup combined with formatting applied through per-work page styles (the "Styles" tab on Index: pages; the CSS there is automatically loaded on all associated pages in the Page: namespace and in mainspace when transcluded using PRP). We need to make that approach more user friendly and reusable (ready-to-use snippets that can be copied), and create better documentation, but apart from the learning curve I heartily recommend it.And compared to general HTML tables we have some unique problems stemming from the fact that we don't write HTML directly; we write wikimarkup that gets parsed and rendered into HTML by MediaWiki, and which is presented inside the skin and site chrome, etc. Compared to your typical CMS we also do far more advanced formatting than, say, a journalist banging out a news story or whatever. And then there is the added complication for Wikisource that we have to split these already twice-abstracted tables across wikipages in the Page: namespace and make them work together when transcluded together. Put together this means tables are one of the most difficult areas for contributors to deal with (which is why the lure of TOC templates is so great: it seems obvious that there must be an easier way to do this). Xover (talk) 12:18, 5 April 2024 (UTC)
@Alien333: Oh, I'm sorry, I misunderstood. I am saying to simply not try to do dot-leaders, using any method, until web standards and web browsers actually support them natively. Everything else I write above is about tables and TOCs in general. Xover (talk) 13:13, 5 April 2024 (UTC)
The example omits the style sheet applied and other stuff. It's just meant to illustrate that the output from the templates is completely unreasonable. Xover (talk) 12:06, 5 April 2024 (UTC)
@Dave314159 It seems the responses above are drifting a little off topic, when it comes to the original question, which I will do my best to stick to. That said, so it is on record, I favor TOC begin, TOC row etc. While I have avoided validating anything for the moment, accuracy wise, the proofreading is good, and a few tips follow.
If you are doing anything with either strange formatting, or strange references, I strongly recommend transcluding the work as you go. In your case, I am not sure the endnotes are working, except those endnotes that you have specifically set the |transclude= parameter for. Because the endnotes are in a different volume, you may be forced to always use |transclude=, and may have to add this to every endnote thus far... That said, I am by no means an expert in endnotes. My main 'experiences' were with The History of Witchcraft and Demonology, which ended up working out eventually, endnote wise, but is far from finished, even with help. But yes, endnotes take an annoying amount of time. Other than that, if you need help with transcluding, ask away, although you seem to have a reasonable grasp of things so far.
Also, your headings do not seem to conform to wikisource styles, as far as I am aware. For example, I believe ==={{sc|The First Things Created}}=== should just be {{c|{{sc|The First Things Created}}}}, and similarly {{c|I<br>{{uc|The Creation of the World}}}} instead of == I: {{uc|The Creation of the World}} ==. Note that I removed the colon ":", because it isn't there in the first place, but thought I should finish this response before doing anything else to the text.
I also believe that manually setting sentence spaces is against wikisource styles (as in, do not force a double space at the end of a sentence, even if the original text does). As you can see in the example for the wsp template, it is for poetry and the like. To save you the trouble, a bot request can probably remove all of these, when the time comes.
I haven't formally (or informally) studied ancient Greek, so can't really help there (or with Hebrew), but otherwise, good luck with the project, and happy proofreading.
Good point about the manual spacing. I was also unsure about it. We in general do not preserve variable spaces; but in my quick scan it looked like they were here used to separate a headword from the rest of the paragraph. It is possible that this is an exception to the general rule. On the other hand, several of the cases did not appear to correspond with such use (extra-wide spaces at seemingly-random places within a paragraph?). But I didn't have time to look closer into it so I can't offer anything specific. Xover (talk) 07:19, 7 April 2024 (UTC)
Latest comment: 7 months ago12 comments6 people in discussion
Sometimes while proofreading, I finish a page where the lines are not joined, but then the last editing "line breaks" in a page create new paragraphs. I don't know why this should be- my usual workaround is just to have no line breaks on the source page for the last paragraph. But this looks messy and it is annoying. Is there some obvious reason that this happens that I can avoid in the future? @Mpaa, perhaps you know the answer, I know sometimes you prefer that lines not be joined, so you have likely run into this problem.
The {{nop}} template needs to be on its own line. Once you've done that, then the problem will go away. However, it is most preferable to remove line breaks within paragraphs as leaving them in is known to cause problems with transclusion. Beeswaxcandle (talk) 07:17, 24 March 2024 (UTC)
@FPTI@Beeswaxcandle I have experienced the same issue before. What seems to happen is that there are extra carriage returns in the text, which are not visible. If you turn on "Generate paragraph (pilcrow) markers, ¶, in the left margin of the Page: namespace to indicate HTML paragraph tag starts." in Preferences-Gadgets they can be seen. I had to log out and log back in to activate the gadget. I don't have this gadget activated normally but it shows what the problem is. After proofreading text I usually run a tool that cleans-up the OCR text, which removes these in-paragraph line breaks Chrisguise (talk) 09:53, 1 April 2024 (UTC)
@FPTI I do not know what the problem is, but I am fine with it as it is transcluded correctly. As the problem is only in Page ns, it might be some glitch in ProofreadPage Extension. Mpaa (talk) 09:57, 1 April 2024 (UTC)
It is not specific of Page ns. In Page ns it happens only when in the footer there is some content inside a div block. The same can be reproduced in Main ns appending a div block to the last line, e.g. something like of course it was only a few seconds. I had leisure<div class="x">vv</div> Maybe it might be the parser that, when it finds the div block, insert a paragraph break at the beginning of the line ... who knows .... Mpaa (talk) 14:26, 1 April 2024 (UTC)
Indeed. This is p-wrapping and a rather dumb regex-based (essentially) parser. The parser thinks that all text has to be wrapped in a p tag, so as it parses the wikitext it is looking for various triggers for when to add an opening p tag, and when it needs to add a closing p tag. The parser sees Page: namespace pages as a single blob of wikitext, with the header and footer merely marked as noinclude sections. When there is a div (i.e. what {{c}} spits out) in the footer the parser notices, and since div elements are not valid inside p elements it decides it has to close the currently open p. To do that it backtracks and guesses at where the paragraph it just closed should start, and decides it's at the first line break character it comes to (i.e. the one between the last and penultimate line). In other words, whenever you have anything in the footer that isn't permitted inside a p element, the parser will put the last text line into its own paragraph hanging loose at the end of the page.The even better news is that the shiny new parser (known as Parsoid) has been designed to be bugwards compatible with the old parser, so it will blithely reproduce this bug in exactly the same way as the old grotty parser. Oh joy!But the moral of the story is the same as the advice given by the old-timers from the beginning: always remove hard line breaks inside paragraphs! Xover (talk) 12:42, 8 April 2024 (UTC)
Actually over the time I tend to preserve new lines as printed. It makes Validation much easier. The page break is just a cosmetic issue in Page ns with no relevance when transcluded. Mpaa (talk) 16:45, 9 April 2024 (UTC)
Hello, can you tell me what tool do you use for cleaning the OCR? The same issue is happening in every page to me. HendrikWBK (talk) 11:29, 8 April 2024 (UTC)
@HendrikWBK: For unwrapping hard-wrapped lines I use a custom user script. You can add mw.loader.load("//en.wikisource.org/w/index.php?title=User:Xover/unwrap.js&action=raw&ctype=text/javascript");// Backlink: [[User:Xover/unwrap.js]] to your common.js to try it out. It adds a link in the left sidebar (in Vector 2010) or to the tools menu (in Vector 2022) labelled "↲ Remove hard line breaks" that does what it says on the tin. This was a hacky little thing I threw together for my own use, so no warranties. The source is in User:Xover/unwrap.js if you want to check what it does. Xover (talk) 12:55, 8 April 2024 (UTC)
Normally I would assume that it's not working because of some limitation in MediaWiki. However, I have used the exact same method for {{insular}}, and it seems to work fine there, so I am stumped. —Beleg ÂltBT (talk) 18:03, 3 April 2024 (UTC)
Latest comment: 6 months ago4 comments2 people in discussion
Over time I've made a number of requests for moves in support of fixing scans with missing pages, etc. These used to be dealt with promptly but more recently the response has been quite slow or they are still waiting (people are doubtless busy). I came across User:Inductiveload/Scripts/Page shifter.py recently. If someone can explain to me how to run this script where it is, or how to incorporate it and run it from my .js (if that's what's required), I'll happily do the fixes myself. Thanks, Chrisguise (talk) 13:47, 12 April 2024 (UTC)
@Chrisguise: If you have to ask I generally don't recommend you try to run this yourself. The script you link is a pywikibot script, and requires the technical chops to operate 1) in a *nix command-line environment, and 2) a big complex bot framework. It's not something you run on-wiki or in your browser, and it isn't really an interactive tool. Page moves due to updated source files in an existing index are especially challenging because there are so many factors you have to take into account to avoid messing stuff up.Page moves and a few related tasks have indeed slowed to a relative crawl lately. Inductiveload isn't active any more, and I am too busy IRL to be able to pick up the slack (and usually my head is too fried to take on anything complicated), and we used to take care of the bulk of these. If you have any such tasks that are especially bad in blocking your progress then please feel free to try bugging me on my talk page and I'll try to prioritise them.PS. If you want to make it require fewer brain cycles to run a page move request like that, specifying the necessary moves in the way the software expects (vs. what's logical for a human being) can help. The moves are specified as "page range to be moved" (123-345) and "offset" (+4, or -3, or...). And the destination page cannot exist, so when it's not a single unified shift you'll need to give these in batches with, typically, the last pages in the first batch: that way you make room for the later page moves. This way of thinking about it is really non-intuitive for humans, but it's what the computer requires, and it's often what takes the most time and effort when doing bulk page moves. Xover (talk) 14:17, 12 April 2024 (UTC)
Yeah, it's a bit more that than plain "anyone". Don't get me wrong, it's not rocket science, but the "computer geek" factor is pretty high. Xover (talk) 14:50, 12 April 2024 (UTC)
Latest comment: 6 months ago4 comments2 people in discussion
There is about 40-ish pages that I still need to upload for the document, but I can’t get the clickable pages to appear on the Index page and I keep seeing “Error: No such file” when I try to fix that. Can someone get the 40-pages red linked down there? WeatherWriter (talk) 16:07, 17 April 2024 (UTC)
@WeatherWriter: Why are you uploading this as individual image files? Proofread Page is designed to work with multi-page media formats like PDF and DjVu. Xover (talk) 16:28, 17 April 2024 (UTC)
Because it isn’t a PDF document, but rather a StoryMap. NOAA is weird and doesn’t like to create standard PDF-style documents. WeatherWriter (talk) 16:48, 17 April 2024 (UTC)
As always, posting about seems to have fixed the problem :). Looks like I was having the same problem as many of the people in an earlier discussion. In the same way as discussed there, I tried purging everything possible multiple times, and after about half an hour of nothing happening the pages suddenly appeared! Qq1122qq (talk) 14:05, 13 April 2024 (UTC)
Yes - I purged and hard purged (not sure of the difference) everything I could, and crossed my fingers. I imagine there's some very slow running process buried somewhere which only runs intermittently. I've never had this issue with .PDF files so I wonder if it's a .DJVU only issue? Qq1122qq (talk) 18:42, 13 April 2024 (UTC)
I don't think so, since Chrisguise reported that he had mostly had that problem with PDF's.
But had you purged it on commons, specifically? asking because there there is no easily accessible "purge" button like here, needs to be done through a gadget or some other things, and you could have missed it. Sorry for insisting, but it looked like we finally had a solution, and I want to be sure before abandoning it. — Alien333 (what I did & why I did it wrong) 08:31, 14 April 2024 (UTC)
I purged on Commons, then purged in WS, and the page fixed itself. This has also fixed 3 more projects I've created in the last few days.
I was given the magic incantation to purge a file on Commons via someone helpful on Discord - if you want for example to purge File:Methods_of_Operating_the_Comptometer_(1895).djvu
@Qq1122qq: If you go to the Gadgets section of your Preferences, down under the heading "Interface", you'll find a dedicated "Purge" Gadget that does just this for you (and which exists due to this kind of problem). I see the description talks about a purge "tab", but that's just a holdover from when Monobook was the default skin. It really appears in either the sidebar or the Tools menu, depending on which skin you're using. Xover (talk) 05:10, 18 April 2024 (UTC)
Sorry about that - I find the threaded wiki-style comment pages like this one very hard to parse so I had no idea it was still a live discussion. Qq1122qq (talk) 18:43, 13 April 2024 (UTC)
@Ineuw: It's an ongoing flakiness in MediaWiki and the integration between Commons and other projects. This file should be fixed now through black magic and dread incantations to the technology deities (or something to that effect). Xover (talk) 17:21, 19 April 2024 (UTC)
Thanks for the replies. After posting, I remembered that this is happening everywhere on the web. Systems and software are constantly being updated. I gained patience, understanding, and acceptance. In the meanwhile, I am sure to find something else to do. — ineuw (talk) 12:07, 20 April 2024 (UTC)
Latest comment: 6 months ago21 comments7 people in discussion
I have uploaded on commons a djvu file. It looks fine over there, but here and on wp the file is registered as being 0x0 (not as empty, though), and trying to link it just gives a link, not the file. Am I the only one having this problem, and can someone tell me how to fix it? — Alien333(what I did — why I did it wrong)19:48, 21 March 2024 (UTC)
@Alien333: So here's exactly what happened. I uploaded the file (revision 1) and it appeared as 0x0 just like you described, and the Index page's pagelist threw an error that I can't remember now. So, what I did was I uploaded another version of the same PDF, just directly taken from IA instead of modified by me in a certain way, and then uploaded that with chunk upload (revision 2). It initially seemed to be a working file, but then I deleted and recreated the Index again, to find out that this revision was now broken. So I reverted to the previous revision (revision 1) as revision 3, and somehow that worked.
TLDR: Revision 1 was broken and when I uploaded revision 2, that revision appeared to work at first. But then it reversed out of nowhere, to where revision 1 was working now and revision 2 magically broke in its place. . . . If this is confusing to you, it's exactly as confusing to me so don't worry. I have no idea what the issue was. ☹️ SnowyCinema (talk) 07:15, 22 March 2024 (UTC)
The error was "Invalid interval" (at least for me). What I think is that the file got broken during some sort of transfer from commons to here (since it was also broken on wp). On top of being 0 x 0 pixels, Invalid interval probably meant it was 0 pages long (even an empty pagelist got an error). The file was still registered as being 1.something MBs long (here, on commons, and wp). I do not think it was an issue with the file I uploaded, as on commons everything looked good. That's what I know. I also had a vague suspicion of it being linked to the server switch on wednesday, but it might have nothing to do with it. Alien333(what I did — why I did it wrong)07:29, 22 March 2024 (UTC)
Servers and backends and deployment are quite finicky, and there are a million mysterious things that can go wrong all of a sudden. It's hard to really say exactly what happened without talking to someone who has the keys so to speak. Our global multisite system is bound to be quite complex on the backend, I'll say that much... SnowyCinema (talk) 07:41, 22 March 2024 (UTC)
@Alien333@Beeswaxcandle@M-le-mot-dit@SnowyCinema I have experienced (and reported) the same issue. Although the problem sometimes seems to fix itself, I have a number of cases where it has not (>6 months, which I assume means never), and a number of cases where random pages in an otherwise functioning PDF are completely unreadable on WS. I've basically given up uploading PDF files (even though I have paid-for PDF editing software - Foxit) and default to DJVU. Converting PDF's to DJVU with 'Pdf 2 Djvu Converter' generally works OK. Chrisguise (talk) 13:17, 26 March 2024 (UTC)
I just had this same problem with Index:Florula Mortolensis.djvu, where the pagelist is broken and the file appears as 0x0 on WS (but works fine on commons. I used the IA upload tool.) Purging the caches didn't help, but hopefully it fixes itself, as it seems to. Cremastra (talk) 20:54, 27 March 2024 (UTC)
Indeed, after having the problem with Index:Poems Coolidge.djvu, clearing the cache on commons and then here fixed it. Maybe doing it in the other order doesn't work.
Thanks. I hadn't previously delved around the gadgets section on Commons. Doing the two purges has fixed one of my problem files (a DJVU, unusually). I'll try it on a PDF. Chrisguise (talk) 08:18, 13 April 2024 (UTC)
@Mpaa: It's looking for id="headerContainer" in the HTML output of the parsed page, and {{header}} et al are no longer outputting that id. I've now changed it to look for the common class name (ws-header) instead, so it should no longer be tagging edits like that. Xover (talk) 20:46, 27 April 2024 (UTC)
@FPTI: You follow the red link, delete the pre-filled content of the text box (the author template), go to the Advanced section of the editor toolbar, pick the button with the icon of an arrow swooping down and to the right, and then fill in the destination page. OR, you can simply type in #REDIRECT[[Author:Maurice R. Coons]]. Xover (talk) 06:52, 29 April 2024 (UTC)
Help is asked for with a CSS selector
Latest comment: 6 months ago6 comments3 people in discussion
I am trying to configure my User:Ineuw/vector.css to view the main namespace text with the following selector, and I am unable to get it right. Can someone in the know, help me with this?
/* main namespace page */
#mw-page-base #bodyContent{
font-family:FreeSerif !important;
font-size: large !important;
}
Adding a comma in between the two ID selectors should get it to work, but in all namespaces. I'm not sure how to select for a specific namespace. Cremastra (talk) 21:35, 1 May 2024 (UTC)
@user:Cremastra Thanks for the help, but like you said it affects everything. I tried every selector of the main namespace just to target the page content. I am now trying other selectors. — ineuw (talk) 07:39, 2 May 2024 (UTC)
@user:Alien333, Hurrah, and thank you! It is as I hoped. Interestingly my vector.css uses .mw-editform for page namespace selectors and not .prp-page-container. — ineuw (talk) 20:02, 2 May 2024 (UTC)
Not sure if this is what you were wondering, but prp-page-container is for viewing (in page ns, next to the image and all) and mw-editform is for editing (in all namespaces). — Alien333 (what I did & why I did it wrong) 20:11, 2 May 2024 (UTC)
Yes, this is exactly what I wanted to understand. I must return to the Wikimedia kindergarten. Thanks again.
I updated {{MathForm2}} and these uses should be migrated over, using |talign=end or suitable equivalent to obtain the same rendering as the RH approach.
The objective desired is to de overload uses RunningHeader, so that pending changes to it are potentially easier to do without unexpected side effects.
ShakespeareFan00 (talk) 09:37, 29 April 2024 (UTC)
@ShakespeareFan00: I have no idea what it is that you're saying here. Quite possibly that's just me being dumb, but could you try explaining again with a little more detail; especially as regards what is the problem that exists now, and what is the outcome you're looking for? Xover (talk) 11:11, 29 April 2024 (UTC)
The current problem is that RunningHeader is used in some works for formatting a three-cell table, or even more specfic an equation+number format. (something it wasn't technically designed for.).
The desired objective is replacing these usages with something more appropriate ( I'd updated {{MathForm1}} and {{MathForm2}} as they were intended for the use case concerned ).
@ShakespeareFan00 And then there is {{Numb form}}, and then there is .... I think these template are known by 0.1% of the users and used by 0.01%. I would like to see pro/cons of the different solutions before spreading templates. I think we already have enough. Less templates and more general-purpose is the way to go IMHO. Mpaa (talk) 21:12, 30 April 2024 (UTC)
As I said, I do not know. So for now I would not do anything.
It seems to me that rh is the more natural choice for a user, it is short, written as the text flows, anybody is familiar with it, etc.
This syntax is verbose and not intuitive: {{MathForm2|talign=end|2=<math>x = \frac{3}{4} \times \frac{\cos \beta}{4 + \pi \sin \beta} \times l ,</math>|1=(2)}} and I do not know if that template is a good implementation or not. Mpaa (talk) 08:55, 1 May 2024 (UTC)
For whatever that's worth, I have on my todo list to look into {{MathForm1}} and {{MathForm2}} and {{Numb form}} and {{disp right}} as well as the several other templates that are used for this kind of thing, with a view to finding one sensible replacement for all of them. {{numbered equation}} is a very embryonic first draft of an approach to explore what would work well (not a proposed solution/replacement; just a start at figuring out what the replacement should look like). And hopefully we can usefully abstract out some of the generic parts of {{rh}} to reuse for this and a couple of other use cases (quoted letters with salutations and dates, signatures, etc.). Xover (talk) 09:10, 1 May 2024 (UTC)
These are some possible uses of Running Header outside of Page namespace that should be taken into consideration:-
If you ever have to transclude an index whose name has both single and double quotes, you can replace the quotes in the name with " or ' as appropriate. —CalendulaAsteraceae (talk • contribs) 20:07, 4 May 2024 (UTC)
Unspaced page end for references
Latest comment: 6 months ago7 comments6 people in discussion
I also wasn't sure if I should modify it myself, but when looking through the help on footnotes, I read the following Other parts of the footnote should use the "follow" parameter with the same name. The position of this text does not matter to the footnote but adding it to the bottom of the page, where the footnote would normally be, is the most obvious position. Make sure that it is a part of the page that will be transcluded with the rest of the footnote: ie. not in the header or footer of the Page, not in tags, and—if sectional transclusion is being used—in the same section as the rest of the footnote (so all parts are transcluded into the same page in the main namespace). While perhaps correct that the position of the ref-follow does not matter to the footnote, unless I have misinterpreted something, a ref-follow placed at the end of a page with paragraphs created in the conventional sense (i.e. pressing enter), will incorrectly insert a line break at the start of the transcluded page. Thus, ref-follows should be placed at the top of the page body section (which also helps with sectional texts, as the ref-follow would typically be transcluded with the first section on the page).
I just tried using a nop, but that didn't work. The solution I would use is to bring the footnotes back from /63 and leave that page blank. Beeswaxcandle (talk) 01:28, 27 April 2024 (UTC)
I agree with Beeswaxcandle: this is just exactly a case for a pragmatic solution rather than chasing a "perfect" one. Move the contents of the ref to the page where it's referenced, and leave a comment on both pages to explain what you've done (<!-- comment here -->). Xover (talk) 09:31, 27 April 2024 (UTC)
I don't think you should worry that much about the leading space (I don't think it copies anyway). If you're really insistent on getting rid of it, you can put {{nop}} into the {{hws}} and {{hwe}} templates; if you put e.g. {{hwe|{{nop}}|Governor|hyph=}} on the first page and {{hws|Governor|Governor|hyph=}} on the second page, the text should transclude correctly without a leading space (the first transcludes as "Governor", the second doesn't transclude). (FWIW, as an alternative to M-le-mot-dit's solution of putting a comment into the first <ref>, on another page I put a dummy ref-follow in the footer with {{nop}} in it, but both of these give the leading space.) Arcorann (talk) 12:58, 27 April 2024 (UTC)
@Arcorann,@Beeswaxcandle,@M-le-mot-dit,@Xover Thanks all for the suggestions. M-le-mot-dit's seemed to work, a comment in the leading reference tag didn't create a leading space on the transcluded page, without having to move any text. Tempted to call it perfect and pragmatic. Although I never said I was chasing a perfect solution, moving text can be an unnecessary complication, but all sorted now at any rate. Thanks again, TeysaKarlov (talk) 20:15, 27 April 2024 (UTC)
Only a small chance this suggestion is useful, but here goes. For a different purpose (and if I recall correctly with the guidance of Billinghurst), I used a different method of creating notes in History of Oregon Newspapers/The Pioneer Period. (I'm not done, but the first 68 notes follow this format.) It's pretty fiddly and I can't say I fully get how it works, but if you follow the example of those pages you should be able to figure it out. Whether it's worth your time or not, I'm not sure :) I believe that code is intended for a case like I used it in (endnotes on a different page), and I have a hunch there might be a way to adapt it to your use case (though it might involve reworking the entire book). -Pete (talk) 06:57, 8 May 2024 (UTC)
Pages on ProofreadPage look bad, pages in PDF look fine
Latest comment: 6 months ago5 comments3 people in discussion
Now compare Page:The Hindu-Arabic Numerals (1911).djvu/14. It's a combination of IA over-compressing their PDFs, MediaWiki generally doing much worse on extracting images and text layer from PDFs, and MediaWiki having pathologically bad image quality results on some PDFs. The bottom line is: always use DjVu if at all possible. Xover (talk) 05:46, 8 May 2024 (UTC)
That was a single pathological case brought on by waay too aggressive compression settings and over-optimization. There is a general problem with PDFs (as above), and combined with some other issues with the format and tooling I strongly recommend preferring DjVu whenever possible. Xover (talk) 18:15, 8 May 2024 (UTC)
I agree with this. DJVU is also an open format, and though they admittedly might not be all that user friendly, there are more/better free tools available for manipulating them. -Pete (talk) 20:08, 8 May 2024 (UTC)
Help with Score extension
Latest comment: 6 months ago7 comments4 people in discussion
We will never have version 2.25. The odd numbered versions are the experimental ones. It will be version 2.26. In the meantime we need to use the Rehearsal Mark technique, which doesn't replicate in the midi file. Beeswaxcandle (talk) 20:50, 7 May 2024 (UTC)
Thank you! I'd think we could get the repeat in the MIDI file by adding a repeat and overriding the bars so they don't display, but just adding \bar "|" before the repeat doesn't seem to override the ".|:" bar. Do you know if that can be done with some more complicated syntax? —CalendulaAsteraceae (talk • contribs) 21:38, 7 May 2024 (UTC)
Some questions
Latest comment: 6 months ago4 comments4 people in discussion
JS scripts seem to be imported (or executed) in a inconsistent order, which, when multiple scripts add sidebar buttons, causes fumbling every time to find where exactly this time are the buttons. Is there a way to change that?
For poems, it can be hard to determine if there is or not a stanza break on a page break, when there is: a) no clear stanza pattern, b) no special tendency for stanzas to end with sentences, and c) no other edition of the same poem. The first two tend to disappear for more modern poetry and the third is rarely available (and disputable anyway). Is there another way to find it?
Policy says to always implement footnotes and endnotes with ref tags (or something to that effect). Often, footnotes use asterisks and I couldn't find a way to make ref use them. It could be done by manually putting an anchor on the footnote, and a link to the anchor, and transcluding the two separately, but it's quite a bother. Someone probably has encountered the same issue before, so there must be a better way to do it lurking around somewhere.
For (2), you have to use your best judgment; there's no way around that. I face the same issue in poetic dramas I've transcribed. The points you've listed are the same ones I use to make that call when it's unclear. The fact that you're giving it thought and attention is what's most important.
With regard to (3), we do not attempt to reproduce the symbols used for footnotes / endnotes in the original copy, for multiple reasons. If you hunt around, there have been several essays written on the subject over the years. Simply let the software generate a number. from the ref tags, and don't spend time fretting over the displayed number versus an asterisk. --EncycloPetey (talk) 14:56, 8 May 2024 (UTC)
Just wanted to mention that large numbers of references all marked identically with * isn't a good UX, considering the way we transclude entire chapters as flowing text with the notes at the end, typically. For situations like see Note * on page ... it is possible to link etc. MarkLSteadman (talk) 17:13, 8 May 2024 (UTC)
Regarding #1, the short answer is no. There is a longer answer involving using custom events to chain them or manually coordinating menu position, but they're not good solutions. Xover (talk) 17:05, 8 May 2024 (UTC)
Using fromsection correctly
Latest comment: 6 months ago2 comments2 people in discussion
At the following page, I am trying to use fromsection to limit the included section on the starting page, but for some reason nothing is included from the first page. Can I get some assistance on how fromsection should be used correctly?
Found the problem, it was due to the {{c}}, not fromsection. You opened it before the section break and closed it after. I'm not exactly sure how it works, but having the section break included in a template probably causes it not be parsed correctly, so the pages tag resorts to the usual method with unknown begins, it starts at the end of the page (which is why you still had the [70] pagenumber on the left, because it techically includes something from that page, it's just empty). — Alien333 (what I did & why I did it wrong) 13:50, 10 May 2024 (UTC)
Score music needed
Latest comment: 5 months ago3 comments2 people in discussion
Put {{sdelete|G7}} at the top (for more info about this, see WS:CSD and Template:Sdelete). Note: When you create a page at the wrong place, instead of deleting the old one and creating a new one, you can move (rename) it by going in the "Page" menu at the top right. — Alien333 (what I did & why I did it wrong) 18:39, 14 May 2024 (UTC)
If you do move a page like this (i.e. quickly after creation), please mark the old title as {{sdelete|M2}}. That way we can delete the wrong title. This criterion only applies within a week of creation. Beeswaxcandle (talk) 10:12, 15 May 2024 (UTC)
The pagelist is a tag and must therefore be opened with < and closed with />. In your case, there was </ pagelist >, which doesn't work (technically, it'd be a closing tag, and then you're missing the opening tag). I fixed it, normally, though you still need to put a pagelist (as of now, it's empty). — Alien333 (what I did & why I did it wrong) 18:24, 21 May 2024 (UTC)
Huh, so much for me. I also, as a matter of routine after that discussion, purged the file on commons, and it might be that, though it didn't appear to have immediate effect. EDIT: Could you translate me the error message? (I don't speak Hebrew). If it's that whole purging thing, the error should be "Invalid Interval" — Alien333 (what I did & why I did it wrong) 18:45, 21 May 2024 (UTC)
As far as the index: You've transcribed library annotations from the bottom. Typically, we don't transcribe writing added after the work was printed, whether library markings, stamps, book plates, etc. --EncycloPetey (talk) 21:17, 23 May 2024 (UTC)
Latest comment: 5 months ago7 comments2 people in discussion
I've been trying to find an event to trigger from JS to change the radio buttons for quality. $("input[value=3]").trigger("click") and $(".quality3").trigger("click") are the only two that appear to work, but they do not really, as the proofread button gets checked but the other buttons not unchecked and on saving the page is still red (or whatever it was before). $("div[aria-activedescendant][tabindex=5]").attr("aria-activedescendant", "14") didn't help either, so that's not what actually changes something. The relevant file appears to be there, and it even looks (from the comments ~l.5400) that it is plainly impossible to .trigger() it. Maybe there's a way to bypass the radio buttons, and directly call some underlying function. — Alien333 (what I did & why I did it wrong) 19:25, 28 May 2024 (UTC)
@Alien333: What is it you're trying to do?In any case, those radio buttons are not plain HTML radio selects, they are an OOUI RadioSelectWidget. You're probably looking for something like
.trigger("click") isn't the same as interactively clicking the UI with a mouse, it's the event that gets triggered under the hood by that UI interaction. Sometimes these are the same thing, but often they are not. To actually simulate a mouse click you would have to use accessibility APIs and it'd be pretty hacky, probably.But if you're in edit mode the quality buttons are right next to the save button anyway. Are you trying to replace both perhaps? Xover (talk) 17:28, 30 May 2024 (UTC)
For save, there is already the alt-shift-S shortcut, that is sort of useless if you have to scroll anyway to change the quality, so I thought I might as well make another shortcut for that. — Alien333 (what I did & why I did it wrong) 17:33, 30 May 2024 (UTC)
You can probably just $("#wpSave").trigger("click") (not tested, may be more OOUI weirdness involved) and assign your own accesskey to kill both birds with one stone. Xover (talk) 18:24, 30 May 2024 (UTC)
Latest comment: 5 months ago3 comments3 people in discussion
So I am working on scanning a large U.S. government online document into a PDF. I currently have 47 pages. I know the Commons allows you to post a new version of a image/PDF. My question is would that screw up stuff over here on Wikisource? So say I uploaded the PDF as-is right now with 47 pages to work on transcribing those (basically 3 chapters done of it). If I then later go back and scan in more images, should I upload that as an entirely separate thing on the Commons or just upload a new version of the PDF? Basically, what should I do in this circumstance. Note, there is probably 200-ish (or more) pages of the document. WeatherWriter (talk) 20:36, 7 June 2024 (UTC)
It's best to transcribe from a complete PDF. Yes, altering a PDF during transcription can create problems. This is one reason we have a setting option on the Index page to indicate a scan needs to be repaired before proofreading. --EncycloPetey (talk) 20:47, 7 June 2024 (UTC)
Latest comment: 5 months ago3 comments2 people in discussion
Sort of silly question, but there, I can't manage to understand what the first letter of the title is, it's in a {{bl}} variant I'm not familiar with. (there is no TOC in this book, so can't use that). It's not a common word either, ending in "idigeigi". Google OCR says it's an H, but it doesn't look like that. A G, maybe? — Alien333 (what I did & why I did it wrong) 15:23, 10 June 2024 (UTC)
Latest comment: 4 months ago10 comments5 people in discussion
On this page, the text layer is offset (for some reason). However, when I tried to generate OCR for the page, it generated OCR from the same (wrong) page! TE(æ)A,ea. (talk) 23:43, 12 June 2024 (UTC)
I've created the text from what I see. If it is the correct page, then you may need to clear your browser cache to see the correct page. --EncycloPetey (talk) 23:58, 12 June 2024 (UTC)
EncycloPetey: What does this page look like to you? To me it looks like the start of the next section. If that is the case, then the text layer and OCR generation bases are both correct, but the images are offset. TE(æ)A,ea. (talk) 01:19, 13 June 2024 (UTC)
Something is very weird here. When the page first loads it shows one page scan, and then the page image reloads and shows a different page. The text layer loaded into the editor seems to be the one for the page that was briefly displayed, and so does not match the new page image. Since I've never visited this page before it can't be a browser cache issue. But it could still be a MediaWiki thumbnail cache issue, or that Proofread Page is getting stale text data from the API. @Sohom Datta: You may be interested in this issue. Xover (talk) 06:15, 13 June 2024 (UTC)
But that's not happening to me, so it's not occurring universally for everyone. FWIW, I'm running Firefox on a Mac OS. --EncycloPetey (talk) 07:25, 13 June 2024 (UTC)
I’ve tried from Edge and Google Chrome on two different computers—neither of which has a cache for the page—and both show the right OCR (p. 389) but the wrong page (p. 391, which is the beginning of the next section). This makes it rather difficult to proofread. TE(æ)A,ea. (talk) 15:02, 13 June 2024 (UTC)
@Ineuw: I don't understand the question. Why do you want to straighten this page? If you're asking how to straighten all pages in this scan then I advise against it: it's a lot of manual work, and the benefits are limited. Xover (talk) 06:28, 16 June 2024 (UTC)
I think it works better to purge it at commons, and also the ?action=purge only works (I think) if you're already in index.php, so it's simpler to use of of the gadgets that do that. — Alien333 (what I did & why I did it wrong) 07:12, 19 June 2024 (UTC)
Scan resolution (question for the technical people)
Latest comment: 4 months ago9 comments5 people in discussion
I'm getting frustrated with the poor quality of the scan image when proofreading A Dictionary of Hymnology. Have a look at Page:Dictionary of Hymnology 1908.pdf/44—the fine print is barely legible, even though I have increased the "Scan resolution in edit mode" to 2000. When viewing the PDF directly, the print is perfectly crisp.
I am guessing that the Wikimedia software takes the scan image at its default resolution, heavily JPG-compresses it, then increases the resolution of the compressed image, rather than scaling up before converting and compressing. This results in high-fidelity images of JPEG artefacts instead of actually usable scan images. I also have found a related task phab:T38597, to replace JPG with PNG in these images, which would presumably mitigate this issue—but this ticket is ten years old and hasn't been touched for years.
Anyway, my question is this: is there any way to improve the scan image inside ProofreadPage? Or do I just have to open the PDF in a separate window (which is what I have been doing)? —Beleg Tâl (talk) 18:06, 2 July 2024 (UTC)
Taking a quick look at the code, the PdfHandler extension generates jpgs which are then retrieved by us. Which jpg is retrieved might vary but it doesn't regenerate the images at a higher resolution if the original conversion is a poor representation. MarkLSteadman (talk) 21:31, 2 July 2024 (UTC)
Let's do a little math…The file as uploaded has 1796 pages and is 194.31 MB, which works out to about 110 kB per page. A modern smartphone photo averages about 6 MB, which means each page image here is somehow 1⁄55 smaller than the photos your iPhone makes. How is that possible given bulk book-scanning rigs are literally a DSLR mounted over a plate with some lights and other gizmos? Well, if you go look at the raw scan images at IA you'll find they add up to 2.7 GB, and even the cropped and colour-corrected images are 1.3 GB. But surely that's non-compressed images? Oh no, these are JPEG 2000 (.jp2) wavelet-compressed images (i.e. the 1.3 GB is already compressed size), which works out to about 760 kB per page. This means the PDF that IA produces takes already compressed images and then compresses them a further 7x.Then we get to the images on Commons. When a page image is requested, MediaWiki essentially uses Ghostscript to "print" that page out of the PDF and into a JPEG file. It does this by extracting the image data out of the glorified Postscript format (PDF is just PS with some sugar on top) into its own internal raster representation and then serializing that into the requested file format, in this case JPEG, including (lossy) compression. Proofread Page always requests thumbnails that are 1024 pixels wide (height is set automatically to preserve the original aspect ratio), which means that for page images that were originally less than 1024 pixels wide the extracted image is then decompressed, scaled up to 1024 pixels wide, and then recompressed before being sent to the web browser as a JPEG. Now the OpenSeadragon embedded in Proofread Page takes over and crams that image into the Page:-namespace viewer (OSD requests 1.5x, 2x, and 3x assets too, which complicates this a bit, but lets simplify for illustration purposes). This multiply-rescaled and recompressed image data is then what OSD and the web browser zooms in and out of and which you're trying to proofread from. That is, you're looking at an image that has been lossily recompressed at least 3 times and upscaled beyond what image data was there to begin with twice.So why is DjVu better? Well, the purely technical advantage isn't all that huge, but as it happens IA over-compresses their PDF files (they deliberately use very aggressive compression settings when making the PDF in order to achieve a small file size). When I make a DjVu file I grab the original scan images (the 1.3 GB zip), extract and convert the JPEG 2000 files to PPM (lossless), and then directly convert them to DjVu with moderate compression settings. That saves one recompression, and the compression settings are a lot less lossy. In addition, the DjVu compression algorithm (also a wavelet-based algorithm), designed specifically for scanned text (vs. JPEG that was designed for general photos), does a lot better at preserving original image data (it's a lot less lossy for this case). And finally, instead of the awful Ghostscript-based method for PDFs, MediaWiki uses the native DjVuLibre tools to extract a single page image, and it does a much better job at extracting the page image. MediaWiki (Thumbor) still rescales the resulting image based on Proofread Pages request, but since the starting image is of much higher quality with fewer compression artefacts, the resulting output is usually also much better. There are pathological cases where the result is bad, but these are extremely rare (usually from some random web service that converts the IA PDF into DjVu, achieving only making things worse).So… When I say I strongly recommend using DjVu whenever possible I really mean "Come on people, why would you ever use the IA PDF?!?! Get with the program and use DjVu because even if you have to bend over backwards and jump through hoops to get that DjVu it's still going to be better!" And it's why I have an open invitation to anyone to ask me to make DjVu files for them, that I try to prioritise as much as I can (which isn't very just now, but…). There are issues with lack of user-friendly end-user tools for DjVu (i.e. you can't view DjVu inline in web browsers any more), and there are big questions about the long-term viability of the format (there's no commercial backing and no significant community around it), but it is still a much much better choice than the current state of PDF and PDF tooling. Longer term (much longer) the new target is probably support for "Collections" on Commons so that we can upload the original JPEG 2000 scans (zero loss) but still get an atomic pseudo-"file" for Proofread Page to work on. But given the pace of development and lack of resources the WMF assigns both Commons and Wikisource this is still a long way in the future so we still need the lesser evil in the mean time. Xover (talk) 08:40, 3 July 2024 (UTC)
Thanks for the detailed info! I knew that IA highly compresses the PDF files, but since I am able to see the page clearly in a PDF viewer I would not have expected that to be the issue. In fact, Wikisource:DjVu vs. PDF claims that PDF has a higher resolution. Most discussions I have seen (here on enWS, and also on commons) seem to take the view that there is no longer any reason to use DJVU ...
Oh, and the "Scan resolution in edit mode" option in Index: pages… It's been a long time since I dug into what that actually did, so I'm very vague on the details, but as I recall its effect was essentially about how big to display the image in the web browser but the image generated was exactly the same. I.e. it's a kind of hard-to-use zoom that's been obsolete for years. I could be wrong, but my conclusion at the time was that the option was useless. Xover (talk) 08:45, 3 July 2024 (UTC)
Brackets for vocal + piano scores
Latest comment: 4 months ago3 comments2 people in discussion
I'm transcribing a score that follows the common pattern of one line of vocal music together with a treble and bass piano part, the piano parts marked with a curly bracket. At the moment, on the pages I've transcribed (2 and 3), the vocal line is also included in the bracket. Could someone fix this? —CalendulaAsteraceae (talk • contribs) 18:41, 4 July 2024 (UTC)
You've got all three Staffs inside the PianoStaff. You need to nest them like this:
Hi, this is sol-fa notation, which is used as an alternate way of representing the melody for those who don't read graphical music. d=doh or the tonic; r=re (supertonic); m=mi (mediant); &c. The lines and colons indicate how long to hold the note. There's currently no satisfactory way of representing this in Lilypond and I have quietly ignored such in the transcriptions I've done. Beeswaxcandle (talk) 02:43, 7 July 2024 (UTC)
Latest comment: 4 months ago3 comments2 people in discussion
I can't figure out how to address the following footnote variation in Index:The Remains of Hesiod the Ascraean, including the Shield of Hercules - Elton (1815).djvu. On page 129 there is a footnote beginning Tis time to sow. This footnote continues onto the following page 130, where there is a footnote within the footnote. I have dealt with such footnotes before where both the main footnote and the sub footnote are on the same page but I can't figure this out. I have tried using the new-ish <refn> but it doesn't seem to support the use of 'name' and 'follow', unlike <ref>, which are needed to cover the spread of the main note over two pages. Any suggestions? Chrisguise (talk) 21:51, 6 July 2024 (UTC)
@Chrisguise The documentation of refn may not look like it supports name and follow, but clicking edit on the template (without touching anything) seems to indicate that name and follow are parameters. I have attempted to use {{refn}} on said pages (and 131 for the additional follow on), and checking your transclusion, it looks like it is working. Thanks for all your efforts, and sorry for Wikisource's (lack of) documentation. Regards, TeysaKarlov (talk) 23:37, 6 July 2024 (UTC)
Thanks for that. I did try refn with 'name' and 'follow' and couldn't get it to work. I guess I must have made a mistake somewhere. Thanks again. Chrisguise (talk) 07:43, 7 July 2024 (UTC)
Latest comment: 4 months ago1 comment1 person in discussion
The first page of the Preface to the Second Edition (two pages below the title page) is missing in the index, but present in the Google Books scan[1]. Can someone who knows add it? I assume, also, that the pages following it in the Index should be moved +1, but I do not know how to do this. Mårtensås (talk) 22:36, 9 July 2024 (UTC)
How to deal with poetry.
Latest comment: 3 months ago4 comments2 people in discussion
@TE(æ)A,ea.: Yes, it does, thanks! I also used all "Purge", "Hard purge" and "Null edit" buttons before, but none of it helped. Did you do the same or did you do anything else? --Jan Kameníček (talk) 21:30, 13 July 2024 (UTC)
Jan Kameníček: I cleared the cache for the file on Commons, and then for the index here. It is usual for files to not render properly on Commons; a cache purge generally fixes the problem. TE(æ)A,ea. (talk) 22:09, 13 July 2024 (UTC)
User:ShakespeareFan00/SPC_Colors
Latest comment: 3 months ago9 comments3 people in discussion
So I was using some rather recent (as of 2023) CSS support, to make the color tables. However, doing this manually seems to be a waste of time. Is there someone versed in Lua that can come up with a way to generate this programitcally. Also having a Lua script that can convert Munsell Colors back to sRGB would be nice. ShakespeareFan00 (talk) 11:26, 14 July 2024 (UTC)
Recurse Steps 3-4, for Red Brown, Cyan, and Key if values present, using z2,z3,..zn x2,x3,...xn etc.
On the last pair of values scale them both by the xn value reached,
write
) <xn>... <x2>) <x1>) <x0>)
This should write a series of nested color-mix functions that mix a pusedo CYMK color. The same approach could be used to make an n-tuple color-mix function that could be invoked from elsewhere, like the Pages of "The Color Painter" for example. ShakespeareFan00 (talk) 22:50, 19 July 2024 (UTC)
(sigh) - Essentially, I was wanting to reconstruct the charts in the PDF linked. It gives various CYMK values (as percentages), and the base colors for the CYMK (also a Red Brown it uses additionaly.) as xyz or xyY values. The template I had written was a first attempt at doing the conversion. However I was subsequently advised (off-wiki) that I needed to scale values in the nested color-mix statements, which gets complicated in pure wikitext. I was therefore asking if someone was able to produce a Lua module to generate the nested color-mix functions, doing the scaling as values as needed. The list of steps above was my attempt to explain what the module needed to do. ShakespeareFan00 (talk) 07:38, 20 July 2024 (UTC)
You generally need to understand quite advanced colour theory in order to make actual conversions, which I most definitely don't. Your explanation above either presupposes a familiarity with colour theory, or it is not a coherent explanation. Either way I do not understand it. Xover (talk) 13:03, 20 July 2024 (UTC)
Xover All I can pull from my gray-matter memory is the need for Y' which is based on the color to be converted (XYZ to X'Y'Z') that can be used to determine the converted color. 'Luminance' belongs to this explanation, as I gleaned from https://python-colormath.readthedocs.io/en/latest/conversions.html A color module that would determine this magical Y' (the scalar ShakespeareFan00 mentioned and a term from linear algebra, as these are all matrices) would allow conversion for many color spaces, not just the Munsell. For additional mind-boggling, look into "out of gamut" (crossed-eyes and dotted Ts)--RaboKarbakian (talk) 15:37, 20 July 2024 (UTC)
I had a partial template already (User:ShakespeareFan00/SPC color but got stuck on trying to work out the respective ratios.. Hence my attempts at explaining it, I'd already converted the xyz information down into RGB colors.. What my template is trying to do figure out how to nest the color-mix functions I was advised off wiki that for each additional color in the mix I needed to rescale the contribution of colors within that contribution to the overall final result. For example:- 30% yellow., 50% magenta, 20% cyan
Starting with the yellow 'lay down';
Yellow is 30% of the mix so color-mix(insrgb,--yellow30%,color-mix(insrgb,<MCK>)(100-30%)
The magenta values then has to be scaled color-mix(insrgb,--yellow30%,color-mix(insrgb,--magenta,50%/(100-30%),<CK>100%-(50%/(100-30%)))(100%-30%)
That leaves the Cyan/ Key mix to consider color-mix(insrgb,--yellow30%,color-mix(insrgb,--magenta,50%/(100-30%),color-mix(insrgb,--cyan20%/(100%-(50%/(100%-30%))),--key0)100%-(50%/(100-30%)))(100-30%)
Trying to write calc functions to do that in wikitext would lead to unreadable code, hence the request for a module to do it. 4 colors mixed and it's already getting hard to read. color-mix() is the 'magic' that makes it possible. ShakespeareFan00 (talk) 16:09, 20 July 2024 (UTC)
And something is still not quite right, because the colors being generated don't match the relevant linked page in the scans.. Time to just give up and conclude "too difficult, because no-one has the expertise to figure out why the color mixing isn't working. " It doesn't help of course that Google Books seem to completely destroy any form of accurate color reproduction on some works (sigh) ShakespeareFan00 (talk) 17:17, 20 July 2024 (UTC)
By putting the \repeat volta 2 { } into every voice, one of them has got out of alignment with the others. I've found that you don't need to do this in all the voices. As long as one voice has all the repeats marked, the others will fall into line. I just notate straight through on the other voices. btw I find your very open style of Lilypond transcribing with minimal comments quite difficult to read, thus I can't tell quickly which voice has the problem. For scores like this, I usually use the Frescobaldi app (offsite) and then copy/paste the final Lilypond text across.
For your second question, unfortunately there is no way to do that through transclusion. I usually go for the pragmatic route of transcribing a score into a single page in the Page: namespace and leave transcriber comments on the other pages on what I've done. See Page:Fugue by Ebenezer Prout.djvu/235ff for an example. For a work that is entirely a score over multiple pages, I've done it directly into the Mainpage like you have with this score. See Essentials in Conducting/Appendix B for an example of that. Beeswaxcandle (talk) 09:44, 30 July 2024 (UTC)
I tried taking out the extra repeats in the piano voices, and this worked fine for the score, but it meant that none of the piano voices played in the repeat in the MIDI. The reason I use this open style is that I find the code easier to read if there are clear delineations between measures and between lines, but I'm happy to add comments! What kind of comments would you find helpful?
I usually put in comments like "end of line 1" or "end of page 3" into each voice, which helps me orientate to where I'm up to. For the Haydn score, I did the midi offsite and then linked a sound file via Commons rather than having it create the file on the fly when opening the page—but that's a very long score. Come to think of it I did have a go at transcluding one score in the orthodox manner Cox and Box (complete)/Rataplan. The jiggery-pokery in Lilypond to get the page continuations to behave was challenging. However, the sound file did have to be created separately and linked at the top. Beeswaxcandle (talk) 08:39, 31 July 2024 (UTC)
As it happens, the process of updating comments in the page namespace and then re-copying them over showed me that the repeats were misaligned because I'd just failed to copy over some of the score. Problem solved! —CalendulaAsteraceae (talk • contribs) 18:11, 31 July 2024 (UTC)
There's no proof of this working outside the template documentation. I would need to be convinced of its utility in real situations. Beeswaxcandle (talk) 08:40, 31 July 2024 (UTC)
Merging duplicated indexes
Latest comment: 3 months ago5 comments3 people in discussion
Hi y'all,
Thanks to Uzume (talk • contribs), we noticed that there is two index for the exact same version of text:
Could someone move the pages (without redirects) and merge the indexes? (afterwards, I could delete the first djvu on Commons and update the main page transclusion and the Wikidata item).
We can move the pages, but what do you mean about merging Indexes? We cannot merge the Indexes. Each Index is keyed to a specific scan file, and in this case the two files have completely different coverage. I also note that more than page moves are required because of current linking, transcription, and project templates. All of those things will also have to be fixed --EncycloPetey (talk) 20:37, 2 August 2024 (UTC)
Merging the Index pages is indeed probably not needed (and I meant w:Wikipedia:History merging), you can just delete it (maybe with redirect here, just in case?).
I deleted the first file on Commons and replaced it elsewhere (on Wikidata).
@VIGNERON: I appreciate you deleting the source DjVu file on Commons. I found another reason to delete that besides it being inferior and redundant. Apparently two of its pages (10 and 11) were swapped. I figured this out because (along with another issue), I found out Page:Shakespeare’s Plays, v.3 (playswithhislife03shakuoft).djvu/22 and Page:Shakespeare’s Plays, v.3 (playswithhislife03shakuoft).djvu/23 are swapped. So I went back to the website (the URL has changed but the content is basically the same) that is the origin of the DjVu and found the pages are also reversed there. Then I checked six other scans of the same volume (each from different original print sources) and in each case they are reversed compared to the website originating the now deleted source DjVu. Thank you, —Uzume (talk) 14:36, 5 August 2024 (UTC)
That happens because on Wikidata, that "main topic" of the song is the play. Our header template pulls that information and links to the corresponding WP article for that subject. If Wikipedia had an article about the Lehrer song, and if that were linked via the usual Wikidata interconnections, then it would point to the article about the song. But since there is a "main subject" for the work andno WP article about the work, the link goes instead to the article about the main subject. If someone creates a WP article about the song, and links it from the WD item, then the linkage would be to that article. In some instances, this subject linking works better, such as a history book about the US Civil War linking to the WP article about the US Civil War. In other situations, it can be confusing, such as in this instance.
Although there is no WP article, I have solved the issue somewhat by using a WD link via WP redirect to the article about the album on which the song appears. --EncycloPetey (talk) 19:19, 4 August 2024 (UTC)
Mass move of pages beginning with
Latest comment: 3 months ago21 comments4 people in discussion
I've split The Rigveda (a versions page for the ancient work itself) and The Hymns of the Rigveda (the translation by Ralph TH Griffith) which were originally one page. Now I need all pages starting "The Rig Veda/" moved to equivalents starting "The Hymns of the Rigveda/", for correct linking (see, for instance, The Rig Veda/Mandala 3/Hymn 1). Can someone who has a bot do this? It is far too tedious to do manually. Mårtensås (talk) 15:25, 3 August 2024 (UTC)
Before making any moves, the Contents listings for our copy need to be harmonized with the Contents listings in the scan. Our copy is divided into "Mandalas", but the published copy calls these "Books". It would be premature to start making mass moves without first determining what moves are required to match the scan. --EncycloPetey (talk) 15:42, 3 August 2024 (UTC)
The following change should at least be uncontroversial
all pages starting with The Rig Veda/ are moved to The Hymns of the Rigveda/
The edition is split into two volumes, each of which has its own indices of hymns and names. I am not sure if they can be merged. The numbering of the books of the RV is consistent and traditional, in the present ed. volume 1 contains books 1–6, volume 2 contains books 7–10. Mårtensås (talk) 17:22, 3 August 2024 (UTC)
@Mårtensås / @EncycloPetey: Can you confirm that this list look correct? Format is old page name / new page name on alternating lines (it's what the pwb move script expects, not necessarily the most logical way). I can run the bot over and move the pages, but I'd rather not move 1k+ pages to the wrong name. I can also run any replacements needed for links etc. due to the name changes, but I'll need instructions for what's needed (I know less than zero about this work, much less this edition of it). Xover (talk) 19:58, 3 August 2024 (UTC)
I note that our copy is the 2nd edition of this work, which implies that there is more than one edition, and therefore may need to be disambiguated in future. Given the large number of pages, using a preëemptively disambiguated page title is likely preferable over using one that may require an additional move in future. --EncycloPetey (talk) 20:20, 3 August 2024 (UTC)
Is it worth it here, though? I'm having trouble imagining someone will be adding additional editions of this text in any reasonable time frame. Xover (talk) 20:30, 3 August 2024 (UTC)
Trouble imagining someone will complete it, or trouble imagining someone will start it? It is starting it that will require disambiguation. --EncycloPetey (talk) 20:42, 3 August 2024 (UTC)
Here's what Griffith writes in the preface: "This second edition of my translation is in the main a reprint in compacter and cheaper form, with some corrections and other improvements in text and commentary, of the original four-volume edition."
In which case I do not see a point in worrying about the first edition. It can be marked separately in article space if anyone comes along and adds it. Further, this would not be the only instance on WS where a later edition is unmarked in article space.
Fair point. Though I have to admit I have trouble imagining someone will even start another edition any time soon. I'll also note that the number of pages doesn't matter that much here: the work is done by a bot so it's more the noise in recent changes and people's watchlists to worry about. For a relatively straightforward case like this I am not overly concerned with a hypothetical future need to move it again. It's not nothing, but it ain't a big deal either. Xover (talk) 06:45, 4 August 2024 (UTC)
Given the information provided by Mårtensås about the earlier edition, I agree. And it is probably worth putting that information onto the work's discussion page, so that any future editors will know that information as well. --EncycloPetey (talk) 19:11, 4 August 2024 (UTC)
The list looks good. Many links to the Rig Veda are probably referring to the ancient work (at Rigveda) rather than Griffith's translation specifically, so I wouldn't move links. Mårtensås (talk) 21:14, 3 August 2024 (UTC)
I meant internal links within the text. |next=/|previous= links, etc. When moving from an ambiguous name to a specific one you cannot generally automatically change incoming links because determining the intended target requires human judgement. Within the text the links all obviously refer to that text so updating the links is just a simple text replacement. Xover (talk) 06:49, 4 August 2024 (UTC)
Alright. The format of the "previous/next" is just [[../Hymn X]]. However each hymn has a header "section" in the format [[../|Mandala X]], where Mandala should be replaced with Book. Mårtensås (talk) 16:03, 4 August 2024 (UTC)
To clarify: (1) Will there be pages like [[../Book X]] with lists of the contained Hymns? If so, then I assume such pages will be converted to use {{AuxTOC}}? And (2) The question is also whether such pages should appear in the page linking sequence in the same order listed on the proposed list page. --EncycloPetey (talk) 19:10, 4 August 2024 (UTC)
Such pages already exist, e.g. The Rig Veda/Mandala 3. They should be converted, but this can be done manually since there are only ten of them. And it is a change independent of the move.
As for the move, considering no further objections have been raised, could @Xover please move the pages according to the list in your Sandbox? (I have made a minor change to it, by removing the superfluous word The before Hymns, so please use the latest version.) Mårtensås (talk) 17:34, 6 August 2024 (UTC)
The word "the" is present in the original title, but is simply an article (and is printed in much smaller size), and I figure it might make the work more difficult to find. The change was also done in order to agree with Hymns of the Atharva-Veda (the original printed title of which likewise has a "The", in the same format as the Hymns of the Rigveda). But I have no strong preference; if you find it so problematic then I can undo it. I just want the subpages moved. Mårtensås (talk) 20:10, 7 August 2024 (UTC)
Moves and cleanup should be Done now, and all pages live at The Hymns of the Rigveda.Since I agree with EncycloPetey and BD2412 on the title issue and you indicate you have no strong preference, I went ahead and used that title.Please feel free to hit me up if you need more mass replacements done (e.g. switching the poems to use {{ppoem}}). Xover (talk) 09:57, 8 August 2024 (UTC)
Gadget-RunningHeader
Latest comment: 3 months ago3 comments2 people in discussion
Thankyou @Alien333 - the symptoms I am experiencing point clearly to a diagnosis of total incompetence - I was expecting a button, but the help page actually talks of a link in the sidebar, which after a few hard purges is now... there! Apologies for wasting everyone's time... CharlesSpencer (talk) 09:30, 8 August 2024 (UTC)
Pagelist tag backend error
Latest comment: 3 months ago6 comments3 people in discussion
I'm trying to populate Index:The Bibelot (Volume 15).djvu, but I can't create a working pagelist. No matter what I try, whenever I click on "Preview pagelist," I get this message: "The following error was encountered while parsing the pagelist tag in the backend, Error: Invalid interval." This doesn't happen when I click on "Preview pagelist" on works with existing pagelists; the tag functions normally for them. I've copied the pagelist tags from these other works into the pagelist field at issue to see if the error was caused by a typo in the tag, but the error still occurs even where the text of the tag is known to be correct. Am I doing something wrong, or is this a malfunction on Wikisource's end? Eureiachthon (talk) 23:19, 6 August 2024 (UTC)
From what I can see (and I'm not an expert), there is an issue with the DjVu file itself. It's showing content that it should not contain, and the file may have to be re-processed by someone who know how to fix it. --EncycloPetey (talk) 23:32, 6 August 2024 (UTC)
Thanks for the explanation. What do you mean by "content that it should not contain," and where should I look for someone to fix the file--here or on Commons? Eureiachthon (talk) 02:46, 7 August 2024 (UTC)
When I look at the file, the default image is the test scan of a black metal frame, and not the front cover of the book that was scanned. The test image should not appear anywhere in the file. There is a Wikisource:Scan Lab where repairs can be requested, but I would recommend waiting here until a couple of technically proficient file editors have had a look. The fix may be simpler than having to actually edit the file. --EncycloPetey (talk) 02:55, 7 August 2024 (UTC)
Long story short there's an ongoing problem with the communication between Commons and Wikisource that makes Wikisource think the file is 0 pages long, and hence any interval you try to use in <pages /> will be "invalid". I have made the requisite voodoo incantations to get around it and the file should work as intended now. I also regenerated it from the source scans to get rid of the test pages at the start and end. Xover (talk) 05:39, 7 August 2024 (UTC)
Latest comment: 3 months ago3 comments2 people in discussion
I can't figure out how to get the alternative score to show up only when it has notes in it. Using \startStaff and \stopStaff partially works, but there are still some lingering staff elements before the \startStaff. I don't want to add
\context{\Staff\RemoveEmptyStaves}
because this will cause problems elsewhere in the score, where there are empty staves that are supposed to be displayed. Help?
This is what's known as an Ossia staff. The second technique described in the Lilypond Notation Reference in section 1.6.2 is the most suitable for what you want to do. It's also got how to make the ossia smaller. For the bracket, this is described in section 1.6.1 under Grouping staves (the first of the selected snippets). However, as doing it will cause problems with the Staff Group already in use, I suggest that you don't reproduce it—particularly if it ends up not starting at a new line of music. Beeswaxcandle (talk) 06:33, 7 August 2024 (UTC)
Latest comment: 3 months ago6 comments2 people in discussion
I created a new document for the first time, but misspelled it.🤦 I'm assuming the reason I can't move it is because I'm not autoconfirmed on Wikisource. Can somebody move it for me (and make sure I otherwise made the page correctly)? Nicknimh (talk) 18:34, 9 August 2024 (UTC)
Done. However – what is your source for this section? Ideally new entries to the Jewish Encyclopedia should be coming from the Index pages, like Index:Jewish Encyclopedia Volume 2.pdf. Joses Barnabas is on page 587. Do you need help figuring out transclusion? Have you looked at Help:Beginner's guide to proofreading, which is very useful? There's no rush, but it'd be good if this section was scan-backed. (That is, transcribed from a scanned version of the book and connected to an Index page). Cremastra (talk) 18:39, 9 August 2024 (UTC)
I created the page by using an existing one (this one, selected at random) as a template. However, I suppose that that must be a really old page from before Wikisource started using this scan & transclusion system. There is a scan, but the autoscan is really bad and seeing as I already written out a page in wikitext, it seemed easier to just post it in directly. Would it be better if I edited the relevant scanned pages and then replaced my wikitext with a transclusion? Rather than doing that, is there a way I can just keep the existing page and link to the scanned PDF so that someone can easily verify it's correct? Nicknimh (talk) 18:47, 9 August 2024 (UTC)
"Rather than doing that, is there a way I can just keep the existing page and link to the scanned PDF so that someone can easily verify it's correct?"
That is a little unorthodox, and while it might be a good interim step, I wouldn't really advise it. Scan transcription is really the preferred method here. If the automatic transcription isn't working, it's just going to be a bit of a pain. I'm not sure why it isn't; it seems to have just frozen for me.
But there's good news!
This is a collaborative project, so people can help out. You don't have to do all of it; if you want you can just proofread the section that you want. Cremastra (talk) 18:58, 9 August 2024 (UTC)
poem tag formatting carrying into included footnotes
Latest comment: 2 months ago4 comments2 people in discussion
I can, of course, circumvent this behavior by simply not using the tag, but want to determine whether this behavior can (or should) be corrected for future editors. --EncycloPetey (talk) 21:04, 4 August 2024 (UTC)
@EncycloPetey: You mean the line-breaking behaviour? Yeah, I'm guessing that's due to the interaction between the <ref>...</ref> and <poem>...</poem> tags; the poem extension is either inserting <br /> before handing the ref tag's contents off to the cite extension or applying itself to the cite extension's output. I had never noticed this behaviour before so it could well be a recently introduced bug (or at least change in behaviour), but it may also just be that since they changed the line height and other vertical whitespace recently this issue became a lot more noticeable. The observable behaviour appears completely logical from a software point of view (even if obviously non-intuitive from a user point of view).This kind of problem is specific to the interaction of MediaWiki extension tags, so any other way to format the poem should avoid it (replacing the <ref>...</ref> tag's behaviour is a lot harder). Xover (talk) 06:18, 13 August 2024 (UTC)
I worry, though that this is an error, and thus might be "fixed" later, in which case the currently applied formatting will be broken when the interactive behavior is fixed.
Is there a way to find out whether this is (a) deliberate, or (b) accidental, but not going to be changed, or (c) unintended, and going to be fixed? --EncycloPetey (talk) 07:55, 13 August 2024 (UTC)
Hmm. Surely I've used footnotes inside poems before, including before {{ppoem}} was an option, and this behaviour should, I think, have been fairly obvious. Maybe we could try to look for older texts that uses the same construct (<ref>...</ref> inside <poem>...</poem>) that are broken now (but, presumably, were not broken at the time)?To figure out whether there's a software bug when we're 1) not sure whether there even was a change, 2) when that hypothetical change happened, or 3) which particular component any such change was in, is pretty needle-in-haystack. We'll need to narrow down the parameters pretty significantly to stand a chance there. Xover (talk) 09:11, 13 August 2024 (UTC)
Help with weird no-lined table formatting
Latest comment: 2 months ago4 comments4 people in discussion
@WeatherWriter: To remove the lines, remove the class="wikitable" in the first line. For other styling, most things you could want can be found in {{ts}}. About the TN (also goes for the 1998 on the left, and the rest): I am pretty sure these were just cases of text getting wrapped, so it really was a single line. I did that first line on that page to show you. — Alien333 (what I did & why I did it wrong) 17:44, 14 August 2024 (UTC)
I've set all the table on that page in a different way to Alien333. Choose whichever makes sense to you and continue in other places. Beeswaxcandle (talk) 17:51, 14 August 2024 (UTC)
For multi-page tables like this, I like to use index CSS, because that way I can just set up the formatting once and not have to worry about it afterwards (plus it's easier to change later if needed). —CalendulaAsteraceae (talk • contribs) 15:55, 15 August 2024 (UTC)
Inline image
Latest comment: 2 months ago4 comments2 people in discussion
Latest comment: 2 months ago4 comments3 people in discussion
Hi, I have a pretty basic question. When proofreading, should I stick to what literally is on the scanned page including mistakes and old grammar, or should I update the text to the newest conpemporary grammar? All information around here makes me believe the former is desired, but there are some books that have been updated to the new, current grammar. Should those be further "updated" to the original text? Wuopisht (talk) 12:05, 19 August 2024 (UTC)
Grammar and spelling shouldn't be changed with the exception that very minor and obvious typos can be fixed (such as "thanks, buddy" for "thanks,buddy") and {{sic}} or {{SIC}} can be inserted for misspellings that are retained in the text. If there are newer editions of a work and those have revised grammar and spelling, then they can be transcribed separately with their new changes. Ultimately, there will be some kind of edge cases or exceptions for virtually any rule, but the bias should be toward retaining the original text as much as you can. If you have to explain a lot about the text to make it intelligible, we do have annotated editions of some texts and you could use that format as a way to give instructions or editorial clarifications. —Justin (koavf)❤T☮C☺M☯12:17, 19 August 2024 (UTC)
I would also say that "very minor and obvious typos" should be very extremely minor. An obvious spelling error like "speling" is not minor and should be preserved (though you can use {{SIC}} for this, so long as it's actually a typo rather than a work that predates standard spelling). An error like a missing period, which is as likely to be an error with the scan file as it is to be a typo in the publication, is borderline IMO. Spacing around punctuation, which could be thought of as more of a "standardization" than a "correction", is fine (and see WS:MOS for more examples of standardization that we are ok with). —Beleg Tâl (talk) 13:15, 19 August 2024 (UTC)
If it helps, there are also the <math></math> tags, which might provide both a simpler syntax, and subjectively, an improved rendering of the equations. For example, {{math|1 - y<sup>2</sup>}} yields 1 - y2, compared to <math>1-y^2</math> which yields (where you no longer need spaces around operators, if you want them, and the negative sign is no longer diminutive). Ultimately up to you though, but if you do use <math></math>, and are looking for examples to work from, you can always try many of the pages in Calculus Made Easy, which came through the MC a while back, and was the work of an unregistered user.
Hi TeysaKarlov, I'll probably use whatever makes it look closer the original. I currently am stuck on this page with a symbol I don't know. It looks like a t, but no idea what style it is. Nobody (talk) 05:06, 21 August 2024 (UTC)
That is τ (tau), one of the Greek letters shamelessly stolen by mathematicians and physicists. To render, you can find it in the insert menu among the other Greek there, or it can be represented as I did here with τ or <math>\tau</math>.--RaboKarbakian (talk) 13:08, 21 August 2024 (UTC)
Would it be a good idea to create a decent quality DjVu file, replacing the overcompressed PDF? Also what's the proper naming convention for the index pages? --Ssvb (talk) 13:08, 23 August 2024 (UTC)
(I'm not aware of any naming conventions for indexes)
Thanks! From the purely technical side, I know how to create a DjVu file. As for the name of the index, I wondered if any kind of common unofficial conventions exist for naming indexes of translated books. Should it include the year of the publication and/or the name(s) of the translator(s) for disambiguation purposes? The Wikipedia article From the Earth to the Moon says that this particular translation was done in 1873, but the title page has 1874 printed on it and there was another edition from a different translator also published in 1874.
I would like to also experiment with adding a text layer to the DjVu file. So that it's conforming to the English Wikisource formatting rules (with words hyphenation automatically removed and the {{nop}} templates added where appropriate). --Ssvb (talk) 14:30, 23 August 2024 (UTC)
When you say removing hyphenation, what hyphenation are you talking about? on end of page? on end of lines?
Also, there isn't really a way to automatically add nop's where appropriate. What you could try is checking if the first line of the next page has an offset, but you need for that to differentiate the header from the body, to take care of images and OCR is not a perfect science so there's always a risk of a spot on the page adding an O at the wrong place, etc. In any case, you'll have to manually review each case, so there's not much point. — Alien333 ( what I did why I did it wrong ) 14:40, 23 August 2024 (UTC)
You can remove end-of-line hyphenation to a reasonable approximation, but you'll probably only get about 80% because the rest are properly hyphenated words and you can't detect these without functionally infinite word lists (and books use inconsistent hyphenation rules anyway). This may not be optimal because many contributors prefer to preserve the original lineation when proofreading (to keep track of where they are) and only remove hard line breaks at the end. I have a JavaScript that does a simple regex-based hyphen removal that's "good enough".Detecting when to add a {{nop}} is possible, but it probably requires machine learning ("AI") in practice. I don't think the effort will be worth the investment to make it, unless there's a ready-made tool for it out there that I'm not aware of. Xover (talk) 15:04, 23 August 2024 (UTC)
Machine learning always has an error rate, that is likely larger than that of a human proofreader, given that printing is often inconsistent, and you can't train a neural network for every type of printing/OCR error. — Alien333 ( what I did why I did it wrong ) 15:08, 23 August 2024 (UTC)
@Xover: It doesn't need to be perfect, because proofreading is still needed anyway. The goal is to reduce the amount of time spent by a proofreader on editing text. I found end-of-line hyphenation correction to be one of the most annoying activities during proofreading. I also have a JavaScript myself, which removes end-of-line hyphenation while preserving the original lineation. If a proofreader removes hard line breaks at the end, then the text will be only harder to review for a validator.
Adding {{nop}} markers should be technically possible for a DjVu creator tool, because such tool can look one page ahead and use this additional information for its decision making. And it doesn't need to always guess everything perfectly. Though you may be right and things like running headers may make it harder to do or less reliable. --Ssvb (talk) 16:39, 23 August 2024 (UTC)
@Ssvb: Index pages must have the same name as the File:, because that's how Proofread Page connects the two. Files, though, don't really have hard rules beyond being descriptive. I recommend the schema Title (year).djvu (e.g. Sixes and Sevens (1911).djvu) because it is sufficiently unique to avoid collisions, descriptive, and comports with common bibliographic conventions. The year is the year of publication of the edition of which the scan represents a copy (so probably 1874 for this text). If further disambiguation is needed (relatively rare) one can add, for example, the publisher in the parenthesis (e.g. Sixes and Sevens (Doubleday, 1911).djvu). Others prefer other schemas and there is no hard rule for it, so this is just my recommendation. Xover (talk) 14:58, 23 August 2024 (UTC)
Oh, and "Index:From the Earth to the Moon direct in ninety-seven hours and twenty minutes, and a trip round it (IA fromearthtomoond00vern).pdf" was automatically generated by a bot doing a bulk import of the Internet Archive. It is not a good practice for naming files. It just grabs the title given at IA (very arbitrarily set from some library catalog somewhere) and tacks on the IA identifier. It was made by one person and was interested only in bulk-importing. Please don't replicate this file naming schema. "From the Earth to the Moon" is the sensible title to apply here. Xover (talk) 15:08, 23 August 2024 (UTC)
This perfectly answers all my questions on this matter. Thanks a lot for providing such detailed explanations. --Ssvb (talk) 15:44, 23 August 2024 (UTC)
You can redirect a `/styles.css` to another CSS page (for example if a set of volumes share the same styles) but the redirect page may need to have the "content model" changed to "wikitext" (from "sanitized-css"), which currently requires an admin.
Despite the annoyance of having to ask an admin, it is generally preferable to use redirects for this purpose. @import has security implications that may conceivably lead to limiting that use of it in the future, and with redirects we can use the normal on-wiki tools for this (e.g. Special:WhatLinksHere) that do not work with @import.PS. cf. Class naming conventions, use the _ prefix for class names in per-work styles to avoid collisions with classes from other sources. Xover (talk) 07:25, 8 September 2024 (UTC)
I created a redirect at vol. 2, but not the others since they do not exist yet. Please feel free to grab me, or post at WS:AN, when you create the index for subsequent volumes. Xover (talk) 07:29, 8 September 2024 (UTC)
Hmm. It'd probably also work to create the stylesheet in one of the subordinate indexes, then move it by turns through all of them until it finally ends up in volume 1 / its permanent location, and then manually update all the redirects thus created. But those are all hyper-complicated technical approaches that we can't subject our contributors in general to. Thus the guidance to just request it at WS:AN and let an admin sort it out. Xover (talk) 10:21, 8 September 2024 (UTC)
{{REDIRECT|Index:title/styles.css}} would do the same thing as moving the page would. Hmm. Maybe three curly brackets on each side. But "Move" just deposits one of these wiki directives and displays a suggestion for how to handle things after the move which should be ignored for this case.--RaboKarbakian (talk) 13:54, 9 September 2024 (UTC)
Page links not displayed in Firefox 130.0
Latest comment: 1 month ago2 comments1 person in discussion
I have 'page links displayed' set on and 'page links beside text'. In Firefox 130.0 (64 bit) they are no longer displayed. I've checked Edge (128.0.2739.67 (Official build)) and they display and work fine there. Chrisguise (talk) 12:36, 12 September 2024 (UTC)
Apologies - they seem to have started working again.
Latest comment: 1 month ago11 comments4 people in discussion
We have three author pages for people named Duncan Campbell, and very limited information about any of them. If anyone would like to assist with researching biographical information for them, it would be appreciated.
Correct, but only if the fl. date is correct, and it may not be. That date assumes the one publication we have was not published posthumously, or that it was not published under the name solely to capitalize on someone's fame. Hence, I am not sure whether it is the same person or not. --EncycloPetey (talk) 16:15, 12 September 2024 (UTC)
Huh, interesting. He seems to think that Campbell (d. 1730) and Campbell (fl. 1756) are the same person, as EP suggested above. I'm skeptical, but maybe I'll take a page from Wikipedia and aim for verifiability rather than truth (shrug) —Beleg Tâl (talk) 19:45, 12 September 2024 (UTC)
Actually, no—The Earth's Groans, &c. describes the 1750 London earthquakes, and was available for sale by the author in 1756, so I'm not accepting that it was written by a guy who died in 1730 regardless of what Allibone says. —Beleg Tâl (talk) 23:20, 12 September 2024 (UTC)
Latest comment: 1 month ago4 comments3 people in discussion
There are many misspellings in the pdf format of this page one is on page 176 item 63 and it's written Lime and I believe it should be time. 24.189.233.20523:42, 15 September 2024 (UTC)
To clarify, I am trying to figure out why the empty line (\n\n) is not being parsed as a paragraph break. As you say, I could put a <br> for a line break, but that doesn't answer the question of why the double line isn't being parsed as expected. If I use {{lang block/s}} and {{lang block/e}} instead of {{lang block}}, it works as expected. —Beleg Tâl (talk) 17:30, 17 September 2024 (UTC)
Near as I can tell, this is a product of how Module:Lang has the feature local inline = yesno(args.inline or "yes") and the default value is no to strip out or suppress certain elements. Compare:
Huh, that seems to be it, though I find it very unintuitive that the feature allowing it to behave as expected for a block element is to specify inline=yes lol. Thanks! —Beleg Tâl (talk) 17:48, 17 September 2024 (UTC)
This is a bug in the way the template is coded. I recently fixed {{asc block}}, which had the same issue. Templates of the form
By adding in the "\n" for both DIV and SPAN varaints of lang, you put whitespace inside a SPAN which the parser really really doesn't like. You should be checking for the block variant of the template/module call and only applying the "\n" on the block variant. ShakespeareFan00 (talk) 18:38, 17 September 2024 (UTC)
A side note, but to avoid future reckless editing, maybe also request protection of module? Looks high-traffic enough to me. — Alien 3 3 318:54, 17 September 2024 (UTC)
Mbee-wiki: You can use the “Upload a new version of this file” button on Wikimedia Commons. I have already done this. The new scan had a few extra pages at the front and end scanned (blank pages), so the pages which have been created are moved a bit. These can be moved by administrators if you make a request at the Administrators’ noticeboard. TE(æ)A,ea. (talk) 13:16, 16 September 2024 (UTC)
@TE(æ)A,ea.: Hello. I saw that you already re-uploaded the better source file to Commons and fix the index. Thank you very much for your help 🙏🏼 I think I will just copy-paste the page I have proofread as it's only one page. Mbee-wiki (talk) 06:36, 17 September 2024 (UTC)
Mbee-wiki: IA no longer automatically makes DjVu files from uploaded PDF files. To get a DjVu file you can use an on-line converter or you can make a request at the Scan Lab. (Sorry for bouncing you around between so many different places.) For this, of course, you would need to upload a new file and make a new Index (DjVu instead of the current PDF). DjVu files have a lot of advantages in terms of readability and accessibility on Wikisource, so that will be a good change. TE(æ)A,ea. (talk) 13:50, 17 September 2024 (UTC)
Thank you. Would you mind to suggest a trustworthy online converter for PDF to DjVu? I would like to try make a DjVu file by myself, but some sites I found are too fishy... Mbee-wiki (talk) 03:25, 18 September 2024 (UTC)
Latest comment: 1 month ago5 comments2 people in discussion
My index page status colors of red and yellow changed to beige and pink, and I have no idea what I touched/changed to cause this. This is on en.wikisource but is normal on fr.wikisource. — ineuw (talk) 04:44, 19 September 2024 (UTC)
Latest comment: 1 month ago4 comments2 people in discussion
Hello,
Three indices I have somewhat recently uploaded to commons are showing up as 0 x 0 pages and I am unsure why, as others which I have uploaded in an identical fashion are fine. For example, File:In the Reign of Coyote.djvu, File:Folk Tales of Beasts and Men.djvu and File:Home-Made Toys for Boys and Girls (Hall).djvu, noting that the former two were the raw djvu file from Internet Archive, and the latter, a pdf2djvu conversion, if that matters. While it would be nice if someone fixes the issue(s), I would much appreciate if someone could also explain exactly how to fix the issue(s), in case it crops up again. And sorry if this question has been asked before and I just can't find a satisfactory answer, i.e. if it is just purging files, how and what exactly?
This is pretty normal with files that have recently been uploaded to Commons. If you have made an Index already, On the Index page, see the top right corner where there is a looping arrow? Hit that and try to purge the file once or twice and that should fix it.If you haven't made an index, trying to purge the page at Commons may fix it or just waiting for a while usually does the trick. The simple "soft purge" method is just Ctrl+Shift+R and that sometimes works. There are also hard purges and null edits. Another purge method is adding ?action=purge to the end of the URI. Sometimes, if you do that once or twice, it will work. In fact, that's what I just did and it worked on my end. —Justin (koavf)❤T☮C☺M☯20:35, 25 September 2024 (UTC)
@Koavf Thanks for the quick and detailed response (e.g. I has seen ?action=purge in comments before, but did not know what to do with it). The indices are now showing up as having pages. Thanks again, TeysaKarlov (talk) 20:47, 25 September 2024 (UTC)
Latest comment: 1 month ago2 comments2 people in discussion
I spent a long long time on the internet exerting all my powers of Google-Fu to chase down the English text of Lord Ellinborough's much-ridiculed Somnauth Proclamation, eventually finding it in John William Kaye's History of the War in Afghanistan (1842). Is it OK to extract the Proclamation and put it as a stand-alone document in Wikisource rather than the entire book?
Latest comment: 1 month ago3 comments2 people in discussion
I'm working on Index:Moralreflection00stangoog.djvu and am not sure if I'm doing the footnotes correctly in the first place to be honest. But there are also some pages where the footnote continues on the next page in the footer section and I'm not sure how to handle that.
I don't know if its too much to ask as well if someone could just give a once-over on my work and see if there is anything I'm doing incorrectly. That would be appreciated. Odhar (talk) 00:49, 1 October 2024 (UTC)
Latest comment: 1 month ago5 comments4 people in discussion
It has happened to me, sometimes, that files generated with the same method as usual have their OCR shifted one or more pages towards the start, which is quite annoying. Does someone know why it happens, and, if so, how to prevent it? Thanks, — Alien 3 3 319:59, 1 October 2024 (UTC)
Alien 333: There is a bug report at Phabricator about this. It is years old. I blamed that choice to remove the first page, but the consensus was no on that. I was given another explanation that was something about the title page causing things to back up. My joining the bug reporters has not helped to fix it, that said, if you would like to join I will dig out the report number for you.--RaboKarbakian (talk) 00:59, 2 October 2024 (UTC)
T194861 is the most common cause, where images from Internet Archive are offset. The issue is that the image sequence becomes out of sequence with the xml, often because there are additional images such as the color targets, before the main sequence. MarkLSteadman (talk) 02:48, 2 October 2024 (UTC)
But some of these files, I built back up myself from the raw JP2s (e.g. File:Poems Nealds.djvu), so I didn't use any external metadata, and I'd expect OCR engines to not depend on the metadata that was in the image files that were used to make a document. — Alien 3 3 305:36, 2 October 2024 (UTC)
Background color - Is that important?
Latest comment: 1 month ago3 comments3 people in discussion
From what I see, it does exist there now and it seems like it's a good typographic approximation of the original. If you think that the header isn't really part of the document as such, but is basically metadata, then you can put it in the header section of the page and it won't be transcluded in the main namespace. —Justin (koavf)❤T☮C☺M☯16:33, 2 October 2024 (UTC)
I wouldn't call it the most important thing, but it's not extremely complex to do. Did it for that page, though the colors are approximative. From what you put there, I think you may find the templates {{rule}}, {{font size}}, {{float right}}, and various things listed at Help:Font size templates interesting. — Alien 3 3 316:33, 2 October 2024 (UTC)
PDF images not loading
Latest comment: 1 month ago14 comments5 people in discussion
Do you mean the images aren't loading on each page? I see that too, and it happened to me at multilingual WS as well.
If I click the link to the image, I get an error "Too Many Requests" which sounds deeply suboptimal. It might be something wrong on the commons end, since c:File:The Dream-Quest of Unknown Kadath.pdf won't load properly either (although the PDF itself still seems to be accessible.)
A side note, but that index appears to be an extract, as it's from a collection and the page numbering starts at 76. — Alien 3 3 315:02, 3 October 2024 (UTC)
Strange indeed: for me it's the first two that work and the third that is broken. Probably related to the WM technical infrastructure (using different servers or whatever). — Alien 3 3 306:42, 5 October 2024 (UTC)
Latest comment: 1 month ago12 comments4 people in discussion
Factotum initials are a kind of cheap drop cap which were popular in the Early Modern period. Instead of using an ornamental drop cap (see left), there was an ornamental frame into which the printer could drop any ordinary capital letter (see right).
Template:Drop initial allows for the case of ornamental drop caps with its image parameter. But there's no way to produce the same effect for factotum initials, which has led to Commons categories with dozens of images identical to their neighbour except for the letter in the middle.
So, I would like to extend Template:Drop initial to add a factotum switch, where the image given will be displayed with the 1 letter centred on top of it. But that template is complicated beyond my confidence, so I'm asking for help here. Marnanel (talk) 14:41, 7 October 2024 (UTC)
It's already possible to do that using the existing template, so long as you have uploaded an image to Commons for the initial letter. --EncycloPetey (talk) 18:33, 7 October 2024 (UTC)
Right, but I want to use an image of an empty factotum initial, like the one on the right here, and add the letter in the middle as an actual letter (rather as the type letter would have been inserted into the original). So it would be a sort of combination of the current modes with and without the image.
Otherwise you end up having to upload images for each of the original letters used, which leads to pointless duplication like this:
F
H
C
W
which are all clearly the same factotum with a different initial letter inserted! What we have to preserve here is the factotum itself. Preserving the original initial letter as part of the image makes no more sense than preserving any of the rest of the document as an image: it's the very reverse of what Wikisource is here for. Marnanel (talk) 21:28, 7 October 2024 (UTC)
But you will not be able to control the font style or font size, and therefore cannot reliably place the letter into the image. --EncycloPetey (talk) 21:34, 7 October 2024 (UTC)
Also, there are small differences between the frames in the examples you've given, and not just intensity of shading, but outlines of the leaves themselves. See for example the leaf pair at the frame edge just above nine o'clock. The original is not reusing the same floral frame. --EncycloPetey (talk) 21:38, 7 October 2024 (UTC)
The "alt=" property on a image allows you to provide the letter, and readers who have images turned off will then see the letter, so you're objection indicates you did not investigate the option I mentioned above. This property is already built into the template, and there are examples in the documentation. --EncycloPetey (talk) 21:40, 7 October 2024 (UTC)
EncycloPetey: “This property is already built into the template, and there are examples in the documentation.” This is obviously not true. The “alt” property is a technical measure which is present in all images, and bears no relation to the factotum image issue presented. The style/size issue is not a problem, as {{overfloat image}} is a commonly used template. I also don’t think that these are different images, but rather different imprints of the same image; in which case, we should isolate the best example and keep that image. TE(æ)A,ea. (talk) 23:39, 7 October 2024 (UTC)
Overfloat image is fussy and tempermental. Even the example in the template's own documentation does not have the text entirely within the blank space when I view it, though at one time it did. And the template certainly will not play well with the drop initial template. --EncycloPetey (talk) 00:59, 8 October 2024 (UTC)
And these are not all imprints of the same image. The H and C images are clearly different in multiple ways. The images are all very similar, but they were not printed from the same original. --EncycloPetey (talk) 01:00, 8 October 2024 (UTC)
I know how printing works, and I also know what kind of wear happens. In this instance, the outlines of the leaves along all four edges are different between the H and C blocks. Look, for example at the lower left flower, which is flanked by buds in the C image, but is not so flanked in the H image. They were not printed from the same block. --EncycloPetey (talk) 19:38, 8 October 2024 (UTC)
H was definitely from a different block from the rest. F and C are in my opinion probably the same block. W is either very degraded or just badly inked, but it is at least very similar to the F/C block. Cremastra (talk) 23:48, 9 October 2024 (UTC)
Done You can tinker with some sizing and spacing, etc., but this is at least serviceable, I reckon. —Justin (koavf)❤T☮C☺M☯ 21:51, 1 January 2024 (UTC)
_-_recurring_music_sheet.png)
Support Protection. ShakespeareFan00 (talk) 18:55, 17 September 2024 (UTC)
resolved at task T372470. Cremastra (talk) 17:09, 5 October 2024 (UTC)
F
H
C
W
_of_Panama_page_153.jpg&action=edit&redlink=1){kind=link}
![[2]](https://upload.wikimedia.org/wikipedia/commons/thumb/c/c9/Travels_in_Europe_and_Northern_Africa._A_woman%27s_view_%28IA_travelsineuropen00rose%29.pdf/page11-789px-Travels_in_Europe_and_Northern_Africa._A_woman%27s_view_%28IA_travelsineuropen00rose%29.pdf.jpg){kind=link}