User talk:CalendulaAsteraceae/Archive 8

Please do not post any new comments on this page. This is a discussion archive first created on 28 January 2024, although the comments contained were likely posted before and after this date. See current discussion.

Cleanup...(Part 2)

I recentyl finished cleaning out the 'missing-end' tage linter of block level tags as far as I was able.

Any chance you could add to your list of ongoing projects, cleanup of ns0 works using raw HTML elements that are not recomended?

Like use of the HTML tags BLOCKQUOTE, FONT and CENTER directly.

Many older direct HTML usages in ns0 have templated equivalents now :)

See also:-

• https://public.paws.wmcloud.org/4407/blockquote.txt
• https://public.paws.wmcloud.org/4407/rawcenter.txt
• https://public.paws.wmcloud.org/4407/rawfont.txt

Can you think of other 'clunky' direct HTML coding that a regexp can find, and replace with a (semanticly meaningful) templated alterantive? ShakespeareFan00 (talk) 13:38, 23 January 2022 (UTC)

@ShakespeareFan00: One thing that comes to mind is <p> tags. I don't know how easy they'd be to replace automatically, but they're definitely not necessary, and can be found pretty easily. Do you have a list of non-recommended HTML coding? —CalendulaAsteraceae (talk • contribs) 19:16, 30 January 2022 (UTC)

There are places where <p> tags are preferred, such as in headers where a paragraph break is needed in text set as a parameter value. Without the tag, the text overlaps the bottom margin. --EncycloPetey (talk) 19:22, 30 January 2022 (UTC)

Good to know; thank you! —CalendulaAsteraceae (talk • contribs) 19:23, 30 January 2022 (UTC)

Subsequent to the above the links may have changed.

In addition you might want to consider if the wikitext layer for Syntaxhighlight or {{HTML}} is a good fit for some <tt> and <code> and <nowiki>...</nowiki> styled source wikitext in documentation. ShakespeareFan00 (talk) 07:18, 14 May 2023 (UTC)

Thank you..

For resolving some of those long-standing LintErrors.

You'll be working through the 2,000 or stripped tags in Page:Namespace? (They don't transfer over into main namespace suggesting they are mostly unpaired templates, or header, footer content related?.

The strategy I had at one point was to resolve all the mis-nesting, then the stripped tags, and then the missing ones. When all the tags were matched up, it would be a lot more straightforawrd to resolve all the obselete formatting, given that a large proportion of that task could be automated.

You might find some of my Template tweaks useful in resolving even more LintErorrs and formatting glitches:-

Such as:- https://en.wikisource.org/w/index.php?title=Template:Span_tag&diff=prev&oldid=12337463 use the text parameter with the wikitext you want to place inside the span. An example would be

{{span|dpb|ac|text=Centered Text}}

can be used as a replacement for {{center}} and {{c}} where the content is inside another span (such as a {{float right}} or {{float left}} or various sidenotes. If inside a ref tag, remember to use {{blockref}} tag at the start of the REF tag, so that when certain limitations in Mediawiki are removed the relevant markup can be changed back.

https://en.wikisource.org/w/index.php?title=Template%3AHeading&type=revision&diff=12307430&oldid=11131499 headings can now have an anchor set directly by using the heading template. Use the title or @ parameter, to override the default that would otherwise be generated. I implemented this so that headings with complex formatting could have sane anchors. :)

{{*/s}} and related may be useful for lists you want to class, I wrote the templates as work-around for not being able to style Mediawiki style lists directly, so that I could apply Indexstyles/Templatestyles to lists, and have list items, with multiple blocks/paragrpahs as needed. (NB. You might want to update the documentation to be more comprehensive). (NB you will need to use explicit {{*/e}} at the end of a list or sublist though.

Hope this helps. (If you want I can I think generate a list using pywikibot of 'Stripped Tag' pages that are nominally validated (ql:4)?) ShakespeareFan00 (talk) 08:56, 25 May 2022 (UTC)

@CalendulaAsteraceae:

The following PAWS results set https://public.paws.wmcloud.org/User:ShakespeareFan00/ns0_missingend_transcluded, has some entries in that I didn't feel comfortable updating. Perhaps you'd care to take a look? ShakespeareFan00 (talk) 18:29, 26 May 2022 (UTC)

@ShakespeareFan00: I got a 404 for that one, but I would be happy to take a look at those pages. Would also appreciate a list of validated 'Stripped Tag' pages. —CalendulaAsteraceae (talk • contribs) 20:46, 26 May 2022 (UTC)

Hmm - Try it again now - https://public.paws.wmcloud.org/User:ShakespeareFan00/ns0_missingend_transcluded , I'll see what I can some up with for what you mention :) ShakespeareFan00 (talk) 21:34, 26 May 2022 (UTC)

@CalendulaAsteraceae: On your second point - https://public.paws.wmcloud.org/User:ShakespeareFan00/ql4_ns104stripped.txt - 500 or so pages, and I had to fix some 'funky' titles or the query didn't run correctly. ShakespeareFan00 (talk) 22:24, 26 May 2022 (UTC)

@CalendulaAsteraceae: I also generated some queries for the other quality levels:- https://public.paws.wmcloud.org/User:ShakespeareFan00/ql1_ns104stripped.txt https://public.paws.wmcloud.org/User:ShakespeareFan00/ql2_ns104stripped.txt https://public.paws.wmcloud.org/User:ShakespeareFan00/ql3_ns104stripped.txt

I've not generated a list for missing tags as they tend to be a bit larger. ShakespeareFan00 (talk) 10:09, 27 May 2022 (UTC)

Conversion of articles to <pages> format

Hi, thanks for converting some EB1911 articles to <pages> style, but I found an issue (which I've since corrected) when you edited Echinoderma. By using the "onlysection" parameter, you're assuming every page in the range has a section tag. Page:EB1911 - Volume 08.djvu/906 was missing a tag (now added) and so it's content wasn't displaying. Is there an easy way to check that other EB1911 articles you've converted include pages that might also be missing a section tag? Thanks – DivermanAU (talk) 22:54, 9 June 2022 (UTC)

@DivermanAU: I don't know of a good way to search my contributions in particular, but this search query should get you most of the potentially problematic pages. —CalendulaAsteraceae (talk • contribs) 05:37, 10 June 2022 (UTC)

Thanks for the query - but there's a lot of pages in the report but no indication if any have problems. Would the Source Checker ( https://checker.toolforge.org/?db=enwikisource_p&title=Index:EB1911_-_Volume_08.djvu ) reveal the issue? DivermanAU (talk) 12:07, 10 June 2022 (UTC)

@DivermanAU: The Checker tool (or the new JS-based replacement being tested, see WS:S) will let you find pages that are unexpectedly not transcluded on a single index. I'm not aware of any tool that will give you that kind of info across a set of indexes. Xover (talk) 13:35, 10 June 2022 (UTC)

Incidentally, this is why I think it's best to use fromsection/tosection by default and only break out the onlysection approach in exceptional cases (e.g. where fromsection will be extra tedious, or in cases where you have full control and can plan your transclusions based on it). I've made this particular mistake a few too many times. :) Xover (talk) 08:13, 10 June 2022 (UTC)

I did an experiment by creating Militia (which spans three pages) and deliberately using "onlysection" and leaving one page without a section tag (so that page did not display, as expected). Unfortunately when I ran the Source Checker I mentioned above, it regarded the page Page:EB1911 - Volume 18.djvu/474 as being transcluded even though the text from the page did not appear in the article. I assume the Checker only looks to see if a page is included in a range, and not whether that page has a section tag which is specified in the page range.

I've since added a section tag to the page in question so it appears in the article now. DivermanAU (talk) 21:17, 10 June 2022 (UTC)

Oh, weird. I'm guessing this is caused by the proofreadpage extension (or possibly LST, which PRP uses under the hood for this) generating a dependency on that page as a whole, but then filtering out its contents on display due to the onlysection. As far as MediaWiki is concerned, the page has been transcluded even if none of its contents are displayed anywhere. Xover (talk) 23:10, 10 June 2022 (UTC)

my failure, your clean-up?

I just lose it when I get to adding new authors to the author pages. Me and the alphabet.... Also, I was patrolling one day and I prevented an ip edit from bolding an author. Since then, I have guessed that the bold on the author pages means awards (due to the defense given by the ip editor.

Adding new authors to the author list... I was spotty before that and just don't do it since then. So, my clean-up request:

Could you go through my "contributions" and add the authors I created to the right pages? The whole thing has just been a nightmare to me, including that little ticky feeling when I know I am avoiding something and will continue to do so.

Me and the alphabet, not so good of friends since 8th grade....--RaboKarbakian (talk) 18:21, 3 December 2022 (UTC)

@RaboKarbakian sure! Noting for my own reference that I'm going off of this query and it's been pretty straightforward so far with the exception of Author:Florence Augusta Merriam Bailey. —CalendulaAsteraceae (talk • contribs) 19:35, 3 December 2022 (UTC)

Thank you so much! 'has been bothering me for more than a year now! I am sorry for the number on that list -- mostly it is the magazine, and quite a bit of authors of children's lit. which is for sure the reason they were not here to begin with. I was counting on your abilities with the query, which I also lack....--RaboKarbakian (talk) 21:06, 3 December 2022 (UTC)

Index:Gems of Chinese literature (1922).djvu

Do you have 'something stronger' to standardise the formatting in this? I'll continue lint matching tags, but it might need someone to compell a more logical standardised format. 13:39, 20 February 2023 (UTC) ShakespeareFan00 (talk) 13:39, 20 February 2023 (UTC)

@ShakespeareFan00 Hmm, so my first thought is that the {{p}} tags for ordinary text aren't necessary; normal paragraphs are fine. Once those have been taken out, the headers can be formatted with {{pseudoheading}} and index styles. For example: Page:Gems of Chinese literature (1922).djvu/242. —CalendulaAsteraceae (talk • contribs) 22:44, 20 February 2023 (UTC)

Feel free to proceed in stripping uneeded P then :) ShakespeareFan00 (talk) 22:51, 20 February 2023 (UTC)

Template:UKSI/paragraph and related..

There are a set of wrapper templates that handle various different levels.. However these were developed prior to Lua being available and prior to Templatestyles..

These were created so that there was a standard format/ style for UK Statutory Instruments on English Wikisource. Would you be willing to look into collating the family into a module like you are trying to do with getting headings under control?

Also {{UKroadsign}} and related would possibly be easier to maintain as a Module, albiet in the process it might be plausible to simplify it a little bit.. ShakespeareFan00 (talk) 20:39, 27 February 2023 (UTC)

Yes, I would be happy to work on these templates. Thank you for asking, and I'll keep you posted! —CalendulaAsteraceae (talk • contribs) 01:10, 28 February 2023 (UTC)

Page:Urbiztondo_Ordinance_no._1-_2022.pdf/1

Hi, id like a second view on something...

IS there a template or module to do the indentation and numbering? {{Numbered div}} is a bit cumbersome and should be overhauled. ShakespeareFan00 (talk) 17:47, 2 March 2023 (UTC)

@ShakespeareFan00 IMO it's easier to just use a numbered list and use CSS to get the right kind of numbering. I've done the first page as a demo. —CalendulaAsteraceae (talk • contribs) 03:06, 3 March 2023 (UTC)

Thanks..

Can you also look into modularising {{*!/s}} and others, with a view to making them able to support the id numbers and classing you recommend?

I created those templates so I didn't have to do as much in terms of raw HTML in page namespace.

ShakespeareFan00 (talk) 07:57, 3 March 2023 (UTC)

Hmm. - https://en.wikisource.org/wiki/Page:Urbiztondo_Ordinance_no._1-_2022.pdf/7. We can't set custom list marker types. ShakespeareFan00 (talk) 11:36, 3 March 2023 (UTC)

@CalendulaAsteraceae: The W3C notes on CSS COunter Styles Levle 3 even have a way to do this formatting, but it can't be placed in an Indexstyle as it doesn't recognise <nowiwki>@counterstyle</nowiwki>, Also @Xover: ShakespeareFan00 (talk) 11:45, 3 March 2023 (UTC)

@ShakespeareFan00, I'll take a look! —CalendulaAsteraceae (talk • contribs) 01:29, 4 March 2023 (UTC)

Specfically, the list should go x,y,z,..aa,bb,cc,dd, etc.. not aa, ab, ac .. ShakespeareFan00 (talk) 17:12, 4 March 2023 (UTC)

Yep, I noticed! That's one of the things I will need to look into. —CalendulaAsteraceae (talk • contribs) 20:21, 4 March 2023 (UTC)

Block center an redirects...

Once you've rationalised the headings, Any chance you could look into "pseudo" poems using {{Block center}} <poem>...</poem> and related? In many many instances , they should be using {{ppoem}} which is now reasonably stable. ShakespeareFan00 (talk) 18:02, 12 March 2023 (UTC)

Ooh, yes… —CalendulaAsteraceae (talk • contribs) 00:05, 13 March 2023 (UTC)

Cleanups...

Thanks for the params cleanup.. Any chance you could look into these which need de-liniting?

https://public-paws.wmcloud.org/User:ShakespeareFan00/u_pair https://public-paws.wmcloud.org/User:ShakespeareFan00/i_pair https://public-paws.wmcloud.org/User:ShakespeareFan00/i_space https://public-paws.wmcloud.org/User:ShakespeareFan00/b_space https://public-paws.wmcloud.org/User:ShakespeareFan00/s_space ShakespeareFan00 (talk) 17:06, 3 October 2023 (UTC)

@ShakespeareFan00: I took a look at the s_space scan, and all of the errors were the result of bad OCR. I fixed the immediate problem by replacing the pages in question with mediocre OCR, but your scans will be more helpful if you don't include un-proofread pages. —CalendulaAsteraceae (talk • contribs) 04:23, 5 October 2023 (UTC)

You can skip Vol 6 of Complete peerage (I will consider taking it manually out of the search results anyway), it needs a regeneration of the entire OCR layer anwyay due to page layout changes in the original.ShakespeareFan00 (talk) 08:26, 5 October 2023 (UTC)

Thanks for the cleanups, it helps immensely. Any objections to essentially helping proofread the text portions of Index:The University Hymn Book.djvu, as you'd done already? ShakespeareFan00 (talk) 13:42, 5 October 2023 (UTC)

The objective with the i_space and i_pair lists was so that those tags when balanced, could be automatically converted over to using "appropriate" wikitext. <U>...</U> tags can be replaced with {{U}} once balanced as well.

Once the lints are resolved, a LOT of automated conversions are more feasible. ShakespeareFan00 (talk) 13:42, 5 October 2023 (UTC)

@CalendulaAsteraceae: How good are you with obscure notations?

https://en.wikisource.org/w/index.php?title=Special:LintErrors/missing-end-tag&offset=1960971&exactmatch=1&namespace=104&tag=p&template=all&titlecategorysearch=

This is a highly significant work, but it seems to have a lot of issues with P tags. ShakespeareFan00 (talk) 08:36, 6 October 2023 (UTC)

{{span}} and {{p}}

https://en.wikisource.org/w/index.php?title=Template:Span_tag&diff=prev&oldid=13491919

Nice idea.. but both these templates should be module based so that the giant switch code is instead a set of "/data" and can thus be changed without needing to patch the parse template? (The same approach was used with {{ts}} a while back and it made things more efficient :) ShakespeareFan00 (talk) 17:39, 9 October 2023 (UTC)

Cleanups (Part 3)

https://en.wikisource.org/w/index.php?title=Special:LintErrors/missing-end-tag&exactmatch=1&namespace=104&tag=table&template=all&titlecategorysearch=

In most cases it's simply a case of adding the right closure for a table in the footer... ShakespeareFan00 (talk) 09:08, 11 October 2023 (UTC)

Running header module changes...

You might find this useful in checking if your consolidations broke things :) https://en.wikisource.org/w/index.php?search=insource%3A%2Fwst-running-header%2F&title=Special%3ASearch&profile=advanced&fulltext=1&ns0=1&ns1=1&ns2=1&ns3=1&ns4=1&ns5=1&ns6=1&ns7=1&ns8=1&ns9=1&ns10=1&ns11=1&ns12=1&ns13=1&ns14=1&ns15=1&ns100=1&ns101=1&ns102=1&ns103=1&ns104=1&ns105=1&ns106=1&ns107=1&ns114=1&ns115=1&ns710=1&ns711=1&ns828=1&ns829=1&ns2300=1&ns2301=1&ns2302=1&ns2303=1 ShakespeareFan00 (talk) 16:51, 12 November 2023 (UTC)

Selectors

In Index:The New International Encyclopædia 1st ed. v. 01.djvu/styles.css you use the selector .wst-running-header.wst-running-header-3 > div:first-child. At first blush that looks pretty highly-specified. Is that deliberate? I haven't checked the cascade in play there, but I would have started with .wst-running-header-3 > :first-child. The class .wst-running-header-3 cannot appear anywhere that .wst-running-header isn't also applied, and it can never have any children but the ones we add from code so specifying that we only want divs that are the first child is unnecessary. We might of course need higher specificity to override other selectors (from the TemplateStyles, typically), but my first thought was that it seems over-specified as is. Just a thought... Xover (talk) 07:29, 14 November 2023 (UTC)

@Xover: I wasn't putting too much thought into it. .wst-running-header-3 > :first-child does seem to be specific enough! —CalendulaAsteraceae (talk • contribs) 22:58, 14 November 2023 (UTC)

Edit summaries

Edit summaries are always nice, but when editing code-ish stuff it's more or less a requirement. Both to let other contributors know what you are doing, but also when looking back through history to find when something changed and why. Think of them as turn lights in traffic: without them other drivers won't be able to tell what you're planning to do and there will be a lot more fender benders. :-) Xover (talk) 10:11, 18 November 2023 (UTC)

Good point, I'll pay more attention to that going forward! —CalendulaAsteraceae (talk • contribs) 21:13, 18 November 2023 (UTC)

Header

Just a heads up... I noticed you've sandboxed some stuff for {{header}}. That template is a kind of special case with extremely high risk. My call has been that sandboxing followed by a big bang switchover is too risky. My plan, that I've been slowly chipping away at when time has allowed, has been to slowly migrate pieces of the template to Module:Header until there's nothing left in the template. At that point we'll have much better control and can code around most edge cases or transitions needed. Not sure what you're planning for the sandbox stuff, so this might be entirely irrelevant, but just wanted to give you a heads up before you step on a landmine there. Xover (talk) 11:27, 19 November 2023 (UTC)

@Xover: Thank you, this is a useful heads up. My current plans are roughly as follows:

1. Make sure that Module:Header/sandbox behaves correctly with the current {{header}} code. (It's not quite there yet, but I think I can do it.)
   • I've written code to allow for some changes in behavior, like putting a "by" before override_author and defaultsorting titles without "A/An/The" in front, but it's not enabled.
2. Work on using Module:Header/container for {{author}}, {{portal header}}, {{process header}} and {{translation header}} as well.
3. Once everything looks like it's working, the transition plan:
   1. Take the CSS from Template:Header/sandbox/styles.css and add it to Template:Header/styles.css. This is adding CSS from Template:Header/main block/styles.css and Template:Header/notes block/styles.css but is not changing any definitions.
   2. In Module:Header/container, change "Template:Header/sandbox/styles.css" to "Template:Header/styles.css".
   3. Protect Module:Header/container.
   4. Add the code in Module:Header/sandbox to Module:Header.
   5. Make {{header}} simply an invocation of Module:Header.
   6. Remove the legacy functions from Module:Header.

Thoughts? —CalendulaAsteraceae (talk • contribs) 06:04, 21 November 2023 (UTC)

Why a submodule?

You'll never be able to catch all the weird edge cases in {{header}} (and the umpteen other templates that wrap it; including being called by Proofread Page itself) use in /testcases. People have been allowed and encouraged to throw any old thing into its parameters that seems to "work" (fsvo) over the last 10+ years, and there's a lot of "magical" behaviour that people have relied on. That's why I've been approaching it piecemeal. Pick one bit of the template code, reimplement it in Lua, switch the template over to call that part of the module, and then give it a good long time in production to shake out and notice any issues. Rinse and repeat for the next bit of template code. This is much much safer, much less likely to lead to catastrophic failures that take time to track down and fix (and attendant mobs of community members with pitchforks and torches), and less likely to leave subtle problems lurking undiscovered somewhere.

Once this slow-and-steady process has everything over in Lua, even if it's badly designed by straight replicating the template-based design, you can start iteratively refactoring the module. Switching to use Module:Header/year for example. And at that point it'll be easier to see what to abstract out to a more general-purpose module and what to keep in-core in Module:Header. At that point it'll also be easier to start proposing changes to behaviour and interface. For example, I want to kill |override_author= in favour of an approach similar to the CS1 modules on enWP (e.g. |author1-name= + |author1-display=). Maybe untangle the "please don't remove params" logic that is unlike every other template in existence. Look into some way to make it be smarter about subpages. Behave better when called from Proofread Page. Redo the markup structure and stylesheet to be more modern, scale better on mobile, be more flexible for long titles or chapter names, etc. etc. There's a lot of technical debt and bitrot in there, and a lot of opportunities for improvement too.

Apart from being big-bang-based, your transition plan looks fine to me. Xover (talk) 08:08, 21 November 2023 (UTC)

@Xover: That's a good point; it's easier to figure out how we want to do things if we do them piece-by-piece. In that case, my first proposal is to add require('strict') and declare all the variables, which I hope is pretty straightforward—does Module:Header/sandbox look good to you? After that, I'll work on tweaking the year code so that years actually display on the testcases page. —CalendulaAsteraceae (talk • contribs) 03:39, 22 November 2023 (UTC)

Template:Font-old

Thanks, for making the size relative, I'd based the styles on what was on the page at Metawiki. but you seem to have come up with a series of fized ratios that works much better. I even adapted the approach over on Wikibooks ( I shifted the ratios, to be better conformant with observed behaviour for the FONT tag. ShakespeareFan00 (talk) 22:39, 24 November 2023 (UTC)

You're welcome! —CalendulaAsteraceae (talk • contribs) 22:54, 26 November 2023 (UTC)

I'm not sure about the +1, +2 type behaviour though - See https://www.w3.org/TR/html401/present/graphics.html#edef-FONT ShakespeareFan00 (talk) 23:08, 26 November 2023 (UTC)

I might consider adding a line=param to deprecate some uses of U and S tags. ShakespeareFan00 (talk) 23:08, 26 November 2023 (UTC)

Greek

I may have determined the problem that is causing widespread display errors in Greek, specifically with ellision as seen in κατ᾽ ἔμον which should display as two words with a space between, but is instead now being displayed as a single word without a space. The character ᾽ is a combining character, which has been entered with following space, which means it ought to display on its own, but the software may be interpreting this as a double space (space-below and following space) and collapsing the perceived double-spacing to single spacing, which unfortunately removed the desired space between the words.

The realization came when I spotted that this issue does not happen for text in Header templates (see Hymn to Venus (Sappho), for example), so I'm guessing that something can be done locally to prevent the incorrect collapse of spacing in Greek. --EncycloPetey (talk) 03:00, 29 November 2023 (UTC)

Huh, glad you figured it out! —CalendulaAsteraceae (talk • contribs) 03:49, 29 November 2023 (UTC)

@EncycloPetey: I'm pretty sure this is specifically a problem with GentiumPlus, because Hymn to Venus (Sappho) doesn't wrap the Greek text in {{Greek}} and it looks fine, and "κατ᾽ ἔμον" looks fine, and "κατ᾽ ἔμον", but "κατ᾽ ἔμον" has the display bug. —CalendulaAsteraceae (talk • contribs) 03:29, 30 November 2023 (UTC)

Is there a means to prevent the selection of GentiumPlus for display? It's affecting multiple works across en.WS. --EncycloPetey (talk) 04:11, 30 November 2023 (UTC)

@EncycloPetey: Find another suitable font and edit Template:Greek/styles.css. —CalendulaAsteraceae (talk • contribs) 04:12, 30 November 2023 (UTC)

Woah. Hold yer horses!

First of all, Gentium Plus is not any old random font that looks pretty. It's a font specifically designed for this sort of use by SIL, and it is actually maintained by SIL. And it's one of the few custom webfonts made available through the i18n machinery in MediaWiki, which means it is actually available for all visitors unlike any other font (we can't use, say, Google's fontcdn or other third-party webfonts due to the Privacy Policy). Fonts for this kind of purpose are also not replaceable willy-nilly: there are a whole host of concerns—some straightforward, some subtle or obscure—that need to be addressed. Complicated by the fact that Greek text can appear in a whole host of different contexts. Why do you think the testcases are so extensive?

The way to address this is to start by figuring out whether the font is really what's causing this. Or is it something about what we're inputting, or how MediaWiki is treating the input. Eliminating the web browser and local fonts and font rendering engine etc. And if it's the font we need to figure out whether the behaviour is actually incorrect, and if so in what way (this is not always obvious). Some such flagged issues have turned out to be a matter of preference between multiple valid renderings and behaviours. If it's a matter of preference we can request a font feature flag from SIL to switch between the two, and run some suitable community process to decide. If it's an actual error we can report it as a bug and wait for an updated release of the font. Once either have been released we can request that the language team at the WMF update the WMF copy of the font.

This stuff is complicated, and there are usually no quick fixes, is what I'm saying. Xover (talk) 08:03, 30 November 2023 (UTC)

"Dictionary of Spoken Spanish" Edits

Hey, thanks for helping out with cleaning up the Dictionary of Spoken Spanish! I do have some questions about your edits, just in case I should take on the changes as part of my own standards:

1. On the TOC page, I noticed you removed the dot leaders. I usually add these; is it better practice to leave them out? I also usually omit links to the page numbers; what's your rationale for adding them?
2. On Page iv, you used the {{c}} template on each centered line instead of just once; is there a technical reason for this or is it more for clean formatting?
3. Also on Page iv, you removed {{bc/s|width=700px}} and {| cellpadding=0 {{ts|al|w100}} and added{| style="text-align:left; margin:auto; border-collapse:collapse;" instead. I'm not too knowledgeable on CSS styles; should I make this change on the rest of the tables?
4. I noticed you made the CSS page; if it's not too much to ask, could you explain what it does?

I appreciate any information you can provide! I want to do my best on this work since (for some reason) it's the most-viewed work on the whole site.

—SpikeShroom (talk) 20:32, 14 December 2023 (UTC)

@SpikeShroom: Thank you for reaching out, and for working on this bizarrely popular work!

1. Yes, it's better practice to leave out the dot leaders; they're not really supported in HTML and they add a lot of cruft to the page for not a lot of benefit. @Xover has a very detailed explanation of this, but I'm not sure if it's available as a stable page or if I've just seen it in previous discussions. The page number links were just for my convenience, so I could check the formatting of chapter headers for consistency.
2. Using {{c}} on each centered line was just for cleaner formatting; it doesn't make a technical difference.
3. Yes, please do make these changes to the rest of the tables. It's better to use text-relative units like em, rather than px, for block widths (see H:PXWIDTH). You don't need {{bc}} to center a table, you can just use margin:auto; (the relevant part is margin-left:auto; margin-right:auto;). If you do this, the table will be centered on the page, and adjust its width automatically to fit the contents. I also just find {{ts}} overcomplicated and hard to read; I had to go to Special:ExpandTemplates to figure out what it was doing.
   • I'll work on implementing more of the formatting with index CSS, so you don't have to do so much of it manually.
4. The CSS page applies styles to all pages in the index; see Help:Page styles. The styles that are there right now affect elements with the wst-toc-table class—that is, the table of contents, which uses {{TOC begin}}. They apply the formatting that was being done inline with templates, like making the chapters use small caps and adjusting the spacing of the appendix subsections.
   • It's better for screen-reader accessibility and improved copy-pasting to use text-transform:uppercase; and use normal capitalization on text that's written in all caps for stylistic reasons.

—CalendulaAsteraceae (talk • contribs) 21:43, 14 December 2023 (UTC)

Did someone set out a nice big soapbox and then invite me to get up on it? Oh, yes, precious, they certainly did! 😎

But since I've been kinda overdoing it with the soapboxing lately and my throat is getting sore, I'll spare you the extended version here and instead just link to WS:S#Orley Farm Contents+Illustrations Lists.

Regarding centering… using {{c|…}} on a single line is reasonably robust, but the second you put more than one line into a template parameter you 1) run into heightened risk that something will go wonky, and 2) you make the wikicode much harder to read (to keep track of where it starts and ends). That's why we, for most templates, have block variants like {{c/s}} and {{c/e}} (short for {{center/s}} and {{center/e}}); {{fine block/s}} / {{fine block/e}}; {{hi/s}} / {{hi/e}}; and so forth. These start+end variants are a lot more robust, can even be used across Page: pages if needed, and it's a lot easier for someone reading to see where the centered (or whatever) content starts and ends. Xover (talk) 06:58, 15 December 2023 (UTC)

translation license

I had need to debug {{translation license}} and noticed its output is… well, not to put too fine a point on it, it's pretty yuck. :-)

Why does it need this many nested divs (not to mention the spans)?

            <div class="licenseContainer licenseBanner dynlayout-exempt">
              <div>
                <div>
                  <div>
                    <div>
                      <p>
                        <span class="imageLeft">
                          <span class="mw-default-size" typeof="mw:File">
                            <span>
                              <img …>
                            </span>
                          </span>
                        </span>
            </><!-- close omnes -->

Two of those useless spans are added by MediaWiki (for some darn reason) but we exacerbate it with that extra wrapper span instead of just setting the class directly on MW's image output.

And why does the styling need to be using selectors more akin to XPath than CSS?

.licenseContainer > div:first-child > div:first-child > div:first-child > div:nth-child(2) > div:nth-child(2) > div:nth-child(2) {
  …
}

So before I go start tearing out stuff there (a layout table? surely not!), am I missing something or is this just replicating an old pre-Lua markup structure and styling? Xover (talk) 07:13, 15 December 2023 (UTC)

@Xover: Yep, this is just replicating the pre-Lua structure; tear away. —CalendulaAsteraceae (talk • contribs) 20:19, 15 December 2023 (UTC)

Cleanups-

Index:Beowulf (Wyatt).djvu Can you take a look over this, it's generating a lot of paired Missing/Stripped lint concerns.. I suppect it's due to a SPAN based template being used instead of DIV based one, but it's widespread over the work in question. ShakespeareFan00 (talk) 10:25, 22 December 2023 (UTC)

@ShakespeareFan00: I fixed the stripped tag error on Page:Beowulf (Wyatt).djvu/126 by editing pages 125, 126 and 127 to use {{lang block/s|ang|class=_saxon}} at the start and {{lang block/e}} at the end instead of using a separate {{lang}} invocation on each page. I don't really feel like doing the rest of the pages right now, but hopefully this is a helpful model; let me know if you do this and there are still stripped tag errors when you're done. —CalendulaAsteraceae (talk • contribs) 08:21, 23 December 2023 (UTC)

Noted.. I'm converting currently, but it will take a while. ShakespeareFan00 (talk) 16:10, 23 December 2023 (UTC)

Template:Physical Revue TOC

I feel like this template is a bit unwieldy where it appears on a work edition page; what would you think if I were to replace it with a link to a new portal Portal:The Physical Revue? —Beleg Tâl (talk) 14:56, 17 January 2024 (UTC)

Sounds good to me! —CalendulaAsteraceae (talk • contribs) 15:42, 17 January 2024 (UTC)

@Beleg Tâl: I've created Portal:The Physical Revue and linked it from the relevant songs. Feel free to add to and edit it! —CalendulaAsteraceae (talk • contribs) 04:20, 20 January 2024 (UTC)

Awesome! —Beleg Tâl (talk) 20:45, 20 January 2024 (UTC)

Form

I'm curious about this removal, which would have classified the song by its period. --EncycloPetey (talk) 07:29, 20 January 2024 (UTC)

I made Category:Cantigas d'amigo a subcategory of Category:Medieval songs, since Wikipedia says it's a genre of medieval lyric poetry. —CalendulaAsteraceae (talk • contribs) 07:30, 20 January 2024 (UTC)

But that conflates genre and time period as categorization, which is why I did not do that. These items are now categorized by genre explicitly, but not by form/date. And while the cantiga originated in the middle ages, it's still written / performed even today. It's a genre. --EncycloPetey (talk) 17:26, 20 January 2024 (UTC)

That's a good point; thanks for checking my work! —CalendulaAsteraceae (talk • contribs) 19:51, 20 January 2024 (UTC)

Template:Biblecontents2

Please re-check, your update is causing it to throw out unmatched P tags (according to Special:Linterrrors)ShakespeareFan00 (talk) 08:51, 20 January 2024 (UTC)

@ShakespeareFan00: Fixed now? —CalendulaAsteraceae (talk • contribs) 20:11, 20 January 2024 (UTC)

Template:Citation

Could you consider converting this to a module? I've got a really hard to track down LintError in Page:Biodiversity Assessment of the Fishes of Saba Bank Atoll, Netherlands Antilles.pdf/37 and the current {{citation/core}} template is too complicated to maintain/debug. ShakespeareFan00 (talk) 12:42, 22 January 2024 (UTC)

Oof, yeah, I'll take a look. —CalendulaAsteraceae (talk • contribs) 15:03, 22 January 2024 (UTC)

@ShakespeareFan00: I've mostly finished working on the module. The only remaining issue is Wikisource:Scriptorium/Help#Lua equivalents of #time and #dateformat?. —CalendulaAsteraceae (talk • contribs) 18:45, 25 January 2024 (UTC)

Thank you!

Big thanks for WD version now being a module, etc.!!--RaboKarbakian (talk) 14:30, 22 January 2024 (UTC)

You're welcome! —CalendulaAsteraceae (talk • contribs) 15:02, 22 January 2024 (UTC)

Comment on “Early Results of Slanted Recession of the Lateral Rectus Muscle for Intermittent Exotropia with Convergence Insufficiency”

Throws a missing SPAN lint. But thanks for updating the citation template to Lua should be more strightforward for someone to track down the glitch. ShakespeareFan00 (talk) 20:53, 26 January 2024 (UTC)

Fixed; thanks for pointing out the issue! —CalendulaAsteraceae (talk • contribs) 00:25, 27 January 2024 (UTC)