In T363009#9731654, @Dzahn wrote:

@Izno Wanna tell me what you dislike about it? Would there really be a difference whether people sign something here or on a wiki?

The first part was T54582: TemplateData: Allow hinting to specify auto-numbered parameter names in some fashion.

Sample edit

Don't forget Unicode has other plans for the length of other character sets. :)

I'm not particularly motivated by the idea that anonymous (soon temporary) and logged-in users should be treated differently - the only change made in the default summaries is to remove the talk page link, but in general editors want to be able to communicate with other editor, even logged-out/temporary ones, and there's no technical reason to distinguish these cases as I see it.

Edit summaries have a length limit and IPv6 strings of hex numbers hit it. It's also practically useless to try to talk to the vast majority because they rotate as quickly as they do. Which may or may not be improved by temporary accounts (no one on the project has tried to demonstrate how many temporary accounts will be created in comparison to how many IPs edit though I think there has been an estimate or two).

Today, the example page doesn't have the reported issue. Are there still some pages that do?

I agree, I think that satisfies the desire for this task. (I think I'd like to see "template used on page" to load a gadget so that we can avoid adding some inevitable categories that will duplicate template links 1:1, but that's a separate feature request.)

In T361785#9688067, @hgzh wrote:

I think this is caused by https://en.wikipedia.org/wiki/User:Headbomb/unreliable.js which applies #ffdddd background color as inline style to a.external when a link to wikipedia gets matched. plainlinks overrides the background color, but the inline style persists and gets chatched by the night mode rule.

In today's world this can be done using a module like Module:Diff, but it would be a little rough to have to copy-paste two separate versions of a page into that input.

In T31860#5407147, @Aklapper wrote:

There should be a time slider that allows a user to set the period of diffs that are highlighted, with the slider defaulting to either the time of the user's last visit to the article, or a reference point the user has set.

That would be Revision-Slider nowadays.

display only a link to them on mobile devices

I suggested something like this in T124168#7493247.

Both would tend to lighten them slightly or be a wash (see a commentary I've put together). It isn't fundamentally the overhead setting up the boxes that makes these heavy, it's the fact that people will put 100s of links in each navbox that makes them heavy. Or even 20s of links in each navbox and then having 5+ navboxes on a page. (Or worst, 5+ navboxes with 100s of links.)

@media all and ( max-width: @width-breakpoint-tablet )

Well, ignoring that we can access no LESS design tokens (and IDK if these tokens are in CSS yet, and couldn't do anything with them outside MediaWiki:Minerva.css anyway yet), tablet breakpoint is 1120px, which has probably been too wide anyway for what this template does. I'll opt for the onwiki-common 720px and take care of enwiki myself. These tables are (or should be!) marked as presentational anyway so not like we're losing anything for screen readers.

Probably the task in general is worth a user notice. "Minerva, the skin used on mobile, has had some CSS in place to make some templates display better at mobile resolutions. Skins shouldn't decide how templates display, so some of this CSS will be removed this week. If you use one of the set of affected classes somewhere on your wiki (put list in description, with a link to Special:Search maybe for each class), you may need to add TemplateStyles to support the same display." or something like that... I don't really like the description of how to fix it inline since the changes may be pretty different (or not needed as with e.g. .collapsible td) so maybe leave that to a description in this task or another task to point users to.

In T360683#9675614, @Izno wrote:

1. Ban its use in main space. All valid uses transitioned to specific "semantic" templates and the rest removed (e.g. sports already has some that could be called "semantic").
   • 500k total transclusions, 175k of which are in main space. This plan seems infeasible on any useful timeframe.

Somewhere in order of preference, but which better consensus than my opinion would be needed:

In T361407#9674209, @Krinkle wrote:

In T361407#9674162, @Bawolff wrote:

Personally i was a fan of <source>....

T39042: Remove <source> syntax from SyntaxHighlight (GeSHi)

I would not oppose declining this in favour of also declining that.

Likely caused by T205341: Replace reset.css rules with ResourceLoaderSkinModule `normalize` feature.

The work I know of has been due to trying to find a home for hacks.less CSS/LESS currently living in Minerva skin repository. T358071 (As I have suggested I think in the community discussion subtask, I think MF is the appropriate place for hacks, but that was only true for the mobile-related hacks, rather than the dark-theme hacks now also being added in hacks.less.)

Or I can just wait for variables/add the extra rules in the sheet. :)

In T359894#9663697, @stjn wrote:

Another alternative is for .cs1-visible-error to implement itself via standard error class while resetting some of its properties.

Ok for multicol I've now added it to the to do page, I'll go add the styles now.

Drop .collapsible td. This is antiquated code that has been replaced with mw-collapsible.

This isn't what I said, so I'm not sure how you came to this conclusion. I said that there is a replacement, but that we still have many uses of the old .collapsible. And that I didn't know precisely what the line of interest was intended to do. It's probably still fair to remove it ultimately, but it's a risk that you take that no one will be super-concerned with it.

I had hoped to be having this discussion in the community discussion task just so we could be very clear before making to-be-actioned tasks......

.content .reflist { column-gap: 2em; } doesn't apply to elements that are display: block

This is not a true statement. I think this still merits removal for the reason I expressed in the discussion task.