I answered at [ https://www.mediawiki.org/wiki/Topic:W9ku8t7v6yh0mam8 | mediawiki.org ] and it is not really a MediaWiki problem.

It is an issue who will monitor edits in which namespaces and trigger tagging.

• On global level this is soemthing around parsing, or similar to mentioning users (notification).
• On site level a cloud tool would be required, which could listen to changes on all WMF wikis.
  • That cloud tool could maintain a database which keeps track of all media identifiers, both site local and commons, via pageid to be robust against file moving.
  • A database could provide image descriptions in a particular language for other pages with transclusion of the same image.
  • Editors could use the text from another page, or use it as base for (automatic) translation and refining in the own language.
  • A JavaScript gadget may be written to be triggered by blind people which would equip the HTML document by alt texts if database knows any image description in an appropriate user language, or might try to provide an automatic translation on the fly.

Tagging edits as requested in this task is not the solution for the goals. It might be one single piece in the jigsaw puzzle for one particular approach. It is not really required. Tagging itself is no solution, and it is pointless without other tools to set the tags, and to evaluate tagged edits.

If at least one tag has been established, revision history will be displayed with an additional checkbox for each revision since registered users will be able to edit tags for this revision.

This should be done on global level, if ever.

It is confusing to use well defined HTML elements for something different.

This is simply not true. The following post has opened a doubled thread for <resourceloader module="ext.gadget.mybookgadget" /> here.

The comment before is cross-posting at T241524 so I feel free to answer by C&P here as well.

• The proposal is also creating a risk of undiscovered vandalism.
  • Malicious users, even every anonymous user can trigger all available gadgets and increase client load of all visitors. Since not useful in current situation, no gadget will have any visible effect and attack may be detected years later.
• It is not clear what is taken precedence over explicit user opt-out. Will the proposed element override user option?
• There are not many templates which have a reasonable connection with a specific template.
  • A template for coordinates may trigger the local adaption of WISWOSM. That can be administrated by the new syntax of T204201 and narrow such gadget calls to pages where coordinates really occur.
  • If a gadget is related to a particular template it is up to inventors to tailor the usage precisely that transclusions and gadget loading are matching well.
  • It is even possible to define a dummy template which creates no wikitext output at all, and has the one and only purpose to trigger gadget loading. That is exactly what the proposed new element shall do, but limited to exactly those gadgets which are defined for such assignment.
• There will be only very, very few templates which are related to a gadget.
  • The total number of gadgets is not very large, and it is an exception that a template is related to a possible gadget in that wiki.
  • There is not a huge number of applications with relation of content and JavaScript. Table sorting, collapsible elements. Those as well as extensions like Graph and Kartographer are under MediaWiki control.
• Management of gadgets shall be kept limited to interface administrators by maintaining Gadget Definitions specifications.

This entire proposal has high risks of misuse. It has no benefit which would not be available via extended Gadget Definitions, but creates double work and complicated additional framework for a single exotic purpose.

• The proposal is also creating a risk of undiscovered vandalism.
  • Malicious users, even every anonymous user can trigger all available gadgets and increase client load of all visitors. Since not useful in current situation, no gadget will have any visible effect and attack may be detected years later.
• It is not clear what is taken precedence over explicit user opt-out. Will the proposed element override user option?
• There are not many templates which have a reasonable connection with a specific template.
  • A template for coordinates may trigger the local adaption of WISWOSM. That can be administrated by the new syntax of T204201 and narrow such gadget calls to pages where coordinates really occur.
  • If a gadget is related to a particular template it is up to inventors to tailor the usage precisely that transclusions and gadget loading are matching well.
  • It is even possible to define a dummy template which creates no wikitext output at all, and has the one and only purpose to trigger gadget loading. That is exactly what the proposed new element shall do, but limited to exactly those gadgets which are defined for such assignment.
• There will be only very, very few templates which are related to a gadget.
  • The total number of gadgets is not very large, and it is an exception that a template is related to a possible gadget in that wiki.
  • There is not a huge number of applications with relation of content and JavaScript. Table sorting, collapsible elements. Those as well as extensions like Graph and Kartographer are under MediaWiki control.
• Management of gadgets shall be kept limited to interface administrators by maintaining Gadget Definitions specifications.

This entire proposal has high risks of misuse. It has no benefit which would not be available via extended Gadget Definitions, but creates double work and complicated additional framework for a single exotic purpose.

Do one thing, and do it well.

• Please use a class mw-page-position-top.
• That does say: This block shall be arranged before introduction section.
• Nothing more.

Those local ambox and hatnote and infobox are decoration classes. They define how this particular wiki is decorating things. That is not a statement on the exact position where to arrange it in a page.

• Structural global semantics are one thing.
• Local decoration is something different.

The ambox and hatnote are providing a particular decoration. They were never limited to certain structural declaration and effect on order of elements. Those are definitely the wrong class names to introduce a new capability. Nobody knows which templates in 1000 WMF wikis are equipped with such classes and which expectation and usage are supposed there.

I don’t think that this proposal should be implemented. It is a rather hard break in usual cooperation of interface, page content and JavaScript gadgets.

• T204201 Extend MediaWiki:Gadgets-definition capabilities would solve the JavaScript loading issues much better. It serves much more cases and is flexible enough, e.g. by loading JavaScript if a certain template is transcluded.
• Introduction of new <elements> will destroy the separation of HTML content, as provided by wikitext, and JavaScript which is part of current interface and basically independent of HTML. CSS is quite close to HTML presentation.
• Gadget administration should be done in a central place, not scattered over source code of templates or even articles.

I don’t think that the recent proposal should be implemented. It is a rather hard break in usual cooperation of interface, page content and JavaScript gadgets.

• T204201 Extend MediaWiki:Gadgets-definition capabilities would solve the JavaScript loading issues much better. It serves much more cases and is flexible enough, e.g. by loading JavaScript if a certain template is transcluded.
• Introduction of new elements will destroy the separation of HTML content, as provided by wikitext, and JavaScript which is part of current interface and basically independent of HTML. CSS is more related to HTML presentation.
• Gadget administration should be done in a central place, not scattered over source code of templates or even articles (book content).

Yeah, I guessed something like that from a few examples, but they had various causes.

Do one thing, and do it well.

• I did suggest a class mw-page-position-top.
• That does say: This block shall be arranged before introduction section.
• Nothing more.

Those local ambox and hatnote and infobox are decoration classes. They define how this particular wiki is decorating things. That is not a statement on the exact position where to arrange it in a page.

• Structural global semantics are one thing.
• Local decoration is something different.

MediaWiki might publish a global Style Guide for such issues.

• Then projects do need 12 months at least to implement all such features.
• All class names are to be prefixed by: mw-
• A global dictionary of selectors exposed to the public and supposed to be utilized should be distributed immediately by MediaWiki design department.
• A .mw-page-position-top can be applied where desired then.

It is not acceptable to take the local customs of enwiki and use them as unpublished global standard for 1000 WMF wikis. This is a kind of cultural domination and patronizing style which can be observed as well with so-called global templates and global modules. It actually means: Everybody has to do everything exactly as enWP does. MediaWiki is supporting enwiki only, and all others are obliged to follow enwiki community decisions.

• T132308#3263787 For the record: Software and services that are used on hundreds of wikis are not required to abide by the policies or guidelines of any individual wiki.

dewiki has no problem to equip all their related templates with an mw- class.

• Our Template:Hinweisbaustein does know |POSITION=oben which actually means position=top and could trigger an mw- class with one edit only.
• The box design for top position is omitting top and side border lines.

I am looking forward to get informed about the MediaWiki selector dictionary with all these mw- classes.

Ah, and two similar resolved issues wrt author, might help:

• T160845 wrong author listed and wrong first/last name for the one author listed using ISBN lookup
• T203361 Publisher is incorrectly parsed as an "author full name" and split into first and last name when generating from DOI

In T249293#6889884, @Jack_who_built_the_house wrote:

I would also be happy if the page properties potentially brought about by these magic words were introduced to frontend somehow (via a config property, for example), so that user scripts (including mine) could use them.

• Each mw.config.get(wg) is blowing up the size of the delivered HTML page and adding bandwidth.
• In most cases, when reading an article, none of these page properties are needed.
• Delivering a pile of unnecessary page properties increases resource consumption for all readers and users.
• If a particular script is interested in a particular property which happens in 0,0001 % of cases that may be retrieved via API rather easily.
• The number of mw.config globals might be reduced for rare site properties.

There are only two values permitted: yes and no.

A change in strategy is needed. Apparently that has been started some years ago already.

• id= should not be used for anything else than to enable the reader to jump to a certain point in document.
• Elements to be decorated or manipulated shall be accessed by mw- classes only.
• If necessary, the mw- class might have one member only. That is the replacement for elements formerly addressed by id=.
• Implementing this consequently no conflict between headlines and MW system activities could occur any longer.

Migratíon to this paradigm is on the way, as see.

For example, by prepending it with h- for heading,

That is no help at all.

Actually I was wondering why Lua did not get fragment identifiers but JS/CSS received them. There are small doc texts possible on JS/CSS code pages by MediaWiki:Clearyourcache but those are short and constant and may be kept free of duplicates easily.

We do need user defined so-called anchors, and they are spread in millions of occurences over the wiki projects.

Just to clarify: Line numbers are working with Lua, see mw:Module:Gerrit, but no id= yet.

The major purpose is to have as first step some results from incategory: and any keyword search.

• Based upon the result set a bunch of pageids is collected.
• Then some maintenance operation takes place, perhaps a bot run.
• Finally, with the original query the search is repeated on the pageid set. If successful no results should be present. If still matches are showing up it needs to be polished again.

Any namespace limitation is pointless if a set of pageids is provided.

There is some confusion over the particular targets.

1. request of 2011 has been for entire code pages: On [[MediaWiki:Common.js]] and other [[*.js]] and [[*.css]] pages
2. idea has been to equip <syntaxhighlight> blocks with unique id=

The first one is feasible.

There is no English word saneitize, and at least confusing.

Yes, thanks, meanwhile I helped myself with action=templatedata now. However, I do need displaytitle from pageprops as well, so I wanted one API call only. Now I make two calls every 50 templates. I can save decompression of result now, consuming more bandwith.

In T266200#6708828, @Krinkle wrote:

@PerfektesChaos That sounds great. What is your question or concern in relation to this task? TemplateData is available for the API, always has been, and is as far as I know also the only supported way of accessing the data. Nothing about that has changed as far as I know.

It has been taken into account to drop storage of TemplateData, be it compressed or not, within DB.

• That should be kept available.
• The track is rather confusing, distributed over three tasks, and I do not see a clear solution.
• One issue is complaining about violation of pageprops for graph and kartographer, considering to drop storing at all. Others seem to think about a new pagepropsBlob for large things, while pageprops is kept for usual tiny things.

In T266200#6708828, @Krinkle wrote:

If you are finding corrupted or wrongly compresed/decompressed data somewhere, please file a high priority task and explain how to reproduce the issue. If you suspect it is related to this one, feel free to mention this task there.

I found that yesterday. Not in DB tables but API. That’s why I looked for this task here. Just try. This issue is already on rails, somewhere. Probably comp/decomp is okay at DB, but result is not yet ASCII-encoded for safe JSON.

Sorry, but I do want to access TemplateData via API. May be within pageprops or anywhere else.

German Wikipedia has disabled this feature as well.

• Our quotes are already enclosed in quotation marks. Nearly always.
• Mostly they are provided via template, which is adding language dependant quotation marks and supporting semantics by <blockquote>, and indenting.
• A certain amount is (mis)using : for indentation and enclosing in quotation marks. Those would get different appearance.
• We abandoned direct usage of <blockquote> which comes from newbies via VE and will be replaced by template. Quite often that one is just misused for indentation, got a stupid grey bar now but no quotation at all.

Apparently depending on browser.

• Firefox works as expected.
• IE10 has no memory.

I do use old browsers frequently to observe how things are working for people with not most recent equipment.

I return to initial state, usually eternity.

Meanwhile I discovered that apparently there is no memory?

• If I select a particular duration and ask for diff or preview during edit, it falls back to eternity. Seems to require to set interval on final Publish only.

A URL parameter would be nice, like &watchperiod= on action=edit.

Please note that the third item (purely decorative images) has been resolved meanwhile.

• aria-hidden="true" role="presentation" will serve that, see T204618.
• You will find on contemporary mw: some examples: The photo of Hackathon participants does not give any clue and will be suppressed, and the icons in the title bars of the blocks are redundant to the textual headline and shall be ignored.

There is no need to grab sourcecode.

• Both nonewsectionlink and noeditsection are pageprops.
• They might be retrieved via API in advance, leaving the tool immediately.

It does not matter. It might be in 2 or 5 or 10 years.

These namespaces are reserved for future usage, and no community shall be permitted to put any single page under a name that will require struggle and moving and efforts later.

On #2: Yes, somebody complained they were confused by the new widget and won’t use it ever. But they want to clean up the writing desk and do not want to change habits.

• It is not worth to introduce a preference on its own, but might be triggered by negative number of days.
• 0 does usually mean infinit or maximum or default.

Maybe; one reason more to avoid confusion and attempt clear structures.

When I discovered VPS subproject list my expectation has been that the full structure of the cloud architecture will be reflected.

• Wikidata-Query-Service is a VPS as well but not yet under any umbrella.

Minus 2:

• https://fountain.wmflabs.org/ does not work, no VPS for now.
• dawiki not yet configured, at least not as VPS.

All VPS subdomains but no other should be derived from Cloud-Services > VPS-Projects rather than from Projects root.

Dealing with JS/CSS resources is no longer sysop business.

• It is up to IA to decide what shall happen with sanitized and deleted code.
• If you do not want a page to be deleted do not ask for deletion.

Dealing with skin resources is no longer sysop business.

• It is up to interface administrators to handle JS/CSS pages.
• The ability to delete JS/CSS pages was left to react on threats as quick as possible, e.g. privacy violation via beacon just discovered.
• People should think twice before asking for page deletion. Once something was deleted it always needs further explanation and requests to get back content which might be copyvio or legally questionable or harrassment or TOU conflict or whatever. It is up to interface administrators to decide upon JS/CSS page content.

It is the intention of this procedure:

• Every regular sysop may delete malicious software immediately.
• A sysop account may be hijacked, somebody may be persuaded to undelete or to forward the bad code. All misusing pages could be exploited then.
• Once the code page was deleted experienced staff has to be asked for undeleting or even copying presumable malware, investigating thoroughly first.
• Recovering malware is doing harm. A suspicious resource which is unavailable for a couple of days until checked and released again does not cause any damage, it is just a tool being temporarily unavailable.
• If there is no good reason, a sysop should not delete pages.

UPDATE 2020
Within the recent five years two more aspects arrived:

1. CSS selectors
2. TemplateStyles

• For generation of any URL, both internal and external, %2E in <a href="...%2E"> would be sufficient. It will be copied by context menu mechanisms as-is and hide the dot from any mail client and office program, whichever URL might be mutilated otherwise. However, that is the same URL as before.
• For our pages themselves they should declare their own URL by appending _ which will be kept in address input field of the browser and may be copied there. If wiki server is called with superfluous _ that is removed anyway.
• On footers, by Special:CiteThisPage etc. the URL of this page, if any, shall be propagated as %2E terminated, but currently ending with digits anyway.

At least there should be a global dictionary for selectors, even if no CSS implementation is delivered.

• German Wikikipedia calls that thing .deviceProperty-not-narrow, another Wiki might define .nohandheld and a third one is using .nosmallmobile as selector.
• This is affecting exchangeability of templates, gadgets and page programming components.
• If there is no global CSS implementation distributed on all pages, every project which likes to use that can specify an implementation by site CSS (MediaWiki:Mobile.css or via MediaWiki:Minerva.css) or TemplateStyles on pages where such elements occur.
• The target is to focus on the main track of the story, if there is a lack of space on a small device. Decorative nice images, the joke of the day and distracting gimmicks shall be omitted. Looking at https://www.mediawiki.org/wiki/MediaWiki the photo is superfluous and does not solve the problem of the visitor, show it on desktop and tablet, but do not fill the entire first screen of a handheld.

The failure first.

Okay, thanks for your efforts.

This may be a misunderstanding.

In T237700#6169065, @Esanders wrote:

If they are obliged to sign with every small wiki individually, they might get problems to track their business distributed over 100 small wiki user talk pages.

You can just turn your local talk page into a soft redirect to the desired one. We also have global notifications now.

In my home wiki I could help myself: Template:WikiLayout/styles@dewiki

• However, that goes for my own wiki only.
• I am heading for a global solution for 1000 WMF wikis and outer space.
• That is what Phabricator is for.

“encourage TemplateStyles adoption”

• That is putting quite a lot of efforts on source editors.

Please compare:

[[File:note.png|class=nomobile-narrow|16px]] NOTE:

with

<templatestyles src="WikiLayout/styles.css" />
[[File:note.png|class=deviceProperty-not-narrow|16px]] NOTE:

The latter is local only, my proposal could be exchanged among wikis.

German Wikipedia does not want that threads in a talk archive might be continued.

• Talk archive pages are in the related talk (or other) space.

Are these pages marked in anyway that would let us know they are archived?

There are various templates, but all of them use both __NOEDITSECTION__ and __NONEWSECTIONLINK__ to prevent thread editing by accident.

German Wikipedia does not want that threads in a talk archive might be continued.

• Talk archive pages are in the related talk (or other) space.
• Talk archive pages are marked as terminated by any archive template and those request explicitly __NOEDITSECTION__ and __NONEWSECTIONLINK__ to prevent thread editing by accident.

DiscussionTool must not invite to reply in a terminated thread.

Strongly opposing.

Well, MediaWiki's Language class is expecting internal queries within PHP infrastructure only. There all codes might have been doncased already.

Thanks. BTW, you might check the heading comment of rMW/ maintenance/markBadBlobs.php – it looks like copied from rMW/maintenance/cleanupImages.php and is bearing wrong purpose, several brion, and 2005-2006.

Please consider a link to a user page with the same nick but a different wiki.

• Imagine a steward or a small wiki support hand. They might link to a user or user talk page at meta: or at their home wiki.
• If they are obliged to sign with every small wiki individually, they might get problems to track their business distributed over 100 small wiki user talk pages.

A main user page or main user talk page should be linked at least, not only subpages. That goes for version history and transparency; it would be easy to channel all complaints into a subpage, then request deletion of the subpage and keep a clean main user talk page version history. German Wikipedia has specific rules on deleting main user talk page (never deleting, might be moved, but moving can be tracked).

Well, the first guess does not change any page:

• If namespace is even and 2, 4, or >90 and there is NEWSECTIONLINK but no NONEWSECTIONLINK then regard this as a forum.
• If namespace is odd and there is NONEWSECTIONLINK then do not permit to extend a thread.

The new keywords to equip rare exceptions where the guess fails may be introduced at a later stage. Therfore no page can be changed for those now.

That might happen by incident, but you cannot prevent on millions of pages that people introduce inappropriate syntax, declare a regular article as disambiguation page or empty the entire content. Projects are patrolling on a regular base and fix such mislead stuff, or search and rescue the John Hopkins University once a quarter.

Well, it seems to depend on browser technology.

• Your example browser is ignoring the orphans/widows rule. They take the wording avoid too seriously.
• I learnt Firefox is not good at this for now.
• https://caniuse.com/#search=orphans
• I recall a browser which divided the S properly as expected, but I forgot which.

The development is proceeding; we should revisit this every spring.

• At least there was progress in keeping things together since 2015.
• Now we are grizzling that it is kept together too much, we want it to get split again.

In T245890#5951850, @Alsee wrote:

The magic words would each override the namespace.

They should be able to override the guess.

• However, the guess should be taken from both namespace and NEWSECTIONLINK (NONEWSECTIONLINK) to avoid tremendous insertion of new magic words within existing pages.

But yes, as I suggested above, a __NOTALKPAGE__ might revocate continuation of closed discussions in archives, and NOTALKPAGE should override EXPECTTALK and namespace and any other.

Nowadays it may be overcome by applying

.mw-category-group {
   break-inside: avoid;
   break-inside: avoid-column;
   page-break-inside: avoid;
   page-break-inside: avoid-column;
   -webkit-column-break-inside: avoid;
   orphans: 3;
   widows: 3;
}

including some fallbacks for CSS2 and older or less smart browsers.

There is a magic indicator: __NEWSECTIONLINK__

That’s entirely right. For a single image (like an icon) the effect is the same.

(+) Some Screenreaders do generate a TOC already from document headlines in the same way as MW does.

Well, on a wiktionary only pages in main namespace are SENSITIVE_ALL, but templates and categories do behave like every other wiki.

Just a reminder:

• The Flag needs three states:
  • IGNORE
  • SENSITIVE_ALL
  • SENSITIVE_2ND or IGNORE1_SENSITIVE
• For hastemplate: and incategory: IGNORE1_SENSITIVE is appropriate.
• On a Wiktionary, linksto: is SENSITIVE_ALL for main namespace, but any other linksto: is IGNORE1_SENSITIVE.
• Common text search is IGNORE.
• When accessing the database tables, it is no problem at all, capitalizing first character of the title part (not namespace) will deliver IGNORE1_SENSITIVE entries.
• There are config variables indicating which page names might need SENSITIVE_ALL.

Strong support for scheduled termination.

• It is not a good idea to maintain current doublespeak in eternity. One day this has to be discontinued.
• It has been discouraged for a couple of years now, it might be 2021 when support for source ends.
• Throwing a maintenance category for a year is a low level action to give opportunity for a smooth change of affected pages.
• It is most confusing in a syntax description that one thing known with an entirely different meaning is an alias for something with another major identifier.
• Even more if <source> is occurring within the displayed HTML code.

German Wikipedia migrated soon after renaming all encyclopedic articles and active project pages.

• Some recent C&P from enwiki, some user pages or archived discussions might still use source but can be cleaned up by bot easily.

The localized version is inserted, e.g. Datei: or Fichier:.

• Localization is blocking C&P transfer into other language wikis.
• German Wikipedia is removing the namespace name by JS tools when source code is edited to enable smooth exchange with other wikis.

The namespace name is redundant and superfluous in gallery and has been a less smart concept there.

• Nothing else than a media ID is permitted at this position, and therefore it is pointless to ask whether it is called Image: or File:.

VE etc. must not insert any namespace name.

• It is cluttering diff and disabling interwiki exchange. At least confusing, no benefit.

This task stalled.