Could you elaborate what "internationalized programming facilities" are, apart from translating documentation on mediawiki.org which is already possible? That would help to understand better what kind of feedback you are looking for in this task. Thanks!

It refers to i.e. T150417, which has been already declined.

The ticket does overlap with T150417. But T150417 was created to go along the corresponding 2016 Community Wishlist Survey proposal.

This ticket try to have a broader scope, but if that was the general idea that was declined, not the Community Wishlist Survey proposal, then this ticket may indeed be closed as well, and I apologize for the misunderstanding.

Otherwise, here are more details.

As explained in T148598, the idea is to implement internationalization facilities (if they don't exist yet) in compilers used in the WM infrastructure, enable contributors to localize them (possibly through Translatewiki), and let them use localized version if they wish.

So having more feedback on known state of the art regarding internationalization in compilers would be interesting here.

Someone in an other ticket (but I didn't retrieve it) seemed to tell that you can localize template magic words and parser functions of templates. Is that the case? If yes, where could I find documentation about it?

Regarding Javascript, there's babylscript. Could it be possibly be integrated within WM environment? If that would be conceivable, what would be the required steps for such an integration?

Regarding Lua, I already began some work for an Esperanto port, and I plan to target a more "internationalize the official interpreter" approach. As said in T148598, so far I worked with Lua 5.3. Would it make more sense to take into account a specific forked version used within WM environment, or should an approach like "focus on introduce that in the last official version, the WM's interpreter will soon or later be updated to follow the official Lua releases"? Or yet an other approach, as the official interpreter currently can't manage Unicode characters, while Luajit can, would it make sense to go with Luajit?

Regarding Python, well, I didn't looked yet at it regarding possibilities to internationalize it, and any feedback on this topic would be welcome. I think Python might indeed also make sense here, as Pywikibot is a popular way to make little scripts. That also make this main use far easier from an integration point of view, as the running environment can be the end user one or some confined server like Tools.

Regarding PHP, that's seems less interesting for average end user, as this is mostly used for Mediawiki core and extensions, which you might expect to require far more technical investment for the end user than the typical use case of previous languages. But still, feedback on this point of view and on the possible internationalization projects of PHP are welcome. :)

Well, I don't know if there is any other programming language which is of heavy use with WM environment, do we have some statistics about programming language rates by the way?

In T150933#2805012, @Psychoslave wrote:

The ticket does overlap with T150417. But T150417 was created to go along the corresponding 2016 Community Wishlist Survey proposal.

This ticket try to have a broader scope, but if that was the general idea that was declined

The broader a scope the less fixable a task becomes.
This task mixes so many things that I miss criteria when exactly this task could be called "resolved". I'm tempted to decline this.
Once requested information has been gathered, specific 'actionable' tasks could be created if there is consensus and interest.

Someone in an other ticket (but I didn't retrieve it) seemed to tell that you can localize template magic words and parser functions of templates. Is that the case? If yes, where could I find documentation about it?

Regarding Javascript, there's babylscript. Could it be possibly be integrated within WM environment? If that would be conceivable, what would be the required steps for such an integration?

Regarding Lua, I already began some work for an Esperanto port, and I plan to target a more "internationalize the official interpreter" approach. As said in T148598, so far I worked with Lua 5.3. Would it make more sense to take into account a specific forked version used within WM environment, or should an approach like "focus on introduce that in the last official version, the WM's interpreter will soon or later be updated to follow the official Lua releases"? Or yet an other approach, as the official interpreter currently can't manage Unicode characters, while Luajit can, would it make sense to go with Luajit?

These all sound like very broad general questions to ask on a mailing list (like wikitech-l) or on IRC to me, before creating tasks.

Well, I don't know if there is any other programming language which is of heavy use with WM environment, do we have some statistics about programming language rates by the way?

We don't and I would not know how to create statistics for hundreds of different on-wiki venues and SVN and Git repositories in several places. From the top of my head: Python and others (Bots / Pywikibot); Native Mobile applications (Objective-C, Java); MediaWiki Core: (mostly PHP; also JavaScript, CSS, LESS and other languages); MediaWiki extensions (PHP and JS); User scripts and Gadgets (JavaScript etc.); Templates (Lua); MediaWiki Skins (CSS/LESS/HTML); External desktop applications for interacting with Wikimedia content like Huggle (C++) , AutoWikiBrowser (C#), WPCleaner (Java), ...; Offline applications like KiwiX (C++); Testing / QA (Ruby and Cucumber).

In T150933#2805747, @Aklapper wrote:

These all sound like very broad general questions to ask on a mailing list (like wikitech-l) or on IRC to me, before creating tasks.

@Psychoslave : Did you discuss the general topic on a mailing list? Link welcome. :)

Not yet. Would you have list(s) where you would advise me to discuss the topic?

https://lists.wikimedia.org/mailman/listinfo/wikitech-l and https://lists.wikimedia.org/mailman/listinfo/mediawiki-i18n would be my guess

AFAICS the broad general questions that I mentioned above have not been brought up on the mailing list but merely a description of this task?

Sorry if the message on mailing lists didn't match you expectation, I didn't thought you had question in mind when saying "What are the general questions that you are referring about?".

If you have explicit questions, whether general or specific, please let me know.

IMO , and I live in a country with poor English skills, and 9-5 my job is to build a tech community for Wikjmedia Indonesia, the 'build tools' are not a significant problem. In fact, the i18n of complex tools often causes problems, as the error messages are not as well maintained in non-English, and are less google-able. The focus is usually to provide error codes , and i18n-d documentation.

Our biggest problem is Phabricator is mostly? English UI. IRC servers use lots of custom messages in English only which apps cant translate. And the Translate extension is hard to use, and most pages on mediawiki.org are not translate, or the translations are very old.

Thank you @jayvdb for your precious feedback. Do you have more information on how 'build tools' internationalization can cause problems in you experience? If you have concrete example, it would be great.

I understand the difficulty with searching in online resources. With command line tools, LANG=C can help to fall back to log message in English when one want to copy/search some log message. Do you have cases where such a "fall back to unlocalized log message" are missing?

For Phabricator, I don't know if there are any translation ongoing effort. Translatable documentation is available on Mediawiki, I did it for French, but I'm afraid I can't help regarding Bahasa or any other Indonesian language.

As for the maintenance of translations, maybe some notification improvement might be investigated.

For Phabricator, I don't know if there are any translation ongoing effort.

--> T225 etc