I purged the page and it shows as fully translated now (also at ): same for you? Maybe something happened with one optional message (which is counted when translated, but otherwise should not count for statistics).
Last edit: 07:06, 3 October 2016
I think we should confirm something here, per Translatewiki.net languages#Enabling a new language for localisation at translatewiki.net:
- A language code from ISO 639-3 or the Ethnologue website (for example 'de'): hoc
- The language name in English and in its own language (for example 'German – Deutsch (de)'):
- In English: Ho
- (And what's the autonym of it?)
- Is the language RTL or LTR (written from right to left or left to right): Clearly LTR
- If the language is commonly written in more than one script, please say in which script the language should be enabled...
- Devanagari is always OKay as a number of modern OSes & browsers are having its support, not sure if Warang Citi (ISO 15924 Wara) could also be added or not, do you have a font for Wara?
- (For MediaWiki) Should the language fall back to another language other than English? If so which?
2. Autonym: 𑢹𑣉𑣉 (hoc)
4. It shall be enabled in Warang Chiti script. We have fonts for Warang Chiti.
5. The fallback language shall be English.
By this answer of 4th question, is this means that only Wara should be used? Or should be 1st Wara then 2nd Deva? This is so important because I've judged incubator:Wp/hoc, many contents are still Devanagari? Things should be decided here now, as changing it after decision, even possible, could be hard due to Wikidata compatibility (IMHO).
That is because there is no available input method for Warang Chiti. Most Ho language books are printed in Warang Chiti. Nowadays it is quite rare to see Ho language printed in Devanagri or any other script. Maybe we should wait till we have proper input methods. Also, Devanagri is seen as a foreign script by the Ho people because it contains so many sounds that are not present in Ho and lacks the most common sounds of Ho language.
Here's what ISO/Ethnologue writes: Devanagari script [Deva], used in Bihar. Latin script [Latn]. Oriya (Odia) script [Orya], used in Odisha. Telugu script [Telu]. Warang Citi script [Wara], no longer in use.
Given that the Warang Citi script is no longer in use, creating a user interface in it, is useless.
Which living script(s) will you be localising in?
The information in Ethnologue about this issue may be incorrect.
This writing system is encoded in Unicode, and the proposal for Ho encoding, written by Michael Everson, says that it is in use. According to that document, Warang Citi is in use.
Hello. I'd like to request: 1. to add my language code 'gur' ISO 639-3 2. to get "translator" rights for my user account.
Done Have fun. I somehow missed this request, so it took a few days longer than possibly expected.
Is it possible to have the list of items to translate or to review, presented in a random order ? It appears that the order is always the same. May be there is a reason. But presenting the texts in a random order improves the covering of the operations and brings surprises in the translated subjects making them atractive.
I can imagine some people would like to do it in random order, especially for review. For myself it helps to do it in order, so I don't need check context for every message. For example two different software could use different wording for some term, and inconsistent translations would be hard(er) to notice if related messages did not appear together.
Christian, thanks for the idea. Do you have some specific case in mind, which makes it bad for you to be shown always the same messages? Maybe some especially "boring" messages in some message group you monitor? Because of course you're going to see again only the messages you decide not to translate/review, no?
Oh no specific example in mind but ... better : could it be a Yes/No option for the person who translates or rereads: yes I want a random order,or no I prefer to follow the original flow of messages to reinforce the coherence of the translation for groups of messages.
Wouldn't be nice to have the option to add the messages one _reviews_ to his watchlist? Probably in [Special:Preferences#mw-prefsection-watchlist|Advanced options]]. Where could I request this feature?
Does mw:Internationalisation wishlist 2017#Activity reporting and engagement express this need reasonably well?
Yes to this "Translators should be able to stay on top of new translation work easily". But the example given in the tracking task T159642 (duplicate of T62904) talks about watchlisting entire message groups. It might help for some messages groups that I care of. But I sometimes translate or review random messages. I may not have a big interest in some projects, but I do in the messages I translate/review.
I was asking for the reviewing part.
We should probably automatically watchlist all the messages reviewed by the user who has selected "Add pages and files I edit to my watchlist" in the preferences.
Last edit: 11:25, 25 May 2017
Hi. Some months ago Apertium machine translation stopped working for Asturian language.
Is there something I need to do to have it again? It was really useful to speed up our translations.
I reported a few days ago that Apertium is broken for everyone: phabricator:T165949. The issue may already be fixed in the code, but translatewiki.net was not upgraded yet, I think.
I missed your report in phabricator, I hope it all will be fixed soon.
Apertium works now. If you were seeing empty suggestions, this was the reason. If it is not appearing it could be
- Apertium service temporarily unavailable
- Missing Spanish translation (if I remember correctly Apertium only does Spanish to Asturian)
"Traditional Mongolian Script" ([w:https://en.wikipedia.org/wiki/Mongolian_script] ) are used in mongolia and Inner Mongolia of China to this day to write Mongolian , which based on this classical vertical script. In Mongolia, the Cyrillic script writing system is used. Until 1990, only elders used to read and write the Mongolian script. But in the 1990s, secondary schools began to intensify the teaching of the Mongolian script.In China, the ethnic Mongol population was nearly 6 million according to a 2010 census. Many of them still receive nine-year compulsory education in Mongolian using the traditional Mongolian script.Mongolians living in Inner Mongolia are not familiar with the Cyrillic script and use the Mongolian script writing system only. They are unable to read Mongolian Wikipedia in Cyrillic. Data: A valid ISO 639-1 or 639-3 language code:mon  Unicode range:  The language name in English and in its own language: mongolian -ᠮᠤᠨᠭᠭᠤᠯ ᠬᠡᠯᠡ writing-mode: vertical-lr. Now, Internet Explorer, Safari, and Chrome support CSS writing-mode: vertical-lr. Default should still be "mon" in Latin alphabet. Please support "Traditional Mongolian Script".
Hello, following this phabricator entry, I would like to know what action are now required to make advance the project of adding phabricator translation on translatewiki.net.
Are you seeing less 504 errors, recently? I think I see less errors in the logs.
Yes, less errors since I reported back in March 2015. Anyway, I just tried both links and only Special:Watchlist gave me 504.
Special:MainPage doesn't work at all with MobileFrontend, though.
Is there any way to make MobileFrontend opt-in through the preferences (for testing)? Or can I get the source code of Special:MainPage to test on my own wiki? (Currently there’s no link on Special:Version, please link to its Git repo there if possible.)
Please amend When using Firefox at computers other than my own, I will type in "translatewiki + Shift + Enter" which the browser then redirects to http://www.translatewiki.net/ but that URI can't be found. Although the days of a "www" subdomain are long ago, it's still worth maintaining these to redirect to the proper https://translatewiki.net/
What version of Firefox is this? I've never seen such a behaviour.
Every version since at least 2.x has had this. Have you tried it?
DNS/Apache config needed? Who has the ops access to twn server?
I am not convinced this is worth the hurdles of adding DNS records and redirects and making sure HTTPS is not broken.
If you think that adding an HTTP redirect is too difficult then that is definitely a problem. I'm 100% certain that there are incoming links which include a www subdomain so if you want incoming viewers to think the site is still functional, I would recommend adding it.
Redirects are hard nowadays, yes. See for instance phabricator:T133548. Closed.
Well, several dozen of them certainly are. Much harder than one to be sure. I don't see how no response--not even a 404 page--is preferable but that's where that incoming traffic will go: nowhere http://isup.me/www.translatewiki.net
This user massively uploads terrible machine translations, including to Russian language (which is stated as his "native language" on his userpage), see User talk:Irus and (for Russian-speaking users) . All his edits must be rollbacked, his new pages (Wikimedia:Wikiedudashboard-articles.none/ru) must be deleted. His messages already displayed on Russian Wikipedia special pages, it's a clear vandalism. (If this page is not administrator's noticeboard of this project, please, move this thread to that page).
User:Nemo bis, what about reverting and deleting? I can do first, but I don't have rollbacker rights, can you grant it to me?
Deleting doesn't have any effect, you'd better retranslate any newly translated message which you think it worse than nothing. As for rollbacks, are you sure everything needs to be rolled back? How many messages did you check? I can make you sysop temporarily, but tell me how much time you need.
Until all his new messages are re-translated, English versions are much better then Irus's versions. I sure that many of his edits need to be rolled back, maybe all /ru-edits. I check many of his /ru-edits and almost all of this is nightmare horror. I think this weekend will be sufficient.
Ok. If you can't check every single translation before rolling back, try at least to sample them. Sometimes, or in some message groups, there might have been some logic.
Deletions don't have any effect because FuzzyBot will reimport the exported translations on the very next day. One would need to first remove the translations with a patch to the MediaWiki extension or other project. It's vastly faster to retranslate.
Could webm be added as a file format to be displayed? I linked a file used on Wikimedia Commons on the page Translating:Proofreading the messages but it does not display as a video, only as a link. And even on the file page no preview of the file is available. I guess this is just a configuration issue, and adding webm in the right place is enough.
Could you send a patch against the gerrit translatewiki repo?
Unfortubately not, since I do not know exactly which parameter need to be changed.