[haiku-i18n] Re: Proposal: languages to enter (and leave) the repository
- From: Adrien Destugues <pulkomandy@xxxxxxxxx>
- To: haiku-i18n@xxxxxxxxxxxxx
- Date: Sun, 23 Nov 2014 09:45:14 +0100
On Sat, Nov 22, 2014 at 11:33:02PM +0100, Niels Sascha Reedijk wrote:
> Just to move back a little bit: I think our goal should be to ship
> maintained, complete translations with Haiku. This is because we want a
> good user-experience out of the box. This means that I think we should not
> ship unmaintained or very incomplete translations. [3]
To make this clear: the goal here is to try to not remove anything. So
if you are using one of the languages that are not included in the list
below, you are welcome to help making it active. For this you need
someone to act as the language manager. The work includes:
- Being subscribed to the haiku-i18n mailing list
- Validating/reviewing strings for that language (anyone can submit
suggestions, but the language manager(s) must review them before they
are included in Haiku).
To become language manager for a particular language, just send a
request to the haiku-i18n mailing list.
>
> However, I do think that it is good for exposure and for the validation of
> work to ship as many languages as possible with Haiku, especially during
> these pre-release phases.
>
> I have come up with two criteria
> * The strings have had to be updated in the last six months
> * At least 80% of the strings have to be translated
>
> Now what I propose is that the exported languages will be determined by
> reaching either of the two criteria. After the beta-period both criteria
> would be required to be part of the release.
I think there is a need for some flexibility and special casing. Some
translations (en_UK for example) are only variants of others (en in this
case which is actually en_US). This means they only need to translate
whatever is different from the "main" translation for the language, and
it is unlikely that they reach 80% translated strings.
There are languages which can be included even if they are incomplete. I
know people using Esperanto (on Linux, but that applies to us) and they
don't expect to have all apps translated. People using Esperanto can
always fallack to at least one other language, and our Locale Kit
handles that fine. Generally, this makes an incomplete translation
better than no translation at all, even for final releases. English is
not everyone's second language, and I can imagine that in some eastern
Europe countries, some people may appreciate a partial translation to
their native language + a fallback to Russian, for example.
So, what nielx suggests is a good base rule, but we can allow exceptions
if there is a good reason to do so. Please ask on haiku-i18n if you
think such an exception should be made for your language.
I think we have failed to include languages earlier in some cases. We
used to have an almost complete catalan translation, but never included
it. In such cases it is expected that people lose interest and give up.
So, if you are in such a situation, please just ask for inclusion of
your language here.
> One remark is about Na'vi, which shows the problems with applying just one
> of the criteria: it has only 1% of the strings translated, and it is
> doubtful that it actually works in Haiku since it might not be
> ISO-compliant, so I would suggest that after consulting with Jessica, we
> drop this language. (see how I can make the case without mentioning it is a
> language that doesn't exist?)
Actually ISO allows using the "art" (artificial) language code for such
cases. Na'vi would then be coded "art-nvi". We can include it, however,
it would not be available in the preference panels because ICU doesn't
know about it. It may still be usable by manually editing the
configuration files, I'm not even sure if that would work. In any case,
we need a more complete translation for this to be of any use.
> [3] We should, however, see whether we can deliver all languages as
> optional packages that can be installed by the users. That would require
> some work in the build system though.
This is a good idea. The language data uses quite a lot of space and
allowing to remove some languages from there to save space on an Haiku
install sounds good. It would be nice to have a ticket to track this :)
--
Adrien.
Other related posts:
