From India. Registered on Pootle. Saw some body had already taken up the translation into Hindi language
So, I want to translate Haiku into my native language “Kannada”. But, this language is not shown in the drop down list. How to add it?
Also, since support for Indic languages is not yet available in Haiku, in what script should the translated matter be written? How will it be converted back to the targetted language at a later date?
I see that on the “localization mailing list” only European, Chinese, Japanese, Indoneasian languages are mentioned…But Hindi language is not mentioned, though it is being actively translated…
I may be wrong but I think that an admin has to create the language and name someone admin for this language. That person will be responsible to review and to validate other translators contributions in the language.
So for the starter it is a bit intimidating and if you add to that complexity of Indic languages maybe none stepped in.
Each language or variant is important. It is just a problem of manpower. To translate a language correctly, you need at least few persons that are bilingual and have a minimum of computer knowledge.
I would open a ticket for enhancement on Haiku bug tracker but before that perhaps could you open a discussion about how Kannada should be written. This would also allow you to see if there are other interested who could assist you in that task.
This is an open source project. Things are important to the people contributing to it. If you want to contribute a translation, we can arrange for that language to be added to Pootle. How you enter the translations there is up to you. I have no idea if the font necessary to display the language’s script is working on Haiku or if special input methods are needed etc.
It may be that Haiku isn’t considered important by Hindi speaking developers.
Thanks to history, a large proportion of educated Indians speak English. It is, after all, the language of the courts in India, and of proceedings in Parliament. Although I haven’t looked it up, I would expect that the proportion of English speakers in India is significantly higher than it is in the other Asian countries you mention.
This means that many people (perhaps most) who buy a computer in Karnataka (the state which uses Kannada) are likely to speak English in addition to Kannada. And therefore there is less pressure to provide translations into Kannada (or indeed any Indian language) than would be the case for most Asian languages other than India.
This is not to say that translations should not be provided. However, there is likely to be little serious demand, and there may not be enough translators to make the exercise practicable.
As usual with languages, things are a bit complicated.
On one side, there is the translation. This “only” needs someone with knowledge of the language and of english as a source language. In the case of some languages, we had a single person do the entire translation over a few days or a few weeks. So, this is the easy part.
Then there is text rendering. In Haiku we have good support for latin, cyrillic and greek alphabets. These all work pretty puch the same and the writing system was somewhat simplified to fit the constraints of printing press and then early computers. Text goes left to right, you can get away with no ligatures, one character pretty much always maps to one glyph.
Other languages such as arabic, chinese, japanese, korean, hindi, use different scripts that use different rules. Sometimes more complicated, sometimes simply different. This will often need to adjust some things in the code. This work has so far largely not been done because you need someone who understand the language enough to know what needs to be done, speaks english or another common language with other members of the development team, and also understands a little bit about the software stack (freetype, harfbuzz, etc) needed to make things happen. So far it’s been quite difficult to find all 3 in the same person, or at least get the involved people to work together.
Finally regarding the mailing lists: the main i18n mailing list is for coordination between the different languages. The main topics there will be:
Timelines for the releases, so the translators know when to get their last minute translation fixes in,
Announcement of changes of language manager for a language
Possibly asking for technical questions when a string cannot be translated correctly due to limitation of the translation system (handling complex plural rules, for example, or string manipulations done in the code that don’t work due to a special case in some language)
Discussions for the work on one specific language happens usually on the corresponging mailing list for that language. There, the translation team for each language has more freedom on how they organize themselves (maybe they don’t like mailing lists or want to use some other tool for coordination. Maybe it’s a single person and a mailing list is not needed). The translation teams usually also get a wiki page on the trac wiki where they can explain their own process in their own language. The result is that each team organization is a bit different and sometimes there is little contact with other parts of the work in Haiku (especially if things are working well, there’s no need for any discussion).
I think this explains why the main mailing list is so quiet. But I would welcome more discussions, as I assume there are some translation problems and they just don’t always get reported. Being so fluent in english myself, I will sometime not notice missing translations at all in the french localization. Culturally, we’re quite used to mixing in english words in french and so it often does not create that much of a problem. This can be very different in other languages.
