Is localization welcome?

Cdc

Localisation would definitely help GrapheneOS adoption. I believe that nowadays it could be piece of cake using even proprietary AI model for GOS translations and such. I am also thinking about isn't there a way to use an official translation and add "a few" missing parts? (Just a thought)

DeletedUser370

Cdc I am also thinking about isn't there a way to use an official translation and add "a few" missing parts? (Just a thought)

The translations for the texts that are imported from AOSP are already present in GrapheneOS. They wouldn't need to be retranslated. They could be retranslated but that seems like enormous work for little benefit, unless the original translations are of poor quality.

VEECCdYKYi

DeletedUser370
But there are a few strings that are untranslated. Such as the exploit protection settings. Translation if these strings would be welcome.

DeletedUser370

VEECCdYKYi But there are a few strings that are untranslated. Such as the exploit protection settings. Translation if these strings would be welcome.

I was only pointing out that because the majority of texts in the user interface are imported from AOSP, those specific texts won't need to be translated again. I wasn't sure the user I quoted understood that so that's why I noted it. I wasn't arguing against translating GrapheneOS-specific strings.

LeslieFH

matchboxbananasynergy

I really don't see the problem with "half of the [GOS] strings are translated and half are not tranlated yet [while all AOSP strings are translated]", because it's still better than "none of the [GOS] strings are translated [and all AOSP strings are translated]".

The lack of translation of GOS strings is a significant accessibility issue. It limits the use of GOS to people who are able to speak English, which is not typical for Android. I always have my Android set to English, but I did not suspect a major Android project would forego any attempts at internalisation.

This is really disappointing, especially since the amount of text to be translated for GOS-only features is not that large (I'm a professional translator by trade), it's only the amount of languages that might be a problem.

de0u

LeslieFH I really don't see the problem with "half of the [GOS] strings are translated and half are not tranlated yet [while all AOSP strings are translated]", because it's still better than "none of the [GOS] strings are translated [and all AOSP strings are translated]".

OK. But if the developers don't want to go that route then the options seem to be working with them on their approach or forking the project.

Meanwhile there is another issue, which is that misleading translations could be worse than none.

LeslieFH This is really disappointing, especially since the amount of text to be translated for GOS-only features is not that large (I'm a professional translator by trade), it's only the amount of languages that might be a problem.

It might be very helpful if a professional translator joined the team. Project leaders have said what they see as the most plausible way forward: MetropleX strcat . Speaking completely hypothetically, might you be in a position to volunteer 10 hours per week to manage translation? If so, mentioning that to the project leadership might move things forward.

LeslieFH

de0u

It's hard for me to see the possibility of misleading translation that would be worse than users feeding the text they want to understand into bullshit machines. And then, there's also the risk of people misunderstanding English because their English is weak, but they're not aware of it.

And sure, I could volunteer for Polish translations.

Also, I think "10 hours per week" is significantly overstating the amount of work needed for localisation of GOS, 10 hours of translating per week would get you about 25000 words translated per month, and GOS is not a book ;-)

de0u

LeslieFH It's hard for me to see the possibility of misleading translation that would be worse than users feeding the text they want to understand into bullshit machines.

Perhaps... but... I can't help recalling a certain Monty Python skit about a traveler's phrasebook that was wrong to the point of causing legal trouble.

LeslieFH And sure, I could volunteer for Polish translations.

I suspect (though I don't know this) that the project may not start off with translations for specific languages before setting up a management framework for ensuring that multiple languages can be kept up to date in a timely fashion.

LeslieFH Also, I think "10 hours per week" is significantly overstating the amount of work needed for localisation of GOS, 10 hours of translating per week would get you about 25000 words translated per month, and GOS is not a book ;-)

Indeed GrapheneOS is not a book. There are often two or three releases per month -- features written by GrapheneOS developers and new code that pops up from Google via AOSP. So I suspect there would be a steady stream of translation work to be done. Obviously to start off with there would be a big backlog, but I do expect there woudl be some work roughly weekly. And I can imagine that coordinating multiple translators, to make sure that translations were done and checked in time for releases, might take something like 10 hours per week.

« Previous Page