Should the names like "Rust" be left as is or should they be translated into language but with the same pronounciation?

[abhik-99]

So I am trying to contribute to Bengali localization and on a brief discussion with @mirwasi and @mgeisler this question arose. This rings especially true when the language does not have Latin characters ( like Chinese, Japanese and Bengali). Would love a discussion on this so that we can reach a consensus and avoid future translation confusion. Example of such words would be "Rust", "Chromium", "Google" and so on.

[mirwasi]

Thanks for opening this discussion @abhik-99! I feel we need to come up with a framework (taking everyone's opinions into account).

My 2¢ (Edit: wrt Bengali):

When it comes to names, I'd follow how the existing programming literature is written in that language. Personally, I'm a native Bengali speaker from Bangladesh and almost all of the programming literature there use the transliteration of the words in Bengali (for example: পাইথন (Python) or জাভা (Java). You can see a selection of the books available in Bangladesh.

Bengali is the second largest language in India (in terms of first speakers), with a significant number of speakers. I haven't read any Bengali programming book from India myself. I was looking at Amazon India for programming related books in Bengali, but couldn't find any. @abhik-99, it would be great if you could share how programming literature in written in Bengali in India.

If a word the name of a library, or a keyword from a programming language, library, I'd try to use the term transliterated in Bengali, while keeping the original term in brackets (I'll admit, I haven't been able to follow this all the time in code reviews)

Here is a screenshot of the TOC of a very popular Python book: https://www.rokomari.com/book/224822 (No copyright infringement intended)

Slightly off-topic (on translation): the lack of having a standard list of programming terms in Bengali is huge problem. Dictionary meanings are often misleading and fail to convey the true meaning. Some people in my country tried creating a list of Bengali terms, but it never got the level of adoption they hoped for.

cc: @raselmandol @sudiptachatterjee

[mingyc]

This rings especially true when the language does not have Latin characters ( like Chinese, Japanese and Bengali).

Please note that different language/region users have different conventions. As of this topic, it does not directly correlate to whether the language use Latin characters or not. For example, it's very common in zh-TW (Traditional Chinese used in Taiwan) to use untranslated English terms, especially in the field of computer science; while in zh-CN, it's entirely different. In Japanese, terms are usually translated, but sometimes English terms/abbreviations are still used.

Thanks for opening this discussion @abhik-99! I feel we need to come up with a framework (taking everyone's opinions into account).

There should not be such framework. Translation should also be localized.

[mirwasi]

Thanks @mingyc for your reply.

Sorry, should've been a bit clearer in my response: what I wrote was with respect to Bengali language. I agree with your general viewpoint.

[mgeisler]

Would love a discussion on this so that we can reach a consensus and avoid future translation confusion. Example of such words would be "Rust", "Chromium", "Google" and so on.

I'm not an expert on what the conventions is around the world — and I agree with @mingyc that it's likely different from language to language.

So the only advice I can give is for each translation to come up with guidelines on this and document them. The documentation could be in the relevant GitHub issue, it could be in a comment in the PO file itself.

The full set of five courses is about 350 pages if you print them... that's a substantial amount of text and consistency is important for a quality translation. Documentation is thus important as well to make sure the translators are roughly on the same page 😄