Personally I don’t care - i Am always set up to see only English, And anybody I work with is capable of doing the same. But I strongly empathise with Antoine’a sentiment.
P.S.: the Wittgenstein quote is taken wildly out of context, it is not about comprehending language but about how weak language is as a tool to communicating certain ideas.
Sorry for stating the obvious, but posters on this forum are by definition biased since they are fluent enough in English to read and write posts in that language. Many (future) users will use forums in their own languages instead.
Do whatever you want guys, I don’t care anymore, I’ll just do my own stuff.
I’ve encountered a fair number of websites that seem to interpret the browser language preference “en-us” as equivalent to “I didn’t know you could set a language preference, so I’ve just left it at default.” So they serve up a page in the language of the country that I happen to be in.
Even when a website does get the browser preference right, I might want to change the language of that specific website to get things like search to work better.
Also, it is sometimes necessary to see the original wording, because natural language translation is never 1-1. Hypothetically, if three different translated texts talk about “produced functions”, “rendered functions” and “generated functions”, respectively, then I might have no way of knowing if they are referring to the same thing unless I can see the original.
Sorry to update this thread but I feel it should be back on !
I have the same feel / experience as many of the posters of this thread.:
I am good enough in english that I read it without “mental fatigue” and I don’t mind it,
but also I agree with this other need of the community : education. And if we want to make Julia a frontrunner programming language for science, research & education, it needs to be translated.
At the minimum for the core language, the core documentation and core modules. Some external books (Thinking Julia) / websites are appearing in French (not following spanish, sorry) but very few.
Yes, the problem of most translation is it’s time consuming and not of immediate ROI.
But, Yes, just a small part of the world speak english fluently and this has to be addressed if willing to reach out to another crowd .
To compare, (if possible), looking at the python documentation, it’s available in English, French, Spanish, Portuguese, Japanese, Korean and Chinese (simplified and traditional). So it’s still missing indian and arabic at the least. (See [List of languages by number of native speaker (Wikipedia)] (https://en.wikipedia.org/wiki/List_of_languages_by_number_of_native_speaker)s ) for why I believe this).
Therefore, the unicode and LaTeX support should help and booster this.
My 2 cts
One thing I can say about this is that (maybe 4 years later from when this was originally posted) the automatic translation that GoogleChrome has built in, for Portuguese, is VERY good.
I do not think that without a dedicated, professional team only to translate and keep updated the page, we (native portuguese speakers doing something else as a main activity) could do better than that.
I cannot evaluate how good is this for other languages, but I guess that for the major ones it should be similar to that of Portuguese.
My two cents is that the reasonable path to providing translations in many languages is to have a builtin deep learning translation, which are really so good now. Something else does not make sense to me.
This is what one initial paragraph looks in the automatic translation to french (my french is not the greatest, mais je me débrouille, and it looks good to me:
L’installation de Julia est simple, que ce soit en utilisant des binaires précompilés ou en compilant à partir des sources. Téléchargez et installez Julia en suivant les instructions sur Download Julia .
Si vous venez à Julia depuis l’un des langages suivants, commencez par lire la section sur les différences notables par rapport à MATLAB , R , Python , C/C++ ou Common Lisp . Cela vous aidera à éviter certains pièges courants, car Julia diffère de ces langues de nombreuses manières subtiles.
Le moyen le plus simple d’apprendre et d’expérimenter avec Julia est de démarrer une session interactive (également connue sous le nom de boucle read-eval-print ou “REPL”) en double-cliquant sur l’exécutable Julia ou en l’exécutant à juliapartir de la ligne de commande :
Not the most idiomatic but perfectly understandable. I especially like the fact that it doesn’t try to translate everything (eg Common Lisp, read-eval-print).