I kind of miss the early 2000's, where a bunch of pretty well-known French and Belgian academics were publishing source code in their own languages (#define to remap keywords and all)
I am a native Swedish speaker and I have had a hard policy since I was a teenager to have all electronics set to English and not Swedish. Why? Because the translations are lossy at best. Often you need to understand both English and Swedish AND have a creative mind for word games to understand what the hell something means.
Example: My daughter had an Android smart watch. One menu option in settings was "omkring". This makes no sense. But if you translate it to English it can become "around" hmm, still nonsense.. oh, but it can mean "about" as in "around and about".
Translation is a fools game. It builds Babels Tower all over again. It's bad enough that we have multiple programming languages that only sometimes interop.
That's not some immutable law of the universe - it's really a pretty recent state of affairs. A quarter century at most
And? Powered flight is barely 100 years old. For most of the world, ubiquitous sanitation, access to running hot water, electricity, medicine beyond voodoo and superstition - are all barely 100 years old, give or take quarter century. For most of the world - the rest of the world got it much later, or in some places, not at all.
Very little in human affairs is an immutable laws of the universe. Most of it is arbitrary, path dependent, frozen in place as we built more things and made more choices on top of past ones, instead of endlessly bickering about which random result would've been most fairest.
It so happened that English became lingua franca, and we've built the last 70+ years of technology on top of it. Do you want to tear it all back down, and call for a World War II rematch, all for the sake of a dice re-roll on which language we name our "for loops" in? Why the hell does it matter?! English won. Time to move on. Build up, instead of tearing down.
EDIT: also, FWIW, languages evolve. English of 2025 isn't all the same as English of 1945, much less that of 1865. That evolution accelerates with globalization, as every culture contributes their bit to the global whole. See the "Internet slang" or various pidgins that popped up in the high-throughput port areas around the planet; this is the sign of things to come.
When I started programming in 1980 there wasn't even a hint that anyone used anything other than English. The first non-English thing I saw was a message from a Prolog compiler (in perhaps 1983) telling me that I had an "erreur syntactique", IIRC.
More significantly, from 1988 to 1991 I worked on a EU Esprit research project, with Dutch, Italian and French partners. They all coded exclusively in English, even the French partners.
YMMV.
There have been many lingua francas in history, often domain specific. In the early 20th century, French and German had become the languages of international scientific discourse. Before that, Latin was the language for academic work and international communication in Europe. If you wanted to get involved, you had to learn the language. That's always the case, because a common enterprise needs a common language.
And today, the degree of international collaboration is so high, and the cumulative literature so great, that it will be more difficult to even make a shift.
(And trust me: programming language keywords in a language other than your own are not an obstacle for anyone.)
The base reason is that Scandinavian teenagers can learn English by sheer osmosis from Anglo culture. (That is not the case in many other places.) Then it makes perfect sense to:
> Because the translations are lossy at best.
Since (my reason at least) is that I can troubleshoot problems using the English Internet corpus. And I don’t need to bother with multiple terms.
Granted I was older before I committed to this practice.
Practical angle: so let's add another datapoint! The same is the case in Poland.
Seriously though, it's just a consequence of history. Computers and software were imported to everywhere around the world rapidly; no country wanted to put themselves at economic disadvantage by blocking sales of hardware and software until they 100% up to specs regarding consumer products being fully explained and operable in native language (in fact, I suspect such regulations came about after computers became a thing in any given place).
> Since (my reason at least) is that I can troubleshoot problems using the English Internet corpus. And I don’t need to bother with multiple terms.
For me it's more of the latter. Translations love to screw terminology up, introducing multiple distinct terms for the same abstract concept, and using them inconsistently (often the result of multiple, independent, half-assed passes at translating the user-facing text, manuals, etc.).
English was the language software authors most likely spoke; English is the language of programming and software industry. English is the language almost all software targets first and foremost, due to market/user population size. As a consequence, most care is put into English in the user interface, it's the one considered canonical, it's the one in which the thing was designed (whether the desigers were native English speakers or not); English is the one where user-facing text is most likely to be consistent and corresponding 1:1 with language used in other programs, as well as with code powering it, etc.
Every other language is a translation, most likely done as an afterthought, as cheaply and quickly as possible (there's little incentives to do otherwise - as long as English text is done well, users will manage). If you understand English, running software in anything but English is setting yourself up for failure.
Similar tendencies can be seen in other European areas. But it seems to be less pronounced as you move outside of the Germanic-speaking parts of Northern Europe.
You don’t have to move far away from the Germanic-speaking Northern Europe to find adults who struggle with English. Including young adults.
Now this is apparently for kids. Certainly kids younger than teenagers have amazing language acquisition abilities. But they also might have less motivation and exposure to English. Now maybe Northern Europe kids have such exposure these days that the current English-dominated programming as we all know it is a mere triviality for them. So scratch those. Not for them. I would still be surprised if there aren’t a large number of kids worldwide that localized programming would help.
[1] Minorities like the Sami could be exceptions here.
As for the ambiguous terminology, this is where the quality of your translation team really shows. When Windows was first translated to Slovene, for example, the team took special care to find the correct terminology (sometimes inventing phrases along the way) and use it consistently. Again, not something machine translations or some random guy for $5/hr on Fiverr would do.
The second is that this all depends on where you're from and what's your goal when using a computer. If you're doing it to become a programmer, then yes, there's no way around English. But for classroom use, I'd argue a localized approach is much better because it directly exposes the "tone" of programming to the students even if they don't intend on ever doing it again in their life. Otherwise they're just learning "FOR x IN...DO..." to pass an exam and not really thinking what it actually means. Again, here I have the "non-coder" kids in mind. You also have to keep in mind that the further you go from Germanic languages, the weaker the similarities to your language become.
OCaml was written in 1996 by Xavier Leroy, Jérôme Vouillon, Damien Doligez, and Didier Rémy at INRIA in France
Learning language - fine, use whatever language you want
Excel - spreadsheet logic breaking when you open it with a different locale - horrible