[traceroute66]

> It wouldn't be bad at all if programming languages were localizable

That is an admirable dream to have, and one that I have sometimes had myself. Unfortunately, as with most things reality soon puts an end to it.

First, the dream is only really a dream you can have with the luxury of a Western language. It is unlikely to be viable once you start looking further East in the direction of e.g. Kartvelian languages (i.e. low number of speakers worldwide and even fewer with relevant specialisms in programming to write compilers) or Asian languages (e.g. Japanese and its particles and other complexities).

Second, when you are writing code you don't really want to have to worry about the possibility of debugging your compiler. Sticking to English means the likelyhood of a more robust compiler because of more widespread adoption. You also remove technicalities such as UTF-8 or whatever else that might cause indirect issues.

Its a bit like science really. Sure you can write your papers in your native language, but the reality is if you're serious about your work being discovered and are interested in being offered bigger and better jobs, you'll want to publish in English sooner rather than later.

[pyrale]

> Unfortunately, as with most things reality soon puts an end to it.

Usually, when a thing gets labeled as a dream, people don't try very hard either.

> Sticking to English means the likelyhood of a more robust compiler because of more widespread adoption.

The comment you quote specifically talks about localizable software, not having one compiler per language. When you I18n a website, you don't write n sites, for instance.

[traceroute66]

> Usually, when a thing gets labeled as a dream, people don't try very hard either.

Some things are hard, but worth trying.

Other things are a dream because the reality is they are unachievable either due to the realities of practical constraints or because the amount of work involved would be hugely disproportionate to the benefit.

For example, I mentioned Kartvelian languages. How about we specifically look at Laz which according to Wikipedia at the last count in 1980 had only 22,000 native speakers.

Writing a Laz compiler ? I'd say if we're being totally honest its a dream, not just "hard work".

[pyrale]

> the amount of work involved would be hugely disproportionate to the benefit.

And you can trust english speakers to properly assess the benefits of I18N, right?

> Writing a Laz compiler ? I'd say if we're being totally honest its a dream, not just "hard work".

As I said in the message you quote, internationalizing a compiler is not the same as rewriting it. If you consider that making languages accessible to other languages means each natural language should have its own programming language and compiler, of course it's going to be hard.

But that's moving the goalposts.

First off, you don't need to port a programming language to every natural language. Some smaller goals would already help. For instance, making it easy for third parties to i18n the language and libraries would be a significant enabler.

[JadeNB]

> Writing a Laz compiler ? I'd say if we're being totally honest its a dream, not just "hard work".

This seems like letting the perfect be the enemy of the good. "We can't possibly localize for every possible language ever" is not much of an argument for why localization itself isn't worth it.

[temac]

IIRC, TI Basic dynamically translated keywords (and probably function names) depending on the current langage.

That's less interesting if you have lots of libraries if you don't translate them.