Re: Internationalization of Lua

On Mon, Nov 21, 2016 at 7:52 PM, Coda Highland <chighland@gmail.com> wrote:

> It's mostly true. I just had my history confused. It wasn't BASIC I was thinking of (I was confusing VBA as Javier alluded to)

I do not think VBA was fully localised, I cannot find any evidence of that. They have tried to localise the API, but that is function and object names, i.e., identifiers; that's not going all the way. And they do not recommend using that, anyway [1].

> it was ALGOL-68 that was translated into Russian.

Indeed, ALGOL allowed different representations. But ALGOL and ALGOL-68 did not have much success except in research. I am not aware of any successful languages derived from them that would also allow multiple representations. That has probably been deemed too baroque and unnecessary.

Which I think proves once again that localisation does not work in real-world programming.

There is a far more profound observation: the mathematical language of equations is chiefly composed of Latin and Greek letters and some universal symbols. It has never been localised to Russian or Chinese, and very likely it never will. I remember attempts by some people to use Cyrillic letters in equations, that has invariably seemed strange and silly. This may be attributed to a tradition, but even then it exists and cannot be just written off. I suspect we may have a similar tradition in programming, in addition to all the other considerations.

Cheers,

[1] "However, due to potential language and COM interop issues it is strongly recommended that you use only English function names in your code."

https://msdn.microsoft.com/en-us/library/bb157877.aspx