[Scilab-Dev] Localization Scilab / Scipad

[François Vogel]

Enrico Segre said on 29/06/2008 10:42:
> For the first launch of scipad in a session, we could make provisions in
> scipad.sci for starting scipad with the desired language instead of
> reading it from the saved preference file.

Hmm, yes this is of course technically possible, but does this mean 
that the Scipad language would no longer be saved in the pref file any 
more? Currently the scheme is that Scipad starts in english the first 
time, then the user can change the language from Scipad, and this is a 
preference which is automatically saved across Scipad sessions. 
Switching languages is not an action the user has to do very often, 
I'm just fine with this. The normal user selects his language once and 
for all, and basta. The fact that he has to do it twice, i.e. in 
Scilab and also in Scipad, doesn't look that much of a PITA in my 
opinion. And there are advantages of this fact too (see below).

> However, please note that the current localization set of scilab is a
> small subset of that of scipad. Would really have sense to force the
> scipad language to be the default choice for scilab, when more choices
> are available (and the user might have already chosen one)? Imho the
> problem stays even if the scilab set grows, as long as the two language
> sets don't coincide.

I agree on this word by word.

> Perhaps what you want is that at first run of scilab (no scipad
> preference file existing yet) the default scipad language matches that
> of scilab?

I guess this is what Sylvestre was after. On this, OK, why not, 
instead of using the default english fallback we could use the Scilab 
language if the localization files exist in Scipad. We would have to 
map Scilab languages to Scipad languages, and have a fallback in case 
there is no match.

So yes, technically possible.

After the first launch, the scipad language should still be saved in 
the pref file. Why should the user have to switch Scipad language by 
switching Scilab language? (Note that this is not always possible, for 
instance Spanish in not available in Scilab). And why would we forbid 
different languages in Scilab and Scipad?


> At first sight, I'm not sure I understand the point - these examples
> just wrap procedure mc of package msgcat, which is exactly what we are
> using. What has gettext to do?

Ditto.


> ps: what about making the debugger work, first?

This is also my very point of view.

Scipad localization works fine. Honestly I do not see any point for 
changing it. Uniformisation with Scilab or whatnot is a very weak 
argument IMO.

We can fix later what is already working in Scipad, when major bugs 
like the broken debugger get fixed. What is the opteam progress on 
this? We made proposals and didn't read your thoughts on them.

Francois