Jean-Marc Lasgouttes wrote:
> It is more complicated now :)

It needs to :-(

> You could also have taken the approach of fixing the parsing of the
> relevant code.

I thought about this, but I think this is more error prone. Think someone uses 
the language code for something and is not aware of the variety (which is by 
no means standard, as opposed to the language tags, but only aspell 
convention, AFAICS).

> What shall we do if we want a translation for old german
> and one for new one? (does this even make sense?)

In theory yes ("Schluß" vs. "Schluss"). But I do not think anyone will set up 
a po file for old spelling. This would be purely cosmetic[1], whereas a spell 
checker that fails to check old spelling passages is really, really annoying.

> > PS. I know that WordLangTuple is a triple now, strictly speaking. But
> > I'd keep the naming for now.
> 
> I am not sure why this tuple does not contain a Language* directly.

Yes, indeed, this would make things much easier.

Jürgen

[1] Well, given the emotions that are bound to the old spelling, I think there 
are people who would massively contradict to this statement.

Reply via email to