From: Daniel Stenberg <daniel_at_haxx.se>
Date: Thu, 19 Sep 2002 09:47:49 +0200 (MET DST)

On Thu, 19 Sep 2002, Florian Mösch wrote:

> > Or more simply add it in the temporary source:
> > << _msgstr_(MSGID_... ,"Hello World\n") >>
> > or
> > << /* _lang_("Hello World\n") */ _msgstr_(MSGID_...) >>
> > as a substitution ?
> > That way, you will be able to see which english message is associated
> > even when debugging source. I don't really see where the problem is.
>
> This is a very nice idea I think. And I cannot see any reason why Daniel
> seems to simply ignore the ideas.

I don't ignore it, and I refuse to accept your remarks saying so over and
over again. I repeatedly respond and tell you *why* I don't like the ideas.
If you don't read those mails or don't don't agree with my reasons, well that
might not make you happy but it sure doesn't make me having ignored them. I'm
against this custom preprocessing of the C code done this way for this
purpose. Be it simple or be it advanced. I want plain and normal C passed to
the compiler. I can't see why you can't instead just use a plain comment in
the source code.

That's my opinion. Others may think differently.

This is not ignoring. This is declining to use that idea, since I don't agree
the required hack would be a nice one. Thanks, but no thanks.

Yes, the code is slightly less easy to read now since the full text string is
not present in the code, but I've never seen you doing much fiddling in the
Rockbox source code anyway so why is this such a big issue to you? All the
major contributors seem to accept/approve of it.

> Thi sdiscussion is abrupted by his announcment and call for translations.
> Sorry for that.

You confuse me and my brother. Björn was the one who called for translators.
(Not that it matters who made it because both of us tend to agree on this
issue.)

This discussion could've ended much earlier with a phrase similar to: "show
me the code". You've repeatedly complained on me and the methods I've
suggested, but contrary to you, I've written heaps of code to back my
suggestions up and voila, now my suggested approach works, builds and runs.
What did you do in the mean time to show us the greatness of your method?

Personally, I'll trade in *any* Rockbox solution for a better one when such a
one is offered. Better of course being very subjective.

Please bring it on into the light and we'll evaluate your approach.

> > Still, it might be nice to have a script that searches the code for
> > character strings, just to find non-translated strings.
>
> Absolutely. Thats what I suggested. But Daniel doesn't seem to like the
> idea.

Totally wrong, and please don't put words in my mouth. If you have tools that
extract existing strings to translate from source code, then please use it.
It could be a cool tool. I guess not as usable now though, as most of the
strings already are extracted, fixed and subject for translation.

A tool like that doesn't need to imply a certain file format.

> And I won't write a script that isn't fully integrated in the translation
> process.

If the script is usable and popular, we could indeed use it. But until you
show us how it works and its benefits, we can't judge that.

Until you back your complaints up, you're just a whiner.

Show me the code.

-- 
Daniel "Bagder" Stenberg -- http://rockbox.haxx.se/

Received on 2002-09-19