> What is/was the aim behind defining an extra math.bind? To have the
> same math definitions in all languages? There are menu- (and therefore
> language-dependend) definitions in math.bind.....
The aim is to ease maintainence: With only one math file, the math bindings
are always synched. Earlier, the math bindings were duplicated in both the
cua.bind and emacs.bind file, and ever so often, they came out of synch. (This
is also the reason we have the menu.bind file.)
I realize that the bindings are language dependent. Get over it. ;-)
Some time during the 1.1.x development cycle, we will certainly abstract the
menus into some kind of configuration file, and thus allow users to customize
them if they want to. This will hopefully also clean up the area in respect to
the i18n situation.
And because the same thing will happen with the layout files, we need to aim
for a generic solution to the translation problem. It would be best if we in
some way could get this into the framework provided by gettext. This is
clearly an advantage for translaters, because it's much easier to administrate.
So this is an area worth spending some time on.
I know Jean-Marc, Joacim, and maybe others have discussed this issue on the
list earlier, but I admit that I didn't follow the thread closely. Maybe they
can sum up the discussion, and write a short design document about it?
I have a few ideas myself: I imagine that practically all configuration files
in LyX will be scripts in the future. So if we can translate the strings in
scripts in a general way under the gettext framework, we will largely have a
solution to this problem. (I.e. this implies that all configuration files only
exist in the English version. We would not have a de_menu.bind file in the
same sense as we have it today. Instead, all the relevant strings in the
configuration files would be part of the po-files. I feel this is important to
aim for, because otherwise we will for ever have the problem we have now: The
different translations are not synched.)
Greets,
Asger