Re: Menu translations (in Gnumeric and elsewhere)

[Danilo Segan <danilo gnome org>]

Hi Morten,

Yesterday at 22:41, Morten Welinder wrote:

> I added code to Gnumeric to check that menu keys (as displayed underlined
> in menus) are unique in all menus.  A quick check reveals that translators
> need to be much more aware of this problem.

As discussed before, this is not really fatal problem: Gtk+ handles
that gracefully.  If there's more than one shortcut for the different
menu item, Gtk+ jumps to the first one when key is pressed, but
doesn't activate it.  When that shortcut key is pressed again, it
jumps to the next menu item without activating it, and so forth.
So, while this is a usability issue, it's not too bad
(i.e. applications are still sort of usable).


> If you want to lift the code for other projects, it is in function
> check_underlines in file gnumeric/src/wbc-gtk.c

With that aside, I'm glad to see this code.  It really rocks, so one
big THANK YOU for working on it.  I'd love to see it used in more
places.  Have you considered letting other developers know about it
perhaps via d-d-l?  I didn't look at the code, but is it perhaps
something that belongs in egg?

Other thing I'm thinking on working on when I find some time is
processing Glade files themselves to check for conflicting
accelerators (it would be trivial if it was not for Gtk+ and lib*
"stock" items).

> Translators: please locate your language below and see if you can make things
> unique.  Apologies for the question marks here -- my terminal in not in UTF-8
> mode.

Please try to use UTF-8 terminal (in gnome-terminal or xterm, it may
be sufficient to use UTF-8 locale, eg. by setting "export
LC_ALL=en_US.UTF-8"): many translators won't be able to see what
messages need to be fixed (eg. those using exclusively Cyrillic
script, like Bulgarian or Russian, or Greek).  It's too much to
require of each translator to compile Gnumeric from CVS, and my guess
is that we're talking about HEAD here (not gnumeric-1-2 branch).  
We (of the Serbian Translation Team) are sort of lucky since Serbian 
Latin script can be used to determine what needs fixing in the
original Serbian Cyrillic translation.  Still, others will appreciate
if you rerun these checks with correct output.

Still, thank you very much for working on this.

Cheers,
Danilo