I don't think we use Q_ anywhere in PackageKit... do you mean do something like this? That seems like a lot of work...

diff --git a/client/pk-console.c b/client/pk-console.c
index 1d8df81..add804a 100644
--- a/client/pk-console.c
+++ b/client/pk-console.c
@@ -1930,8 +1930,7 @@ main (int argc, char *argv[])
 
                } else if (strcmp (value, "group") == 0) {
                        if (details == NULL) {
-                               /* TRANSLATORS: the user needs to provide a search term */
-                               error = g_error_new (1, 0, "%s", _("A search term is required"));
+                               error = g_error_new (1, 0, "%s", C_("A search term is required", "the user needs to provide a search term"));
                                goto out;
                        }
                        ret = pk_client_search_group (client_async, filters, details, &error);

(In reply to comment #1)
> I don't think we use Q_ anywhere in PackageKit... do you mean do something like
> this? That seems like a lot of work...
> 

Sure it's a lot of work, there are +700 strings and +500 of them have comments, but it's quite useless right now if you are not translating with plain text editors (like gedit, vim, emacs and so on) since everything that is not msgid or msgstr is visually stripped in editors like Gtranslator, Kbabel, Poedit ...

Adding the msgctxt ensures that everyone, no matter which editor uses will see your comments and thus will be able to translate properly your app.

Bones,

Tenim un concert a mitja tarda i jo a més haig d'anar a votar a Osona :S

Salut!

El dv 29 de 05 de 2009 a les 07:16 -0700, en/na
bugzilla-daemon@freedesktop.org va escriure:
> http://bugs.freedesktop.org/show_bug.cgi?id=21995
> 
> 
> 
> 
> 
> --- Comment #1 from Richard Hughes <richard@hughsie.com>  2009-05-29 07:16:13 PST ---
> I don't think we use Q_ anywhere in PackageKit... do you mean do something like
> this? That seems like a lot of work...
> 
> diff --git a/client/pk-console.c b/client/pk-console.c
> index 1d8df81..add804a 100644
> --- a/client/pk-console.c
> +++ b/client/pk-console.c
> @@ -1930,8 +1930,7 @@ main (int argc, char *argv[])
> 
>                 } else if (strcmp (value, "group") == 0) {
>                         if (details == NULL) {
> -                               /* TRANSLATORS: the user needs to provide a
> search term */
> -                               error = g_error_new (1, 0, "%s", _("A search
> term is required"));
> +                               error = g_error_new (1, 0, "%s", C_("A search
> term is required", "the user needs to provide a search term"));
>                                 goto out;
>                         }
>                         ret = pk_client_search_group (client_async, filters,
> details, &error);
> 
> 

I've talked a bit to other people about this. The basic message is that msgctx
are different from translator comments. The msgctx is only useful to
differentiate the same mgsid, and the comments should be a help to the
translator when they are being translated. I see the translator comments in
Gtranslator -- maybe it's a intltool issue you're seeing?

Anyway, I don't think converting all the comments into msgctx's is a good idea
for now, sorry.