On Tue, Feb 4, 2014 at 1:25 PM, Kornel Benko <kor...@lyx.org> wrote:
> Am Dienstag, 4. Februar 2014 um 13:09:35, schrieb Kornel Benko <
> kor...@lyx.org>
>
> > > Well, I would like to understand the problem before committing a fix.
> The
>
> > > only way I could reproduce is to go back a few commits, and then I see
> the
>
> > > files deleted. Then, updating to the latest master introduces the same
>
> > > problem.
>
> >
>
> > You are right. Cannot reproduce with recent changes.
>
> > But the fix would also correct a .. misbehaviour?
>
> >
>
> > edit some po-file
>
> > use update-gmo
>
> > -> new gmo in source produced. OK
>
> > use update-gmo
>
> > -> again! new gmo in source produced. Not so OK
>
Of course there is a new gmo, because the user explicitly asked to update
it. What is the problem ?
> >
>
> > The
>
> > OUTPUT "${LYX_DEST_PO}/${_gmo}"
>
> > does not look OK, because the python may 'exit 0' and still not produce
> the gmo.
>
> > OTOH
>
> > "${CMAKE_CURRENT_BINARY_DIR}/${_gmo}.tobeupdated"
>
> > will always be produced.
>
>
>
> Forget it. The misbehaviour is that for all other languages the check is
> done again.
>
> This is *always* the output:
>
> ...
>
What is the misbehaviour then ? It needs to do the check to find out
whether the gmo needs to be recreated or not.
Vincent
