Jiri Denemark wrote:
> > Nonetheless, many of my manual changes highlighted
> > existing (bogus) attempts to translate VIR_WARN0? diagnostics.
> > Those should be addressed regardless.
> >
> > Considering the large number of existing VIR_WARN uses with strings
> > marked for translation, do you want to reconsider?
> >
> > $ git grep 'VIR_WARN.*_('|wc -l
> > 60
> >
> > or would you prefer to remove the _(...) marks on those 60 uses?
>
> Yes, pretty much all of them can be removed. I notice there are also needless
> translations in a handful of files with VIR_DEBUG, and for VIR_ERROR too.
IIRC, there was an agreement on the list that VIR_ERROR should be left
translated since those messages can make their way to end users.
Thanks for the clarification.
I've done the VIR_WARN* normalization (will post shortly)
and will eventually do the following, too:
ensure that all VIR_ERROR diagnostics are marked
ensure that *no* VIR_DEBUG diagnostic is marked
That is consistent with existing trends:
- very few VIR_DEBUG diagnostics are marked, and
- barely 1/3 of the VIR_ERROR diagnostics are unmarked
$ git grep 'VIR_ERROR\s*("'|wc -l
67
$ git grep 'VIR_ERROR\s*(_('|wc -l
110
$ git grep 'VIR_DEBUG\s*("'|wc -l
233
$ git grep 'VIR_DEBUG\s*(_('|wc -l
26