[GRASS-dev] [GRASS-SVN] r52607 - grass/trunk/raster/r.lake

[Markus Metz]

On Thu, Aug 9, 2012 at 10:35 PM, Hamish <hamish_b at yahoo.com> wrote:
>>>>> - G_fatal_error(_("Both seed map and coordinates cannot be specified"));
>>>>> + G_fatal_error("%s", _("Both seed map and coordinates cannot be specified"));
>
> Martin:
>>>> sorry for ignorance, but what is the sense of such changes?
>
> Vaclav:
>> > the older version triggers a GCC warning complaining
>> > about format string.
>
> gcc 4.7 or ?
>
> Markus N wrote:
>> > then it should be changes in *all* GRASS.
>
> better clarify that it is a false positive or real problem before
> cluttering up the entire codebase for a dubious and/or needless
> change.
>
> short of this being demonstrated as a real problem, my vote would
> be to not mess up the entire codebase and tell gcc to be quiet or
> get gcc to fix their test.

+1

I don't get any warning about the format string. IMHO much more
interesting are the warnings in r.lake about not initialized file
descriptors.

Any idea why the warning about format strings is triggered for
G_warning() and G_fatal_error(), but not for G_message()? I assume
that G_message would have been affected as well because the relevant
code seems to be identical. Interestingly, in g.proj, G_message()
seems to trigger the same (bogus?) compiler warning, according to the
change log.

Any idea why this affects only r.lake and not other modules, apart
from g.proj, and there also G_message()?

If this is a real problem, why fix the symptoms and not the cause?
That would probably be the macro defining _().

my 2c

Markus M

>
> 2c,
> Hamish
> _______________________________________________
> grass-dev mailing list
> grass-dev at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-dev