Re: Some bugs found when compiling for maemo

[Philip Van Hoof <spam pvanhoof be>]

Well, maybe this 'detection code' can be done as a script rather than a
little C application that needs to be compiled. Or maybe we can
distribute a header file (per platform) that simply defines it correctly
(for that platform) rather than letting the configure stage do this.

Packagers like Koen Kooi also mentioned this. Because the detection code
isn't easy to get-working on cross compilation situations.

I haven't given this a high priority though. Assuming that eventually
something usable will come out of it. Of course, things don't happen by
itself. So I was wrong I guess ;-)

Please feel free to propose some sort of fix. What happens is actually
quite simple: a iconv-detect.c gets compiled and will run. That run will
create a iconv-detect.h with a define that sets this.

That file will be included in the .c file (e-iconv.c).



On Mon, 2007-02-12 at 12:11 +0100, Sergio Villar Senin wrote:
> Dirk-Jan Binnema nokia com wrote:
> > Index: libtinymail-camel/camel-lite/libedataserver/e-iconv.c
> > ===================================================================
> > --- libtinymail-camel/camel-lite/libedataserver/e-iconv.c   (revision 1573)
> > +++ libtinymail-camel/camel-lite/libedataserver/e-iconv.c   (working copy)
> > @@ -26,6 +26,10 @@
> >  #include <config.h>
> >  #endif
> > 
> > +#ifndef ICONV_10646
> > +#define ICONV_10646 "iso-10646"
> > +#endif /*ICONV_10646*/
> > +
> >  #include <stdio.h>
> >  #include <stdlib.h>
> >  #include <string.h>
> > 
> > 
> > Works for me at least. The problem is that iconv-detect does _not_
> > detect the ICON_10646 on maemo, but the camel code assumes that
> > it does.
> 
> Sure, I used this approach until now, but it's not the most elegant
> solution, isn't it? :)
> 
> Br
> _______________________________________________
> tinymail-devel-list mailing list
> tinymail-devel-list gnome org
> http://mail.gnome.org/mailman/listinfo/tinymail-devel-list