RE: [Bug-gnubg] probable fixes for catalog and gtkfile things

[macherius]

Two issues with your patch on Win32:

1. No <strings.h> on Win32 (gtkpanels.c)

Fix:
#ifndef WIN32
#include <strings.h>
#endif

2. Generated parser files gone
The generated sgf*c and external*.c did not check out anymore with a fresh
CVS co.

Ingo


> -----Original Message-----
> From: address@hidden 
> [mailto:address@hidden On Behalf 
> Of Jim Segrave
> Sent: Tuesday, February 22, 2005 12:30 AM
> To: address@hidden
> Subject: [Bug-gnubg] probable fixes for catalog and gtkfile things
> 
> 
> 
> I've just done the mods to Makefile.in/am to add gtkfile to 
> the top level 4zMakefiles and to add catalog.xml/catalog.dtd 
> to the met/Makefiles.
> 
> While I was at it, I replaced all the C++ style comments with 
> C comments, cleared various small compiler warnings and added 
> CVS Id tags to all the .c and .h files which had them missing. 
> 
> Compiles under FreeBSD now only generate warnings for type 
> punning with pointers (hard to get rid of), an inlined 
> virtual function in a FTGL header file and Pythons use of tmpname()
> 
> For systems where a file off_t or a time_t won't fit when 
> cast to unsigned longs, some printfs will be peculiar, but 
> they used long ints to print without casting before, which is 
> even riskier.
> 
> I can't test that Makefile.am will build a valid makefile, as 
> I have not been able to find a working set of 
> autoconf/automake/autoheader which will work with gnubg for 
> the last year or so.
> 
> -- 
> Jim Segrave           address@hidden
> 
> 
> 
> _______________________________________________
> Bug-gnubg mailing list
> address@hidden http://lists.gnu.org/mailman/listinfo/bug-gnubg
>