Mohan (this-is-cool) 's snapshot gcc33-20030409
Tue May 13 11:16:00 GMT 2003
erik poupaert writes:
> > Perhaps, but casting a short to a pointer will never work. Anywhere
> > in swt.c that a pointer type is truncated or a short type is extended
> > to a pointer is certain to be a bug.
> Strangely, with VC.NET, none of the warnings occur,
That's not so strange. Many C compilers don't warn about this.
> and it seems to work
> (maybe it shouln't...). I've been operating this version of swt for
> months now ... maybe, they don't even really use these calls from swt
That must be the explanation, because this code surely is broken.
> This is the full dump of problems:
> msys-1.0.9-2003.04.18-1, gcc33-20030409 and w32api-2.3, swt
> Feel free to suggest any corrective action. I'd be really happy to to
> compile this with gcc instead of VC.NET.
> gcc -c callback.c
> works, nothing to report.
> gcc -c structs.c swt.c swtole.c
> In file included from structs.h:50,
> from structs.c:13:
> defines.h:23:1: warning: no newline at end of file
> In file included from structs.c:13:
> structs.h:776: error: parse error before '*' token
> structs.h:776: error: parse error before "OLECMD"
> structs.h:776: warning: data definition has no type or storage class
> structs.h:777: error: parse error before "OLECMD"
> structs.h:781: error: parse error before '*' token
> structs.h:781: error: parse error before "OLECMDTEXT"
> structs.h:781: warning: data definition has no type or storage class
> structs.h:782: error: parse error before "OLECMDTEXT"
gcc isn't #including the appropriate headers, probably because it
doesn't have the right include path.
More information about the Java