Re: Gtk+ 2.3 ABI breakage or Gtk-Error **: incompatible build!

[Owen Taylor <otaylor redhat com>]

On Fri, 2003-10-31 at 17:51, Tor Lillqvist wrote:
> Owen Taylor writes:
>  > Hmm, MSVC must pack things different. With GCC everything from
>  > allow shring to destroy with parent will go in the 16 bits
>  > after configure_request_count, so we go from 16 to 17 bits
>  > in a 32 bit space, not 32 to 33 bits in a 32 bit space.
> 
> But note that even if the binary distribution of GTK for Windows that
> I provide *is* built with gcc, I use the -mms-bitfields flag. This
> means that it uses exactly the same field packing rules as MSVC. (The
> obvious reason to use that flag is so that the libraries can be used
> both from gcc- and MSVC-compiled code. I wouldn't want to drop that
> possibility.)
> 
> I haven't built GTK HEAD recently, but presumably Hans's problem would
> occur also with a gcc-built GTK HEAD library being used by a
> application built with gcc with the 2.2.x headers.

I'm not suggesting that we ignore the problem, just explaining why there
wasn't a problem for those people working on Unix.

Do you happen to have a reference for the MS bitfield rules? We
often have questions about when adding a bitfield member preserves
bin compat and having a wider range of possibilities than what GCC
does would be good for answering such questions.

Regards,
					Owen