Re: PATCH new OOo2 file extensions

[Pavel Roskin <proski gnu org>]

Hello, Leonard!

On Sat, 2005-05-14 at 12:00 +0200, Leonard den Ottolander wrote:
> On Sat, 2005-05-14 at 00:13, Pavel Roskin wrote:
> > For example, the "stable"
> > branch compiles with a significant amount of warnings by the latest gcc.
> > This wasn't the case when 4.6.0 was released.
> 
> I don't see *any* warning on the MC_4_6_1_PRE branch, which is the
> branch that should be used for a 4.6.1 release.

What's your definition of "latest gcc"?  According to gcc.gnu.org, it's
4.0.0.  How about this warning:

utilvfs.c: In function 'vfs_mkstemps':
utilvfs.c:211: warning: pointer targets in assignment differ in
signedness

It doesn't even require -Wall.

> > I'm sorry, but I don't have time to port every minor fix to the 4.6
> > branch.
> 
> We have been applying these "safe and minor" fixes to MC_4_6_1_PRE for
> months as we - we being at least Pavel Shirshov, Roland Illig, Jindrich
> Novy and me - expect that branch to be used for a 4.6.1 release.

I appreciate it.

> Instead of complaining about all the effort of back porting (which most
> of the active developers don't seem to mind), could you please release a
> rc for 4.6.1 from the MC_4_6_1_PRE branch?

It depends on how stable it is.  Last time I tried it didn't work well.

> P.P.S. If you still feel that the X11 error when running mc over ssh
> should be fixed before a release of 4.6.1 could you please give us an
> indication of how you think this issue should be fixed? I would expect
> the fix needs to be applied to init_key_x11() in src/key.c but I am not
> quite sure how to do this. Just closing the module, window and display
> as in done_key() and setting them to NULL or 0 does not suffice to fix
> this issue. Do we need an extra parameter to indicate we can't use X11
> related functions?

I believe an X error handler is needed.

-- 
Regards,
Pavel Roskin