Re: gerror.c revisited

[Sebastian Wilhelmi <wilhelmi ira uka de>]

Hi Havoc,

> > > Then the convention for error codes is G_EXEC_ERROR_* members in
> > > GExecErrorType.
> > >
> 
> You used GThreadError instead of GThreadErrorType, we should decide on
> a convention for that. (Tim?)

Yes, as you used GExceError in gexec.c ;-) I'll change that to
GThreadErrorType though.

> > This last part looks like it will be used in a whole lot of functions  and
> > it seems a bit tedious and easy to get wrong. So I included
> > g_propagate_error for that purpose, what do you think?
> >
> 
> It looks like a good idea to me, you might make g_propagate_error
> return whether an error was propagated, so:
> 
>  GError *local_err = NULL;
>  foo (&local_err);
>  if (g_propagate_error (err, local_err))
>    return NULL;

Ah, even better.

> If Owen and Tim agree the gerror.[hc] changes look good to me. I
> didn't review the GThread-specific stuff in any detail.

Ok, I'll commit it tomorrow, if no one objects.

Bye,
Sebastian

PS.: I suppose, you meant to mail both me and the list (as I meant to mail you
and the list and failed to do so;-( the first time) so I'm answering to the
list.
-- 
Sebastian Wilhelmi                   |            här ovanför alla molnen
mailto:wilhelmi@ira.uka.de           |     är himmlen så förunderligt blå
http://goethe.ira.uka.de/~wilhelmi   |