Re: Am I too impatient? was: [PATCH] Wheel mice (try one)

[Owen Taylor <otaylor redhat com>]

Nils Philippsen <nils@wombat.dialup.fht-esslingen.de> writes:

> Hi, Owen, hi rest,
> 
> On 27 Mar 1999, Owen Taylor wrote:
> 
> > 
> > Nils Philippsen <nils@wombat.dialup.fht-esslingen.de> writes:
> > 
> > > A week ago I submitted a patch to this list and still haven't received an
> > > answer to it yet. This confuses me a bit, I was sure that this is the
> > > right list to post patches for gtk+ (or should it go to gtk-list?)
> > 
> > The simple answer is, yes, you are too impatient ;-)
> 
> I almost expected this answer ;-)
> 
> > 
> > We try to look at patches as quickly as possible, but that
> > is not always as quickly as we would like... 
> > Unfortunately, your patches is far from being the oldest patch we
> > have sitting around needing a response :-( 
> 
> This is what I feared... Would it be more intelligent to post such stuff
> also to gtk-list (not just gtk-devel) to have more (potential) testers. So
> when you (i.e. the core team) get the chance to mess with it, other people
> already have tested it and e.g. found problems, flaws?

Experience says:

 - patches posted to gtk-list get looked at only by the
   person applying them to GTK+. 

 - 90% of problems with patches are problems that are
   apparent reading the patch, not bugs in the patch.

If you are submitting a patch that makes substantial
changes to the GTK+ API, then it is good to post
a message explaining the changes and asking for
comments.

But that's what this list is for ... discussion of changes
to GTK+...

> > I'll try to take a look at your patch in the next few
> > days as I get caught back up here. I haven't looked at
> > the code yet, but my initial reaction from your
> > description is:
> > 
> >  - Moving the stuff I put into gtkmain.c somewhere else
> >    is probably appropriate. I'm not sure yet where that
> >    "else" is.
> 
> Me neither.
> 
> > 
> >  - gdk_std_modifier_mask() is in the wrong layer (policy
> >    belongs in GDK), is configured in the wrong way -
> >    the user, not the application, should set this stuff and
> >    the idea of having one "standard modifier" is a bit dubious 
> >    to begin with.
> 
> I didn't mean applications to set this resource, my thought was that
> themes (or a gtkrc resource) would set this, but I wasn't sure how to
> handle this with gtkrc. I must admit the name is misleading, maybe
> {gdk,gtk}_mouse_action_modifier or so is more appropriate. But IANA
> linguist :-)
> 
> >    I'd almost rather actually leave this stuff hardcoded until
> >    we have a way of letting the _user_ set such options.
> 
> gtkrc? I think I'll take a look into this.

My general opinion on the global option problem is that
we should first solve the (harder) problem of setting
widget arguments from an RC file, then the global option
problem will be fairly obvious.

Regards,
                                        Owen