Re: [AD] Remove -pipe

[ Thread Index | Date Index | More lists.liballeg.org/allegro-developers Archives ]


On Tue, 2004-05-11 at 23:40, Ben Davis wrote:

> On Tuesday 11 May 2004 21:59, Elias Pschernig wrote:
> > What would be the policy?
> 
> Eric would have proposed the patch, asked for comments, waited for Henrik's, 
> and quite probably rejected it. Alternatively, if no one complained, the 
> patch would be added to CVS maybe in a few days' time, or more quickly if 
> someone praised it.
> 

Hm, true.. that's how it's done normally anyway - so this was an
exception, since it was part of a quite big patch nobody would have read
through completely anyway (or would you?).

> > Some patches are here for days without 
> > comments..
> 
> Then use a bug tracker or patch tracker. Sf.net provides both.
> 
> > (And stop already about that closebox function, it was 
> > admitted to have been as mistake already long ago :P)
> 
> The close box function was merely an example. The underlying problem evidently 
> persists. Eric assumed he was able to properly consider the ramifications of 
> a patch without anyone else's input.
> 

I think there was lots of input - but only after it appeared in the WIP.
(And probably not when it was/would have posted here, in any case.)

> > Probably there was 
> > also some time between applying it to CVS and the release it was
> > contained in. And it's fixed now - so there was one WIP with that
> > function removed - where is the problem with that? We have CVS -> WIP -> 
> > stable. So adding something to CVS shouldn't be a problem at all, as
> > long as it is posted here - it makes it easy to test, and then discuss,
> > and fix/revert before the next WIP.
> 
> On a technicality, a patch was just applied to the mainline. I am under the 
> impression "mainline" means "not WIP", making it CVS -> stable in this case. 
> Tell me if I'm mistaken.
> 
> If you are claiming that the amount of testing going on is an excuse for us 
> not to consider patches properly, I should just give up now.
> 

No - there just shouldn't be any unneeded policies just making it hard
for someone to send patches. If things can be improved, then it's always
good of course.

> > Just look how many outstanding patches (e.g. KCat's) were posted and
> > scrolling out of my mail reader's window.. if in CVS, they would have 
> > been tested more already.
> 
> As I said, use trackers. The patches won't be forgotten, and if they're 
> appropriate, they will be added to CVS for testing.
> 
> My last post, and this one, are intended to encourage Allegro to be developed 
> more professionally. Thanks for making me feel as if I'm doing the right 
> thing.
> 

Yes, I also want it to be developed as professionally as possible. I was
just thinking, if it gets even harder for something to be applied to CVS
than it is now, then that is more worse than good. Didn't think too much
about it though, I must admit :)

As for the SF trackers - maybe we really should start using them more.
Personally, I always liked the mailing list more, since there's not that
much development going on with Allegro anyway - so right now, it might
be more additional hassle for the few developers than it brings good.
Eric sort of is our tracker :)

-- 
Elias Pschernig <elias@xxxxxxxxxx>





Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/