Firstly, I did not attack your attempts at getting attention to the
project - I simply said that you have to make up your mind about what
the release means. Either it is just a subtle move regarding which
branch we consider stable - in which case we could go ahead and shift
any time - or it is an oppertunity to get some attention, in which case
it is important that we have a release process that produces something
that works.

Neither have I argued against deadlines - I am simply saying that the 
deadlines ought to be about things we can control. I'm saying, go ahead 
and test the current cvs, configure the installers, and make a package 
that will be our release. But then we first release it as a candidate 
and give it a few days (I said a week, but I'm not going to argue about 
two days here or there) to see if any issues arrise. Then take that 
exact package, changing absolutely nothing but the name, and make it the 
release. There is still a deadline: get the release package together. 
Test it for a week.

The only thing that would stop us releasing the whole thing would be if 
critical errors are found in it - that is show-stoppers. Otherwise, we 
go ahead with it as the release even if we are aware of bugs in it.

What is the worst thing that can happen using this strategy? That 
critical bugs are found in the candidate and we are delayed. But what 
would happen under the exact same circumstances if we had gone directly 
to release instead? Well, then we would have released code with a 
show-stopper bug in it as our high profile release. 

You are right about one thing though, which is that these arguments
waste time better spent elsewhere, so I'm not going to get into the the
whole point by point back and forth (though I have to comment that if
you remember our previous releases as smooth, I envy your selective
memory.) I'm just going to remind you who it is here that is on the side
of one in a dispute of ~twenty...

On Sun, Oct 20, 2002 at 10:03:28AM -0700, Ian Clarke wrote:
> On Sun, Oct 20, 2002 at 03:04:51PM +0200, Oskar Sandberg wrote:
> > You interpreted it wrong then, because that is what I meant (or rather,
> > I meant we should set the time back every time we _fixed_ a bug, but
> > that we should not fix any non-critical bugs in the release candidates). 
> 
> This kind of strict release criteria is inappropriate for software at 
> Freenet's current stage of development.  Its only effect would not be to 
> lead to a more stable release, but simply to discourage minor bug-fixes 
> for a week, leading to a *less* stable release.
<>

-- 

Oskar Sandberg
oskar at freenetproject.org

_______________________________________________
devl mailing list
devl at freenetproject.org
http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/devl

Reply via email to