[Exherbo-dev] Exherbo Impressions ~ From Installation to Uptime

Benjamin R. Haskell exherbo at benizi.com
Fri Mar 5 06:02:46 GMT 2010


On Fri, 5 Mar 2010, Sess wrote:

> On Fri, Mar 5, 2010 at 18:10, Malinka Rellikwodahs <aelmalinka at gmail.com> wrote:
> my $0.02
> 
> On Thu, Mar 4, 2010 at 23:44, Sess <leycec at gmail.com> wrote:
> 
> > On Fri, Mar 5, 2010 at 01:48, Ciaran McCreesh <ciaran.mccreesh at googlemail.com> wrote:
> > 
> > > On Thu, 4 Mar 2010 18:38:58 +1300 Sess <leycec at gmail.com> wrote:
> > > 
> > > > {*} Stable exheres. There aren't any. Most exheres are known 
> > > > unstable and marked as such, or otherwise known stable (...for 
> > > > most architectures for most distros) and still marked unstable.
> > >
> > > The whole stable / unstable thing's a Gentooism that doesn't 
> > > really work. When we reach a point where we consider stability to 
> > > be interesting, we'll probably do something completely different.
> >
> > Who doesn't consider stability interesting? I do. (Not as much as, 
> > say, Debian developers. But I do.)
> >
> 
> I'm pretty sure his point here is that major things are going to be 
> broken in Exherbo pretty regularly and therefore most of the exheres 
> are still unstable.  It is mentioned that exheres-0 is actually an 
> unstable api and will be renamed if it ever stablilize leaving 
> exheres-0 as the testing api.

I agree with this, and I think that's what Ciaran meant with:

> > > When we reach a point where we consider stability to be 
> > > interesting

But, my take on the rest...

> > > The whole stable / unstable thing's a Gentooism that doesn't 
> > > really work. ..., we'll probably do something completely 
> > > different.

...is that the biggest problem with marking things with the binary 
distinction of 'stable' or 'unstable' is that it's too arbitrary.  
There's nothing that really marks an obvious transition -- it doesn't 
suddenly become bug-free software.  Someone just decides "okay, it's 
been tested enough".  I would hazard a guess that the threshold varies 
widely per-architecture, for example.  Let alone something like the *BSD 
Gentoos.

I'm kind of curious as to what the something-completely-different might 
be, but not curious enough to force the discussion if no one's actually 
formulated anything.  (...since enough else will probably change by the 
time it's implemented that it doesn't really matter what the idea is 
now.)

-- 
Best,
Ben


More information about the Exherbo-dev mailing list