Re: updated debian development diagram -- comments?
On Sun, 9 Jan 2005, Henrique de Moraes Holschuh wrote:
> On Sun, 09 Jan 2005, Holger Levsen wrote:
> > unstable is described as suited for "...laptops and desktops on non-critical
> > systems..."
> > testing is described as "... can be used for desktop systems that need more
> > stability..."
> > I think this both is wrong. Unstable and testing should not be described as
> > suited for desktops - they are development branches of debian, which are
> > likely to break, which break and... so on. Most of you know :)
> Agreed. Unstable is recommended only for people that "know what they are
> doing". Certainly not for desktop usage, or anything like that.
> As for "testing", well, that one can be recommended to users that need a
> very up-to-date system but who can tolerate the lack of speedy security
> updates... AND who know how to deal with ocasional breakage (yes, sometimes
> it happens even in testing).
Personally I have had much more problems with testing then with unstable
in the past. When its not near freeze large packages (gnome/kde mainly)
tend to migrate partly and it can take a VERY long time for them to become
usable once this happens, also when problem do get through they can take
weeks to be fixed (as oposed to days/hours in unstable).
I am using unstable very happily for a desktop and except for the
ocasional program which got broken for a day or two (very rare) the only
things that get broken are my fault due to setup experiments that should
have never been tried in the first place.
> > good ol' "debian releases to seldom" argument...) - but as said I don't think
> > Debian should propagate this misconcepts.