[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Debian testing vs unstable for home workstation?



On Fri, Jun 11, 2004 at 04:45:06AM -0400, Kevin Mark wrote:
> On Fri, Jun 11, 2004 at 07:29:43AM +0000, Adam Funk wrote:
> > I've been using Debian testing on my home workstation for a while now
> > and am generally happy with it, but I understand there are some
> > disadvantages in comparison with unstable, such as recency of security
> > updates.  I suppose the name "unstable" puts me off, since I work from
> > home a lot and need a functional computer all the time.  What would be
> > the positive and negative effects of switching to unstable?
> > 
> > 
> > -- 
> Hi Adam,
> From the posts I have read, 'testing' is best for people testing debian
> before it is released as 'stable'. 
> 

Testing is a collection point for preparing the next stable
release. Its usually in big cause between releases, get better near a
release (like now).

> 'unstable' packages, after some user
> testing, automatically move into 'testing'. After begin in 'testing'
> and

Packages enter into testing automatically, don't know whats the method
for keeping them back if there is a big problem, I think only big
problems hold them back. Fixes can takes weeks to filter down to
testing.

> some more time, the package goes into 'stable'. In this respect,

New packages never make it into stable. The only things that filter
down into stable once its released are security patches.

Testing is made into stable only in version releases.

> 'testing' is the place where packages go before they inclusion in
> 'stable' debian. so, 'testing', after a new 'stable' has been released,
> is very close to 'unstable'. And before the release of the next
> 'stable', 'testing' is closer to 'stable'. 

Its more like its stabler close to a release.

> 
> It has a few characteristics
> relateing to this: if some part of kde in 'testing' has may bugs, and
> this cases kde to not function, the whole of kde is removed from
> 'testing'. So, groups of packages can be moved in and out as things get
> fixed in 'testing'. But as 'testing' gets closer to being 'stable', in
> tends to have less big shift like this. But the idea is that if you NEED
> kde, all you can do is wait until things are fixed. This can be a few
> days or a few months.
> 
> With unstable, things are being put in all the time and it does not have
> the hugh package shift like 'testing'. Also, bug fixes reported in unstable
> or testing get put into the next unstable package. 
> 

Unstable gets new packages all the time. Most problems that appear
there though are when the program has had major changes or there was a
change in the packaging scheme and then some packages may not be
installable for several days and in such cases if the package is
already installed you won't be able to update it until things are
fixed, but everything already installed will continue to function, you
just need to watch the upgrades to make sure aptitude is not trying to
remove something you need and just wait with updates until its fixed or
hold the packages which have problems at the current version until
things are ok (the = under aptitude).

> And as for security updates, 'stable' is the only one that has this.
> 

Because unstable doesn't need it since it gets the new version quite
quickly and these changes are then ported to stable if needed (only
security fixes, not features). Testing usually suffers for a few weeks
in some cases until the security fixes finally filter through.

> So, most people who do not need the 100% stability of 'stable' chose

Usually only servers, at least if you watch your updates.

> 'unstable' because it is more up-to-date with respect to hardware and
> versions.
> 
> take this with a grain of salt as IANADD. (I am not a debian developer)
> :-)
> 
> -Kev




Reply to: