Message: previous - next
Month: February 2012

Re: [trinity-users] Please use version numbers!

From: leee <leee@...>
Date: Sun, 19 Feb 2012 12:52:37 +0000
On Saturday 18 February 2012 18:38:00 Kristopher John Gamrat wrote:
> On Saturday 18 February 2012 06:16:07 am Lisi wrote:
> > On Saturday 18 February 2012 10:14:17 Werner Joss wrote:
> > > On Saturday 18 February 2012 11:07:56 Lisi wrote:
> > > > I am baffled by this.  Ubuntu _has_ got version numbers.  9.04, 9.10,
> > > > 10.04, 10.10, 11.04, 11.10 etc., and they go in numerical order.  The
> > > > fact that they also tell you when exactly that version was released
> > > > doesn't detract from that.
> > > >
> > > > And Trinity certainly has numbers (in numerical order).  So as I say,
> > > > I am unable to understand the problem.....
> > >
> > > i think the OP had just the problem to always remember the correct
> > > relation between code name and version number, e.g. what was exactly
> > > maverick ? (10.04, 10.10 ?)
> > > - i confirm having this problem also, from time to time :)
> >
> > Me too!    They are even more silly than teh Debian ones - and taht takes
> > some doing!!
> >
> > But the OP actually said:
> >
> > <quote>
> > Now, i have to cope with childish names. Even the download page for
> > Ubuntu uses version numbers, so it´s not even possible for me to decide
> > which is which.
> > [snip]
> >
> > And PLEASE, use version numbers in the future!
> > </quote>
> Looking at the main page at, you see "Debian Lenny" and
> "Debian Squeeze", not "Debian 5.0" and "Debian 6.0". This may be what the
> OP was referring to. The only reason I know the difference is because I
> have been using Debian for just under a year. If I hadn't done my research
> on Debian beforehand, I'd have no inkling of an idea what "Squeeze" and
> "Lenny" referred to. Since I'm not an Ubuntu user, I have not been keeping
> track of their code names, so I could not tell you the difference between
> "Karmic" and "Lucid" except that they are two different versions of Ubuntu,
> but I do not know what those versions are.

I suspect that the reason that Ubuntu follows the Debian idea of using names 
as the primary way to identify each release, as opposed to version numbers, 
is because it's quite a good idea and it's a good idea because it makes it 
easy to upgrade, or not, from one release to the next.

As previously mentioned by some of the other posters, there are several 
concurrent Debian releases, or more accurately, distributions, at any one 
time.  At the top level these distributions are 'stable', 'testing' 
and 'unstable'.

The way the Debian scheme works is that 'stable' is always the current release 
distribution, 'testing' is always the candidate distribution for the 
next 'stable' release and 'unstable' is always the development distribution.  
When it is decided that the 'testing' distribution is fit for release it 
becomes the new 'stable' distribution release and a snapshot of 
the 'unstable' distribution becomes the new 'testing' distribution.

In addition to the 'stable, 'testing' and 'unstable' identifiers, each 
distribution is given a name (with the names being taken from characters in 
the film 'Toy Story').  Whilst each distribution progresses from being known 
as 'unstable' via 'testing' to 'stable' the name of each distribution does 
not change, so for example, the current 'stable' release, 'Squeeze', was once 
a snapshot of 'unstable', which became 'testing' and then 'stable'.

When you identify the Debian repositories from which you install Debian 
packages you can do so by using either the 'stable', 'testing' or 'unstable' 
identifiers or by specifying the name of the distribution e.g. 'Lenny' (old 
stable), 'Squeeze' (current stable), 'Wheezy' (testing) or 'Sid' (unstable).

Now the reason that this is a good idea is because if you want your systems to 
be automatically upgraded when a new release is issued, or if you want to 
ensure that you're always using the 'testing' or 'unstable' distributions 
then all you need to do is use the 'stable', 'testing' or 'unstable' 
identifiers to specify from which pool you want to install packages.

Conversely, if you don't want your systems to be automatically upgraded to the 
next release then you can use the names, such as 'Lenny', 'Squeeze', 'Wheezy' 
or 'Sid' to identify which package pool to use.

There are valid reasons for using either scheme: if you're running 
production/mission critical systems then you won't want to automatically 
upgrade your systems until you've validated the new release and made plans to 
deal with any manual admin-side changes that may be required, in which case 
you'd be better off using the distribution names e.g. 'Lenny', 'Squeeze' etc. 
to identify the package pool to use.  However, if you're just running a 
desktop system that only uses packages installed from the Debian 
repositories, or if you're a developer who be will always be working on 
the 'testing' or 'unstable' distributions then using 'stable', 'testing' 
or 'unstable' to identify the package pools is more appropriate.

Just for info, there are a number of additional Debian repositories, such 
as 'non-free', 'contrib', 'experimental', 'backports', 'oldstable', 'security' 
and 'snapshot' and these are commonly used in conjunction with 
the 'stable', 'testing', 'unstable' or named distributions.

Sorry for the lengthy explanation but it seemed clear that many folk didn't 
understand why there were different ways of identifying releases and why a 
simple numbering scheme isn't used; Debian only assigns a release number to 
each 'stable' release once it is released and then increments that number by 
points as updates (bug-fixes/security-fixes) to that release are issued, and 
whilst the base release number of the 'testing' distribution might easily be 
predicted, a release number for the 'unstable' distribution is simply