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

Re: potato late, goals for woody (IMHO)

I think "Obsolete, Boring, and Broken" is pretty extreme, but I have to
concede it is not far from the truth!

First, I think it is impractical to use any distribution which does not
have CDs easily available.  We have been dealing with this on our own, as
I have mentioned here earlier, by maintaining a local Debian mirror --
{http|ftp|rsync}://debian.bilow.com -- and autobuilding our own CD ISOs --
{http|ftp|rsync|://debian.bilow.com/debian-cdimage -- by running the
"debian-cd" tools as a cron job.  Many of the Debian CDs being sold still
provide slink running on a 2.0.36 kernel -- which reduces Debian to the
status of historical curiosity, something like SLS, in user evaluations.

Second, I think we need to strike a more realistic balance between what is
possible in terms of technical quality and what is genuinely useful to
users.  I don't think that potato, as it stands, is any worse in terms of
technical quality than, say, Mandrake 7.0.  It is not entirely fair to
judge Mandrake and Red Hat, or really any RPM-based system, by Debian
standards because, in my opinion, their goals are lower.  There is not
even any centralized control of RPM package names, for example, so there
is no hope of achieving Debian-like robustness when it is quite common
practice to install Mandrake packages on Red Hat and so forth.

There also has to be an appreciation for the size of the gap that must be
jumped from one release to the next.  The jump from slink to potato, for
example, has proven to be a relatively large one because of the glibc 2.1
issue, and it is essentially impossible to install a potato package on
slink if you need it.  This is probably the biggest jump that had to be
made since the change from libc5 to libc6 some time ago, and even that was
exceeded by the horrendous mess of going from a.out to ELF even earlier.  
On the other hand, most of the packages from woody will install on potato
for the foreseeable future, and that should be reflected in the schedule.

The problem is that squeezing out that last 10% of technical quality is
consuming 90% of the time and 90% of the work -- as is always true.

-- Mike

On 2000-05-03 at 05:41 -0500, Manoj Srivastava wrote:

> 	Perhaps since our distribtuiotn has been variously
>  characterized as having Obsolete, Boring, and Broken flavours? I know
>  I am losing converts at work, since I can't honestly advocate Debian
>  servers, since teh CD's we have of Debian are ancient, and Mandrake
>  7.X CD's are far more, umm, current.
>         I think we can bring our users, even those who are not
>  constantly upgrading to teh bleeding edge, a distributin that stays
>  more current without compromising technical quality.
>        Do you think I am mistaken?

Reply to: