Re: Debian doesn't have to be slower than time.
Quoting James Troup (james@nocrew.org):
> Michael Neuffer <neuffer@sciobyte.de> writes:
>
> > Then you just do a "make world".....
>
> Are you aware that it takes well over a week of pure build time on a
> reasonable machine[1] to recompile all the architecture dependent
> packages in main?
Yes, very much so, but you don't need to update everything
everyday. Look at how (Free)BSD does it:
Just go into a subtree and hit make.
If I had decent connectivity I would have already set up
a dedicated machine for the Debian "make world" project,
that could be used by all developers that want to work on this.
Between 1995 and 97 I poured a lot of my money into Debian to
build up and finance its infrastructure.
2 years ago at the last real ALS I already said and I still stand
by this, that I will pay for what I consider to be a tremendously
important piece of architecture for Debian.
I believe that we must have:
1. a centralized source archive for the distribution.
This can only be done with CVS or something similar.
Maybe we could even use Bitkeeper for all I care.
Besides practical reasons, there are even legal ones
why we need this.
2. a build world mechanism
A build world would instantly get rid of many problems and
delays that we've been having for years. I am convinced that
it would enable us to do at least 2 releases per year (I'm not
stating here that we need more the 1 a year, but we at least
could if we needed to) instead of struggeling to manage to
to one release in 1.5 years.
A new kernel ? A new X release ? A new whatever ?
The screaming has been big everytime. It takes ages until
the last maintainer crawls out of his hole and pulls himself
together enough to rebuild his packages in an environment
of which we don't know how it looked like and how it affected
the package.
Sorry for the language, but I say: screw it.
We can and should get rid of these unnecessary problems.
If somebody provides the bandwith, I will pay for a dual AMD or
Intel box with a gigabyte of memory and mirrored disks.
This should be enough for the first steps of a CVS archive
and a build machine that can be used to work on a build world
for Debian.
Cheers
Mike
--
People often think of research as a form of development -- that it's
about doing exactly what you planned, doing it on time, and doing it
with resources that you said you'd use. But if you're going to do
that, you have to know what you are doing, and if you know what you
are doing, it isn't really research."
--Dave Liddle, The New Yorker, Feb. 23/Mar.2, 1998, p84
Reply to: