[172] in SIPB_Linux_Development
Re: Distribution
daemon@ATHENA.MIT.EDU (Calvin Clark)
Fri Oct 8 17:43:26 1993
Date: Fri, 8 Oct 93 17:42:11 EDT
From: Calvin Clark <ckclark@mit.edu>
To: ghudson@mit.edu
Cc: linux-dev@mit.edu
In-Reply-To: "[171] in SIPB_Linux_Development"
People who are asking for stuff like that get what they get. It's like
getting the hurd snapshot, with a zillion files that look like:
Makefile.broken.~78~
core.~7~
If you tell them ``we don't have anything coherent yet'' and they
persist, just let them log in to quiche and take what they think they
want; if they forget something, that's their problem.
Still, a quick solution would be to have a floppy disk or three labeled
"Kerberos" which correspond to tar files on line. The user then simply
copies the disks or downloads the tar files, and untars them. The only
thing you have to do is make sure that whatever they untar will sort of
be complete and work on the kind of Linux system people are likely to be
running.
You could also have "Zephyr" and "MH" disks if those packages are ready.
Source should be separate; many people will only want binaries and the
necessary configuration files.
Producing such disks and tar files should not be too hard, but I don't
think it should be done until we know everything we put on them builds
from source, without incident, on some known version of Linux, with some
known C compiler, and runs without too many bugs. (We could provide a
list of known bugs in the port.) In other words, somenone should try
building, running, and installing everything on some machine other than
quiche, preferably one that's just been installed with a vanilla
distribution. Since I don't have access to such a machine, I'm not
volunteering for that but I'm willing to prepare some of the sources for
that.
-Calvin
