[RULE] FWD: question about RULE installers

[Michael Fratoni]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thursday 13 February 2003 09:35 pm, Matthew Miller wrote:
> On Thu, Feb 13, 2003 at 07:55:57PM -0500, Michael Fratoni wrote:

> I think it's mainly to make sure there aren't weird kernel version
> clashes. But is really enough of the code to reduce memory useing in
> the first stage (loader) part of the installer? Wouldn't you need to
> change the second stage too? I guess I need to look at your code a bit
> before talking more. :)

Yes, to actually reduce the real memory requirements of anaconda, other 
changes would need to be made. However, keep in mind that our goal was to 
continue to use the stock media.

Miniconda for Red Hat Linux 7.2 would install without difficulty with 12M 
of RAM. (For 7.3, the installer needed a bit more, 12M became a bit 
"iffy".) There is an artificial limit coded in into anaconda that our 
patches remove. (Even the stock anaconda would install in 16M, once the 
hard coded limit was removed.) Additional savings are gained by reducing 
the number of packages installed for a "base" system. Fewer packages, 
less work setting up and executing the rpm transaction. It's really just 
a quick hack taking advantage of the anaconda 'updates' function.

> > A Miniconda installer? No. I never quite got around to finishing
> > Miniconda for the 8.0 release. I am currently fooling around with a
> > new version of Miniconda for 8.0, which should allow rapid release of
> > an installer for the next official release. Assuming of course that I
> > get it finished and it works...
>
> If you're fooling around, I'd suggest at least using the copy from the
> Phoebe2 beta, if not rawhide. That'll save you trouble when the next
> release comes out....

Yes, I have a Pheobe2 system set up for testing as well.

> > I can easily create an anaconda based installer for any given
> > version, but it would require RULE distributing the ISO images. I
> > don't have the bandwith for that, and I doubt the FSF would be happy
> > if we used their servers either. We aren't trying to create a new
> > distro, and that is
>
> I *do* have a new distro, so my goals are a bit different. :) What
> you're doing does sound very cool, though.
>
> > >> Basically: would we be
> > > > best off adopting your installer for BU Linux?
> >
> > Depending on your needs, maybe. ;)
> > The Slinky installer is an option. It does _not_ do everything that
> > Anaconda does, but it does install a working system. On first boot,
> > kudzu
>
> Is it able to do a network-based install? That's key.

Slinky is capable of NFS, HTTP, and FTP installs. (I've even done a serial 
port install with a null modem cable and slattach on an old Zeos notebook 
w/ 8M of ram and an 80 Mb hard drive. It was painful.) For the 8.0 
release, you need to be sure to provide an i386 kernel package if the 
target system requires it. The installer scripts still need a little more 
cleaning up to handle the missing i386 package. I should have another 
version of slinky ready in a few days with several fixes as well as a lot 
of cleanup.

Network installs aren't too difficult with slinky. The only issue is that 
the network must be brought up manually. The installer makes no attempt 
to discover hardware and load modules. (We do now have a bootable 
hardware detection floppy to help with identifying hardware.) A script is 
included allowing the user to define several variables (IP, modules, 
gateway, etc.), which does make it a bit easier to get networking 
established. The slinky ISO image (27Mb) includes an i386 kernel package, 
all the modules contained in the kernel-BOOT rpm package, and PCMCIA 
tools. If the ISO method isn't possible, the modules can be obtained from 
my server. Not ideal, but it works.

> Either way, this looks like a good way to get a modern RH on my Toshiba
> Libretto (which has been stuck at 6.1), even if it turns out not to be
> helpful for the university at large.

It should be useful enough for the laptop at any rate. As for the 
University at large, I'm not sure. Hopefully it will be, if not now, as 
it progresses. Progress to date is slow, and depends on my ability to 
find the time to write and test changes to the scripts.

- -- 
- -Michael

pgp key:  http://www.tuxfan.homeip.net:8080/gpgkey.txt
Red Hat Linux 7.{2,3}|8.0 in 8M of RAM: http://www.rule-project.org/
- --
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE+TH08n/07WoAb/SsRAvPUAJ4j6SN9GpcEmQNDnEhbhdyA3KZwtQCfYfTO
uycpBTO/mQL4ArJETmszA/0=
=m6nG
-----END PGP SIGNATURE-----



_______________________________________________
Rule Project HOME PAGE:  http://www.rule-project.org/en/
Rule Development Site:   http://savannah.gnu.org/projects/rule/
Rule-list at nongnu.org
http://mail.nongnu.org/mailman/listinfo/rule-list