Re: [cobalt-users] Have SUN stopped supporting XTR?

[Travis Ogdon <togdon@xxxxxxxxxxxxxx>]

Gerald Waugh wrote:

> On Sunday 20 October 2002 01:23 pm, Ryan Verner wrote:
> > On Sun, 20 Oct 2002 19:25:43 +0300 aljuhani <aljuhani@xxxxxxxxx> wrote:
> <snip>
> > (PS. Anybody interested in making some kind of unofficial "alternative
> > restore cd" for the cobalts, perhaps debian with a nice web interface on
> > it, that actually works properly? <g>)
> 
> We are working on it!
> We could always use help!


I've been entertaining thoughts lately of doing an 'rpm -qa' and starting at the
top and working my way down the list. The basic list of things that need to be
done:

0. Repackage in a working manner all of the broken/old/cruddy pieces of software
currently installed on the box (and while we're add it toss in ssh, lsof, and a
bunch of other tools that should have shipped on the boxes in the first place).
Fairly easy since almost everything we're talking about is Open Source.
1. Figure out how to reliably flash the firmware so that we may use 2.4 kernels
on the boxes. 
2. Get a working 2.4 kernel (I think Sun has these somewhere in the
depths of their ftp site along with the firmware upgrades). By working I mean
the whole thing, including the front panel.
3. Unroll the os-restore cds (there were instructions posted at some point) so
that we can roll out "new" boxes based on our version. 
4. "Patch" the os-restore cds to a known state (ie. renagade cobalt v1). We'd
likely need to write some sort of shell script that made it easy for newbies. I
have a feeling that Sun wouldn't be happy with us redistributing an OS restore
CD, especially if it's going to contain some of their proprietary stuff. If
we're just distributing a shell script and a tarball of rpms that did the heavy
lifting...
5. Burn our patched OS Restore CD and cross our fingers as we OS restore a bunch
of test boxes.

IF we could get this working in a reliable manner and we were able to document
it I think we could collectively forget that Sun doesn't care about us and move
on with our lives. 

When new vulnerabilties were found in software we could just release a .pkg or
.rpm that patched the vulnerabilty, and not in a cop out manner using the
version from the vendor. There's no doubt in my mind that we could do it faster
than Sun. 

When new software is released with tons of great new features (php, a version of
snmp from this millenia, etc.) we could release packages for those as well. I
suppose we could even have two release branches, a "slightly bleeding edge," and
a "more stable than Sun ever got it."

They're not supporting us in the first place, so what do we have to lose? I
contend that given enough time we could get it working and still have a fully
functional Cobalt UI. 

Thoughts?

-- Travis