Minimal install (was: Re: Features for FC4)

Jeff Spaleta jspaleta at gmail.com
Tue Nov 16 05:08:51 UTC 2004


On Mon, 15 Nov 2004 21:01:50 -0600, Rodolfo J. Paiz
<rpaiz at simpaticus.com> wrote:

> I just found that file, for the first time in my life. 

Really? That's interesting. Are you actively watching and
communicating on the anaconda-devel-list?
Have you read up on how anaconda does things?
http://fedora.redhat.com/projects/anaconda-installer/
http://fedora.redhat.com/projects/anaconda-installer/comps.html


> It's 8000 lines long for Chrissakes, and even if I'm sure I can
> eventually parse it that's NOT the point. You want a comps.xml file?
> Then bloody freaking make it clear to people who are NOT programmers how
> to read and write the thing. Don't go mouthing off about no one doing
> any work.

Oh im sure people find your howto useful as a workaround.... but its
not a solution to the problem of getting a more sensible native
"mimimal" anaconda install option. It's one thing to have a discussion
as to how to rip out things post install to create a better minimum...
its quite another to have a discussion as to how to include a better
minimal option in the anaconda installer.  For the anaconda minimum,
we are running in circles until someone codifies things in the
comps.xml for testing. I want this discussion to go somewhere.... i'm
tired of people asking/demanding that the installer provide a better
"minimal" option than it does. The only way forward in that discussion
at this point is through comps.xml creation and testing that
incorporates a better minimal set.
I'm pretty sure at this point that the people who care about this,
know what that the realisticly acheivable better minimal set basically
looks like. And its certainly clear to me that on the list of
priorities internal Red Hat manhours this issue is in the subbasement.
The only way forward is for someone in the community starts producing
a comps.xml files that other people in the community can test and beat
the crap out of by rolling test images and doing installs, and that
internal redhat developers can quickly diff and merge against the
official comps.xml when its ready to evaluate possible incorporation.

 
> Frankly, I'm not only pissed at your attitude, I'm also disappointed by
> it. Not the usual intelligent, productive, and useful commentary I
> expect from you.

Really? That's amusing since this isn't the first time i've
communicated these opinions, I'm nothing if not a broken record. I'm
pretty sure I've been saying the exact same thing in one of the many
minimal install threads every release cycle of fedora. But I am only
human, I can only repeat myself so many times before even my resolve
is crushed. I consider it a sign of my own personal weakness of will
that i have never responded to any of your posts until this moment. I
will redouble my efforts to make my opinions known in every thread
where you post from now on. I must cure you of the delusion that my
commentary is usually intelligent, productive or useful.

> Can you at least point me to a doc showing the
> structure of that thing, so we get some value out of your contribution
> to this thread?

Sorry, I assumed all the interested parties had read the availlable
documentation at the fedora website already.
http://fedora.redhat.com/projects/anaconda-installer/comps.html
Is that enough?  If not anaconda-devel-list might be a good place to
ask for more specifics...and to beat someone up about providing better
documentation.

-jef




More information about the fedora-test-list mailing list