Re: Now: caos and RH

On Tue, Feb 10, 2004 at 11:53:35AM -0600, Chris Spencer told me:
> First off, please understand that I am a big open source advocate.  I am
> also not new to these parts having started back in 1993 with Linux.
> 
> I do understand the ideas of constant improvement. 
> 
> Nevertheless Fedora is a community project and to make it into the
> perfect distribution will certainly take less than to create an entirely
> new framework.

Hrmm... I agree with the concept, but IMO the Fedora base is flawed and 
tainted with non-standard code overly bloated with patches.

> Nevertheless, I acknowledge that it is perfectly fine and everyones
> right to go and do as they please.
> 
> It just strikes me as a better method to work with Red Hat on the
> community project Fedora since they are doing most of the initial work
> for you.

Well, I started cAos with several other people way before Fedora.us was
released. We evaluated the idea of helping RedHat after the Fedora
consumption, but the goals were very different (as I mentioned). Thus we
stayed with our project.

Since then, we have had quite a number of people tell us that they have been
trying to get code into Fedora, and there are no solutions to do this. They
want to help, but are being turned away. ESR is among them. Knowing that now,
I am glad we stuck to our original direction.

It is possible that this will change in the future, but once again this
should have been the primary direction from the beginning (IMHO).

> Fedora Legacy again promises to keep the distributions active longer
> than what Red Hat intends to.  
> 
> Of course all of that depends on people helping Fedora Legacy and
> working for the common cause.
> 
> The more fragmented it becomes the less stable and reliable all become.

Fedora Legacy is a different problem/solution then Fedora core as you probably
know. I will contribute to Fedora Legacy where/when I can.

> Why is it that there is still just one place to go for the Linux
> Kernel?  We are all free to fork and go our own way...but their is
> recognition that with a single focal point more progress can be made.  

There are other kernels out there, but today Linux is perfect for almost all
solutions. There is no general need to create a fork or a reinvention because
it is being supported, and managed well. If Linus were to charge everyone that
uses his name on every system a royalty, then I would assume that the kernel
would be forked, or another solution would take hold.

> There is value in diversity though.  I am sure that good things will
> come from everyones work even if it is redundant.

Of course. It is all open! ;)

> I still like apt.
> 
> GRAB has a few benefits but overall I prefer apt.
> 
> > More examples include the various httpd's, ftpd's, WM's, editors, MTA's, 
> > etc... ;) The point is that wheel reinventing is a fact of Open Source and it 
> > occurs when there is a need.
> 
> Was their a need though?

I think so... some wanted smaller, faster or more secure, etc...

The need is perceived by the individual. I see a need, where you may not. This
is open source, so we will all still be friends and share developments 
(hopefully).

> > RedHat for many (many) years was a perfect solution to most Linux users. There 
> > was no need to research anything else because it was the "sweet spot" distro. 
> > But they decided to change their support models and leave a very large niche 
> > in the community. Many people in the community see Fedora as the life vest, 
> > but I don't.  
> 
> They gave long term support for a return that was less than what it
> cost.  It was a bad business proposition.  I can't blame them for trying
> to find a balance where they can focus on their business customers but
> still offer something of value to the community.

I agree. I hope I didn't mean to sound like I was faulting them for this path.

I admire RedHat... First for helping make it a commonly used OS. Second for
bringing it to the Enterprise along Sun and Oracle.

> > Fedora is a great solution for what it was designed to do. It was not however
> > designed to be a viable long term production system, it is not in their stated
> > goals(1). Fedora is designed to allow the community to take part in the 
> > development process of their leading edge technology. RedHat has made it much 
> > easier to install and work with then its predecessor (Rawhide), but there are 
> > no plans for longevity, and no migration paths (as it has been described to me 
> > by RedHat). Yes, the community may create initiatives to provide longer term
> > support of the releases, but that conflicts with its design goals as I see it.
> 
> It was designed to include the community support from the beginning.  It
> just wasn't what Red Hat wanted to do with it.  

What do you mean? Even today OSS developers are having a hard time getting 
code into it. I know it was stated that it was a goal, but I have not seen it
in widespread action.

> > Hope you don't mind me using this opportunity to speak my mind, and try to
> > clarify things.
> 
> Heh.  It's your list. :)
> 
> Don't think for a minute that I am against anyones rights to do as they
> please here.

Of course... No prob.
-- 
Greg M. Kurtzer
http://runlevelzero.net/
http://caosity.org/
http://warewulf-cluster.org/