[Pulp-list] How about we just merge these core features into Cobbler?

Mairin Duffy duffy at redhat.com
Fri Sep 12 20:22:18 UTC 2008

Seth Vidal wrote:
> It seems like cobbler has functioning code and a good-sized userbase.
> Active development. Pulp otoh has some ideas and some structure but not
> a lot of functioning code to backup those ideas. It seems like Michael
> is suggesting we take the feature set that pulp wants to achieve and
> implement it as a mode/interface/etc of cobbler. Essentially, folding
> pulp into cobbler.
> I'll be honest it sounds like a fair idea. There's going to be a fair
> bit of overlapping code b/t pulp and cobbler anyway - and it would get
> more exposure to the feature ideas from pulp if people could get to them
> in cobbler. Since I know of many, many, many existing cobbler installs.
> While I appreciate it not mattering what code stuff lives in - from a
> user and 'marketing' standpoint it does matter.
> If I've already got cobbler setup I want to see the new bits  added
> there not have to setup something somewhat overlapping but different.
> So, to encourage and expand the userbase that both projects  are
> ultimately targeting. Why not join forces, converge under the cobbler
> code base and scm and move up from there?

I think I said this earlier, but honestly I don't know if it is or isn't 
a good idea. I leave this up to the folks doing the implementation, and 
if Michael is planning to do the implementation of pulp in cobbler, I am 
all the more happy to see something that actually works getting done. :)

What I am worried about is how, at least it appears to me, the design 
work and user interface ideas are getting pushed off to the side as 
something different / not a part of this.

But we could be arguing in circles saying the same thing but using 
inconsistent terminology. I mean, I think the way Michael is spinning it 
is that he doesn't care about UI, he just cares about code that works. 
That's fair enough although I also thinks that limits the reach of the 
actual code, and I do think that having a UI for this stuff is important 
and it was always a part of the original vision as it was something that 
a lot of people specifically said they needed. So Michael has been 
posing that the UI should be a part of Spacewalk, and not cobbler.

Now, that is I think the one piece I'm struggling with. The entire point 
of pulp was to do the core repo management bits of spacewalk right. We 
had determined quite some time ago that it would be better to do that 
with a clean slate than to try to clean up what is already in spacewalk, 
because to be fair I think in the past four years we've been TRYING to 
do that in spacewalk but have not made much but very small incremental 
improvements. I think, at least at some point, that myself and others 
had the feeling that we wanted to break it out into a different project 
so we could have the freedom and space to make the kinds of changes and 
innovations we needed to make WITHOUT risking the core functionality of 

So now I feel like we've come back in a circle, and what has proven to 
be a losing strategy (fixing what's already in spacewalk) is again back 
on the table.

Does that make more sense, put that way?

> If I dare say it - it sure seems like good synergy! :)
> But in all seriousness it does seem like a good place to collaborate
> well that helps get pulp's ideas into production and helps the existing
> users of cobbler.

Sans UI. :(

> am I really offbase here?

I don't think so and thanks for the fresh perspective. I think you 
helped me realized that my main issue is with the UI bits 'living' in 
spacewalk, because I've been there for, again, four years trying to make 
that happen!


More information about the Pulp-list mailing list