On Mon, Jul 25, 2011 at 09:17:27PM +1000, Simon Harris wrote:
Some ranting and raving as a VERY newbie guy on the project:
On 25/07/2011, at 4:44 AM, Hugh Brock wrote:
[snip lots of stuff other folks have responded to already I think]
OK I wil stop now. You're all going to get bored and irritated - if you aren't at
least one of those already.
My main questions around all of this are, to flog a dead horse:
What are the stated goals (ie epics)?
What is the fastest way to achieve those goals?
I see lots of cool features but very little traceability to achieving some desired,
high-level outcomes. Mostly it appears to be about achieving a vision for a fantastic
ideal.
I know many in marketing would have use believe we only have one shot at this and we
don't want to "piss off" potential customers but IMVHO, if a potential
customer won't give us a second look down the track, it's because we don't
have a compelling story not because we pissed them off earlier.
What I feel we need is real-world feedback. I would hazard a guess that if a random
developer could: $ /usr/bin/ruby -e "$(curl -fsSL
https://raw.github.com/gist/somegist)" and install everything, they would start
contributing back and adding features they NEED almost right away.
I don't think anyone on the team would disagree with you on this last
point. What we have right now is:
* Difficult to install
* Difficult to configure
* Difficult to use
* Much easier than it used to be on all those fronts
We are at least now almost to the point where a random developer could
"yum install aeolus-all" and be presented with a working app. But
that's still a way off from where we would like to be I think -- for
example we don't want to restrict our Ruby app to running only on
Fedora, a not-particularly-friendly-to-ruby platform (at least it's
not commonly the one folks develop on).
One of the motivations for breaking things up -- taking the resource
management piece (Conductor) and the image building piece (Image
Factory) and the n-tier orchestration piece (Orchestrator) and the
policy engine piece and making each of them useful on a standalone
basis -- was precisely to make it easy for random developers to grab
one thing and use it and contribute back to it. I think this is
something we should pursue sooner rather than later -- we just need to
make sure we have the right split. The only thing that makes me
nervous about this is that I have a product delivery sword dangling
over my head... but I'm trying to ignore that part.
--Hugh
--
== Hugh Brock, hbrock(a)redhat.com ==
== Engineering Manager, Cloud BU ==
== Aeolus Project: Manage virtual infrastructure across clouds. ==
==
http://aeolusproject.org ==
"I know that you believe you understand what you think I said, but I’m
not sure you realize that what you heard is not what I meant."
--Robert McCloskey