Open Stack

On Thu, Apr 23, 2015 at 11:14 AM, Chris Dent <chdent at redhat.com> wrote:

> What can and should the TC at large, and you specifically, do to ensure
> quality improves for the developers, end-users and operators of
> OpenStack as a full system, both as a project being developed and a
> product being used?



One thing that I see as an indicator of quality is how a system fits
together.  Does it feel like a well-designed cohesive unit or a collection
of individually well-designed parts?  Current indications are that
OpenStack feels to most people like a collection of parts.

I believe it is the TC's responsibility to facilitate and encourage the
cross-project communication required to achieve a cohesive whole.  Work is
already being done with this goal in mind, for example Thierry leads a
weekly cross-project meeting for exactly this purpose, to facilitate
communication between projects.  I do think he is wearing his Release
Manager hat then, but that further illustrates that it is not the TC that
need to be doing the work, but they should be setting the direction and
expectations, and providing a framework within which to operate.

With all of the talk about the TC needing to 'get out of the way', I also
believe that one of the responsibilities it has in technical leadership is
to make some hard decisions when something is not working.  I know I
sometimes have difficulty knowing when a project I am working on is going
nowhere because I am too close to it.  I need someone with perspective to
let me know what I can not see for myself.  To put it into the Big Tent
metaphor, someone has to occasionally clean the elephant poo out of the
tent.  Fortunately this has been a rare event so far in OpenStack's
history, but the opening of the tent flaps changes that activity from an
up-front one to an after-the-fact one and the TC must not forget that
building a quality product is both including quality parts and assemblies,
and removing the lesser quality ones.

As I said before, the majority of my time with OpenStack has been in
cross-project activities and I get a first-hand taste of what people have
to do to work with it.  It is not pretty.  There already are working groups
addressing two of the most glaring inconsistent areas: APIs and log files.
The TC role here is to encourage and support these groups and prompt others
to form where they will be able to do good work.  I have staked my future
with OpenStack to two areas that both need this attention and I work every
day to provide frameworks for developers and those wanting to do
small-scale exercises, and for deployers and end-users who need to actually
use an OpenStack cloud.  I think that the next level for that vision is
strengthen that from within the TC itself.


dt

-- 

Dean Troyer
dtroyer at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20150424/d613e5e0/attachment.html>