Open Stack

Short answer: Dunno.

I need to spend more time thinking but I would suggest that technical side
metrics do OK, perhaps with the wrong focus. What is the be simplest metric
of quality? Again, dunno, but a hi-res smoothed features vs bugs graph
might illuminate. One could model life cycle costs per feature in a closed
environment more easily, but perhaps in a meritocracy one might see a
greater acceptance of measuring success in terms of net problems solved vs
created. I'm happy to be the first over the parapet on that question in the
new year.

On the customer side: Aside from surveys, which are not statistically
significant samples, how do we elicit valuable user feedback?

I doubt that and I understand that public cloud operators will not give us
verbatim customer feedback. But having the only feedback filtered through
the prism of a provider's dev organisation is surely insufficient.

I will be away from the coalface next week, Tristan-willing, so I can spend
some more time pondering.

My general advice is to step back from OpenStack and think about your
customers. Stop thinking about how many times you have prevented your
customers from seeing problems because of your awesomeness and start
thinking about how the platform can help you if it were to improve.

In the end, if we are competing on masking OpenStack problems we all lose.
Implicit to that is a better Enterprise play, but I'll get there next year.

If we compete on adding value to OpenStack's base value then we win. Well,
some of us will. Especially those than aren't playing the OpenStack game
defensively.

Roland
On 11 Dec 2014 01:05, "Rob Hirschfeld" <rob at zehicle.com> wrote:

> Roland,
>
> These are good suggestions.  Is there a small starting place that we could
> get started on?  I'm very focused on getting the small steps rolling
> because I think that having this group communicating regularly and working
> together addresses the critical gap and make it possible to tackle the
> larger issues over time.
>
> Thanks,
>
> Rob
>
> On 12/10/2014 06:08 AM, Roland Chan wrote:
>
>> I don't know for certain. Metrics from the delivery process are great, but
>> we also need good quantifiable customer feedback.
>>
>>  From the technical side, I would to see full lifestyle information at a
>> feature level: metrics related to the quality of the delivery process and
>> also post-release metrics. That is how many post release problems did a
>> particular feature/blueprint/<insert other grouping below project level>
>> create? Hard perhaps, but worthwhile if we're going to improve the
>> customer/operator experience.
>>
>>  From the non technical side, I'd love to be able to track customer
>> satisfaction. Can't think how though, given the nature of the distribution
>> channels for OpenStack.
>>
>> Roland
>> On 9 Dec 2014 06:54, "Michael Krotscheck" <krotscheck at gmail.com> wrote:
>>
>>  What do you want to measure? I'm building OpenStack's project and task
>>> tracking system, knowing what y'all want to report on would be quite
>>> helpful. Resource that can actually flesh out our reporting engine would
>>> be
>>> even more helpful :)
>>>
>>> Michael
>>>
>>> On Sun Dec 07 2014 at 11:38:38 PM Roland Chan <roland at aptira.com> wrote:
>>>
>>>  What we measure is an interesting question.
>>>>
>>>>  _______________________________________________
>> Product-wg mailing list
>> Product-wg at lists.openstack.org
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/product-wg
>>
>
> --
>
> Rob
> ____________________________
> Rob Hirschfeld, 512-773-7522
>
> I am in CENTRAL (-6) time
> http://robhirschfeld.com
> twitter: @zehicle, github: cloudedge & ravolt
>
>