[openstack-dev] stackforge projects are not second class citizens

Joe Gordon joe.gordon0 at gmail.com
Sun Jun 21 14:59:15 UTC 2015


On Mon, Jun 15, 2015 at 1:20 PM, Joe Gordon <joe.gordon0 at gmail.com> wrote:

> One of the stated problems the 'big tent' is supposed to solve is:
>
> 'The binary nature of the integrated release results in projects outside
> the integrated release failing to get the recognition they deserve.
> "Non-official" projects are second- or third-class citizens which can't get
> development resources. Alternative solutions can't emerge in the shadow of
> the blessed approach. Becoming part of the integrated release, which was
> originally designed to be a technical decision, quickly became a
> life-or-death question for new projects, and a political/community
> minefield.' [0]
>
> Meaning projects should see an uptick in development once they drop their
> second-class citizenship and join OpenStack. Now that we have been living
> in the world of the big tent for several months now, we can see if this
> claim is true.
>
> Below is a list of the first few few projects to join OpenStack after the
> big tent, All of which have now been part of OpenStack for at least two
> months.[1]
>
> * Mangum -  Tue Mar 24 20:17:36 2015
> * Murano - Tue Mar 24 20:48:25 2015
> * Congress - Tue Mar 31 20:24:04 2015
> * Rally - Tue Apr 7 21:25:53 2015
>
> When looking at stackalytics [2] for each project, we don't see any
> noticeably change in number of reviews, contributors, or number of commits
> from before and after each project joined OpenStack.
>

Looks like my previous analysis was a bit off. Stackalytics is less useful
for gathering statistics on contributons then I originally thought.  Both
the UX and REST APIs are very limited.

Instead I looked at the number of commits and contributors directly from
git (looking only the main repo for each project, ignoring clients etc).

Of the projects listed above, all of them have the most contribuors after
joining OpenStack. In comparison projects already in OpenStack saw the most
number of contributors in the two months before the first big tent
additions. I think this is due to the Kilo release.  So it looks like there
is a measurable bump in the number of contributors once a project joins
OpenStack (although I am finding it diffucult to draw any conclusion about
the number of commits). But when looking further into the data we see a
different story.

* Magnums large spike on contributors (10 additional contributors) but when
looking at the contributor diff, the number should really be closer to 5.
* The 5 additional contributors in Murano can be attributed to new
developers from an existing company plus single patches from from two
developers about sql driver and oslo.

It is hard to read into the jump in contributors after joining the big
tent. But there is definitly something going on, just unclear what it means
over a longer period of time.

data: http://paste.openstack.org/show/310710
code: http://paste.openstack.org/show/310711

What really matters should be diversity, it is easy to see a bump in
development as compaines already involved in a project add more resources
too it. IMHO one of the hopes for a project joining the big tent is to get
new companies to join. Thankfully this is where stackalytics is very
useful.  We can compare contributions by company from kilo and liberty.

* Magnum  -- clear jump in corporate diversity for both reviews and commits
(with new companies getting involved)
  * Kilo reviews
http://stackalytics.com/?project_type=all&metric=marks&module=magnum-group&release=kilo
  * Liberty reviews
http://stackalytics.com/?project_type=all&metric=marks&module=magnum-group&release=liberty
  * Kio commit:
http://stackalytics.com/?project_type=all&metric=commits&module=magnum-group&release=kilo
  * Liberty commits:
http://stackalytics.com/?project_type=all&metric=commits&module=magnum-group&release=liberty
* Murano -- Slight decrese in diversity for both commits and reviews
* Congress -- Liberty numbers are too small to draw any conclusions on
* Rally - Slight increase in review diversity (with a new company joining),
commit diversity had no major change (but it is already pretty good).

>From the little data we have so far, here are my revised conclusions:

* Joining the big tent doesn't automatically mean new companies will
contribute
* Projects that were fairly diverse when in stackforge get new contributing
companies after joining the big tent.
* At this point it is unclear to me if the inverse  (projects that weren't
very diverse before, don't gain new contributors) is true as well.

So it looks like joining 'OpenStack' sometimes has a clearly measurable
correlation with a projects corporate diversity.

It will be very interesting to re-analyize the numbers once Liberty is
released.



> So what does this mean? At least in the short term moving from
> Stackfokeystonerge to OpenStack does not result in an increase in
> development resources (too early to know about the long term).  One of the
> three reasons for the big tent appears to be unfounded, but the other two
> reasons hold.  The only thing I think this information changes is what
> peoples expectations should be when applying to join OpenStack.
>
> [0]
> https://github.com/openstack/governance/blob/master/resolutions/20141202-project-structure-reform-spec.rst
> [1] Ignoring OpenStackClent since the repos were always in OpenStack it
> just didn't have a formal home in the governance repo.
> [2] h
> <http://stackalytics.com/?module=openstackclient-group&metric=commits>*http://stackalytics.com/?module=magnum-group&metric=commits
> <http://stackalytics.com/?module=magnum-group&metric=commits>*
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20150621/61f02429/attachment.html>


More information about the OpenStack-dev mailing list