Open Stack

On 14 November 2013 13:34, Colin McNamara <colin at 2cups.com> wrote:
> Not to be contrarian, but 92% of the commits in Havana came from
> non-individual contributions. The majority of those came from big name
> companies (IBM, RedHat, etc).
>
> What I see as a great thing is the increasing number [and diversity] of
> companies committing, especially from end user/operators.
>
> In the operator case, there are examples where an operator uses another
> companies Dev's to write a patch for their install that gets commited
> upstream. In this case, the patch was sponsored by the operator company,
> written and submitted by a developer employed by another.
>
> Allowing for tracking if the fact that an operator/end user sponsored a
> patch to be created further incents more operators/end users to put funds
> towards getting features written.
>
> This is a positive for the project, it's Dev's and the community. It also
> opens up an expanded market for contract developers working on specifier
> features.
>
> My perspective - I work at and operator / integrator. I have my teams
> working on multiple projects including OpenStack. Peers of mine in Silicon
> Valley who have funded major OpenStaxk development Efforts have required
> that code to be released, but have had trouble verifying. The sponsored by
> tag would provide an easy way of tracking, as well as further incent the
> behavior of funding improvements.
>
> My 2 cents.

It's not *at all* clear that the publicity from having commits tagged
as 'sponsored by FOO' are valuable for organisation Foo. There are two
scenarios I can see where this turns up:

a) Operator/Deployer X wants a bugfix/feature and their supporting
organisation Y delivers the work for them.
b) Vendor X wants a bugfix/feature and they contract to another
OpenStack connected organisation Y to do the work for them.

For case a) the reward of getting the work done is it's own benefit.
There is perhaps a tiny bit of kudos they get by upstreaming the code,
but as being a contributor to OpenStack isn't key to their business
model, it's marginal at best: if being a contributor was key, they
would be resourcing the work themselves.

For case b) again the feature/bugfix is it's own benefit - the vendors
users get the ability to use the bugfix/feature and the vendor can
sell more of their product. And again, if being part of the OpenStack
community is core to their plans, they will be doing that!

So - there are intrinsic motivators for doing this work. Do we need to
track the (I suspect) small fraction of patches with this provenance
in an explicit fashion at all?

-Rob

-- 
Robert Collins <rbtcollins at hp.com>
Distinguished Technologist
HP Converged Cloud