[openstack-dev] [cinder] Proposal for Ollie Leahy to join cinder-core
mordred at inaugust.com
Wed Jul 17 20:28:05 UTC 2013
On 07/17/2013 12:44 PM, Avishay Traeger wrote:
> Dan Smith <dms at danplanet.com> wrote on 07/17/2013 09:40:02 PM:
>>> The affiliation of core team members should not come into a decision
>>> like this.
>>> It is assumed that all core team members are wearing their "upstream
>>> hat" and aren't there merely to represent their employers interests.
>> Mark beat me to it, but.. Yeah, what he said. Core members aren't
>> investments the likes of which get you voting shares and they
>> shouldn't enforced as such, IMHO.
> I agree, and didn't mean to imply that there would be a conscientious
> effort to move the project in a certain way, or that people would be
> purposefully voting for the good of their employers. Of course, voting
> should be based on what the individual believes would be best for the
> project as a whole, for all its users. However, a person's view of the
> project's direction is certainly influenced by the customers they meet, the
> use cases they encounter, and so on. Those employed by the same company
> generally will have similar views.
This POV came up when we were talking about foundation board voting. I
fundametally reject its premise. I do not share the views of others
employed by HP. In fact, the ones we'd theoretically be the _most_
worried about (the mega-big companies such as the ones we both work for)
are the ones who are the least likely to have such a situation happen,
because there are eleventy billion different divisions in each company
that do not talk to each other and have different opinions on what the
best path forward is.
On the other hand, the small startups tend to do 18 hour days around a
ping-pong table with beer and have a higher tendency to share a world
view... they also usually don't have the time in their schedules to let
a massive number of their employees do enough work to be considered for
> It's not because of "voting shares", or
> because of people representing their employers' interests rather than the
> project's. It's because those who come from similar backgrounds will tend
> to have similar views of what is good for the project, and a diverse
> population will tend to have a broader picture of the users' needs. I
> think the current Cinder core members provide a nice balance of views and
> backgrounds - people who understand the needs of public clouds as well as
> private clouds, those who interact with customers who are coming from
> certain deployment models such as Fibre Channel, those who deal with
> customers that are iSCSI-only operations, those that want NAS appliances,
> and those who want to go with server-based storage.
> I believe that diversity of ideas and backgrounds yields the best results,
> and that's why I voted with -1. If I were representing my employer's
> interests, I would go with +1, because HP has been pushing for more FC
> support, which is good for IBM. But I personally have invested many many
> hours in Cinder, and I want it to succeed everywhere. That's why I review
> 5,000 LOC patches from IBM's competitors with as much care as I do when
> reviewing my own code, and even fix bugs in their drivers. That's why I
> listen to every feature request and vote as objectively as I can, even if
> I've never encountered the use case for it myself. I want Cinder to
> succeed for every user and for every vendor, and I think that leadership
> with as wide a view as possible is important to that success.
I absolutely agree with your sentiment, and I think the way that you are
approaching the project is exactly the kind of mentality we want! I want
as wide a set of viewpoints and backgrounds as possible. I simply do not
believe that employer equates to world view - and that in general core
teams should use their judgement as to whether the person being voted in
behaves more like you, or more like someone who is actually tied to
company world view.
> OpenStack-dev mailing list
> OpenStack-dev at lists.openstack.org
More information about the OpenStack-dev