dave at neary-consulting
Feb 15, 2012, 9:01 AM
Post #9 of 11
On 02/10/2012 10:45 PM, Jonathan Bryce wrote:
> Now that we've settled on a good baseline of what the mission of the
> Foundation will be, it's time to set down the details of how it will
> be organized to accomplish that mission. Over the past couple of
> months, we've talked to a variety of people involved in other
> foundations (in leadership positions and as developers on specific
> projects) as well as individuals and companies already involved in
> OpenStack. We now have a basic framework for review that we think
> fits with our mission and community. This framework is not
> exhaustive, and there are several open questions and details that we
> have purposefully not taken a stance on yet, such as the details of
> the Technical Committee. We are working on publishing a list of those
> questions and we'll post those to the list and discuss in the meetup
> next week.
Thanks for the proposal.
As I said a few days ago, I think it's important to divorce the
technical goings-on of the project from the financial and organisational
aspects of the foundation - and that certainly comes across as a goal in
the structure document.
It's interesting that you include release management as something which
will transition to the Foundation - in most software projects I know,
release management is a side effect of the development process, and is
generally the responsible of the technical project staff. Is that how it
works in OpenStack too?
I think it's great to see the terms lengthened from 6 months to 1 year
for the Technical Board - that makes more sense to me than 6 month
terms. But in the spirit of "leave technical operation alone", perhaps
it makes sense to separate the changes in the technical board/PPB from
the foundation scope & discussion? It may also make sense to discuss
growing the size of the TB to reflect the increasing diversity of the
OpenStack technical community.
In terms of the foundation governance, it makes complete sense to me
that the companies who are putting money into the foundation have a say
in how that money is spent. Other organisations like the GNOME
Foundation have avoided this in the past, while organisations like
Eclipse and the Linux Foundation have embraced it. Usually, the amount
of money involved dictates what's appropriate.
Since the OpenStack Foundation looks to be setting its scope as an
organisation with a staff, it probably makes more sense to follow the
Linux Foundation type structure. They have a board of 15 members, made
* 7 representatives of 7 Platinum members
* 3 representatives elected by the 14 Gold members
* 1 representative elected by the 111 Silver members
* 2 representatives elected by individual members
* 1 representative nominated from the Technical Advisory Board (their
equivalent of the Technical Board)
The board gets to approve the foundation budget, tell the executive
staff what their priorities are for the year, and do other
board-of-director type things, but they don't get to say what happens in
the Linux kernel. Strangely enough.
The TAB is elected by the kernel hackers at the Kernel Summit every
year, and represents the meritocratic technically savvy group that
serves as a liaison between the different constituencies, the Linux
Foundation and the kernel community. Put simply, they ensure that the
kernel hackers are heard outside the kernel mailing lists, and that
companies investing in Linux get heard inside the mailing lists. The do
things like co-ordinate content for the Plumber's conference, the Kernel
Summit, they talk about how the kernel hackers can do outreach to other
groups, that kind of thing.
Something like this maps closely onto what you're trying to achieve with
OpenStack - the difference being that you're not at the same starting
point. But this type of board & membership/funding structure might make
sense for you.
Like others, I wonder if "Strategic members" is the right way to frame
organisational members. Why not just "Corporate Members"? You could
perhaps define membership levels based on financial commitment - and
there are lots of ways of doing that: straight dollar value (Linux
Foundation has Platinum: $500K, Gold: $100K, Silver: $5k - $20k, based
on # of employees), company size (the Eclipse Foundation sets membership
fees as a percentage of revenues, with a minimum and maximum level), or
some combination of both.
The key point, I think, is that if the technical governance of the
project is truly independent of the organisational governance, then this
is fairly straightforward. What are the financial needs? What value does
Rackspace place on the resources it will be giving to the foundation?
What will the foundatioon actually be doing, and who has an interest in
influencing that? What is the value of being a member?
None of these are easy questions (note: straightforward isn't the same
as easy), but by taking the technical side of things out of the
equation, things get a bit easier.
Neary Consulting - http://www.neary-consulting.com
Tel: +33 982 382 735
Cell: +33 677 019 213