Re: [HACKERS] SAP and MySQL ... [and Benchmark] - Mailing list pgsql-general

On Thursday 12 June 2003 13:29, Justin Clift wrote:
> Lamar Owen wrote:
> > Each Open Source project has its own personality.  I often use PostgreSQL
> > as an example of a well-run OSS project; I do believe that the current
> > model is working well.

> I strongly disagree.  The current model is stable, and thus far it has let
> us putter along without any major community disputes that may potentially
> divide the community.

Stable is good.

> However, it's also had a chilling effect on our community, not letting us
> drive the expansion nor give the right attention to the non-code parts of
> PostgreSQL that are important.

Who is 'us'?

> If we had some kind of voting system in place for core, then we would
> likely have a more active and larger community - generated by the people at
> the top being more involved, enthusiastic, energetic, and giving solid
> leadership and direction.

I really can't see how the 'people at the top' aren't already the most
involved developers in the group.  This project is run by the developers; and
that is in my mind a good thing.  'He who codes the most gets to determine
the code.'

> Getting involvement in this from the PostgreSQL Advocacy and Marketing
> group would be extremely beneficial as well, as it's presently lacking
> vision, coherent plans and goals to bring the vision to reality, and
> consistent effort by all but a handful of members.  Good leadership +
> direction would be welcome there and should be included in the PostgreSQL
> "core" group as well.

Is Marc or Bruce not a part of the 'Advocacy and Marketing Group' already?
Let the Advocacy and Marketing group make their suggestions in the open forum
of Hackers and see what is thought of it, just like everything else that has
been done to this database.  There are five people on core; their skilset is
varied enough to where any single point of view isn't dominant.  The who
Hackers group drives the development.

> It's not that we need marketing... we need *consistent* marketing.

If a group of developers can make consistent advances in technology without a
formal structure, why can't a group of marketers do the same?  The process is
well known and fully documented in the archives of the hackers mailing list:
you want a feature in PostgreSQL?  The first hurdle is to get it on the TODO.
That's a Bruce item, and one that only happens with Hackers (and core)
consensus.  The second hurdle is to get a developer with the chops to care
about it enough to make it happen.  If I had the chops to do it, seamless
upgrading would already be done.  No one with the chops cares enough about it
to make it happen.  'Scratch the itch' the Free Software Mantra states.  The
key is getting someone to itch for it that can successfully scratch it.  You
must market to the hackers as much as to the users.

> Ha!  I've seen more funding and offers of assistance that made sense
> _rejected_ by members of the core group - for reasons beneficial to them
> privately even though the PostgreSQL Community would have benefited -
> than I have seen accepted.

Made sense to whom?

> However, your right in that this also demonstates we have a fairly
> tight-knit Community that will help one another out when needed, and that's
> all good.

Yes, we do have a tight community for the most part.

> > To borrow from another projects model, no one is asking Linus Torvalds to
> > accept a voted-in core team for the Linux kernel.  He is also one who
> > governs as little as possible.

> Through delegation.  :)

If you'll note, core developers aren't the ones with exclusive commit
privilege to CVS.  Many developers have commit privileges.  Many things are
already delegated where it makes sense to do so.  Such as the RPMs.

> People seem to get this confused a lot.  Why are you associating a
> successful method of organising resources (time, people, etc) with only
> commercial software projects?  _Any_ project that grows to various size
> points and wants to maintain it or keep on expanding will have to figure
> out ways of co-ordinating their time, efforts, communications, etc, that
> work for them appropriately at all of these size points.  That's just
> common sense.

Common sense is relative.  Why must a project expand to be successful?  Why is
more always better?  We need more good developers, no doubt.  However, take a
look at how many developers we already have.  Maybe we are already
coordinating well.

Sure, we can always improve.  However, I disagree that we must look to a
different system to do it.

> Commercial Software projects and companies often use "models of
> organisation" that are proven to work, and although neither they nor us are
> limited to just those models, we don't need to write them off as being "not
> good enough" just because we don't like the other places that have employed
> them.

But you miss an important point.  We are all scratching an itch.  If this
ceases to be fun and becomes more like work those of us who are doing this
totally volunteer may just quit.  PostgreSQL is a hobby for me; while it
makes sense for me in that I use PostgreSQL, I thoroughly enjoy knowing that
what I am doing is helping people.  The current organization makes it easy
for people to jump in (as long as they make the attempt to understand how we
do things and the areas where we will not likely change (like the license --
it's BSD and that's that!) and try to work with us).

You are one who jumped in with some documentation (techdocs in particular);
what was your experience in getting started?  It's very informal; I like
that.

So I believe that the current 'by invitation only' core group is a good way
for this project to continue.  Feel free to disagree; choice is good.
--
Lamar Owen
WGCR Internet Radio
1 Peter 4:11