[os-infrastructure] State of the onion

[Alex Hanshaw]

Hi Roger

 Point 6), you base this statement on what? Ingres survived many years
of CA sales guys booking our sales to other products so that could get
bigger commissions. No investment in resources and we came out the other
end with more business and more loyal customers than H&G realized. Even
before they realized this H&G thought "we can make some money here".
Within CA we only posted a tar ball once in a blue moon. We were not
going out of business as a result, we were considered by H&G to be a
viable business in our own right. If I believed that Ingres was doomed
to failure without an Open Source community I'd be brushing up my CV. I
believe Ingres has what it takes to continue to make money without an
Open Source community. I also happen to think that if me implement a
successful Open Source policy we can go from strength to strength in the
market place. If we get our Open Source policy wrong, our customers stop
paying and we go under. That is why we need management guidance, unless
I'm the only one on this list that doesn't know enough about our sales
operations to be able to unilaterally make all of the decisions that
need to be made?
 Item I) in your list is the thing I believe we are getting wrong. "Do
the open source thing" is not a policy that spells out what we should
and should not make available so that we can get the right balance
between community and making money. We have to hit a balance between
giving all of our source away and the old closed source approach. The
equivalent sales policy would be "Do that selling thing". Maybe I'll
sign up for head of sales if that's all that's needed.

 Alex

-----Original Message-----
From: opensource-infrastructure-bounces at lists.ingres.com
[mailto:opensource-infrastructure-bounces at lists.ingres.com] On Behalf Of
Roger L. Whitcomb
Sent: 07 June 2008 05:17
To: Discussions about the infrastructure needed to support a true
opensourcecommunity
Subject: RE: [os-infrastructure] State of the onion

Everyone -
	I'd like to make a few comments from my perspective.  Here's a
little bit of my background:
1) I've just started at Ingres, but I worked on OpenROAD many years ago
(right after the CA acquisition), so I have some experience with the
code.
2) When I first started on OpenROAD I was confronted with the mentality
that anybody but Ingres developers are idiots and it will take 6 months
to make any real contribution to the product.  When I checked in my
first changes after about 2 weeks, and that turned out to be sound, then
a lot of the initial pushback died down.
3) I'm tasked now with redoing the Ingres DBMS toolset, which by any
measure is a LOT of work, so I'm going to NEED as much help / involvment
I can get (but not involvement that is more like thrill-seeking, but
serious nose-to-the-grindstone, with understanding, coding work).
4) My first two or more weeks was largely spent just trying to get an
Ingres build going ....
5) I'm also involved in (just barely dipping my toes in) another OS
project called Audacity, which is a multi-platform audio editing
program, which has lots of history, a fairly active community (for a
small project) and plenty of some of the same hurdles we face (many
platforms, lots of legacy code, bug systems that don't work with each
other, etc.)
6) I don't think anybody will argue that if we DON'T make this OS-thing
work for us, we're not going to be around very long .....

	So, I think we need to keep in mind several points, which I've
been able to glean from where I'm coming from:
A) Both Andrew and David are actually "doing the dance" of making OS
work for Ingres -- Andrew with the DBMS and David with OpenROAD, so I
think we need to respect both their viewpoints.
B) Some of our decision-making (and I haven't really heard anything
about this yet on this mailing list) SHOULD be from the viewpoint of the
OS developer trying to get a foothold into Ingres.  I say this because
even as someone who had some history with the code, it's not easy to
just pick up and start running with this stuff.  In fact, it's not easy
doing that with ANY software project.  So, I think the very WORST thing
we can do it treat the "outsider" as just that, as an adversary, rather
than as someone who MAY become a valuable ally, if only we can get to be
on the same page.  But, on the other hand, maybe this person has an
idea, a viewpoint, a line of thinking that is really good and helpful,
but radically different than what we're used to, and that is hard to
open yourself up to, and probably even harder to incorporate into our
world.
C) It's really HARD to take, what has been for many years, a
closed-source product and all of a sudden open it up to anyone and his
brother, not only from a technological point-of-view, but just from a
personal point-of-view, so there are bound to be a lot of barriers to
overcome.
D) I think David has a good point in that, if we open things up in a
sort-of "let's just throw things up in the air and see what lands
upright" kind of attitude, we're bound to fail.  By that I mean, we need
to have thought things through, looked at everything we CAN look at and
reasonably choose the business model and so forth that will essentially
drive the technological direction.  There are real issues involved and
serious costs and possible side-effects of making wrong decisions.  But
we also can't take so long that we're frozen in fear of what might
happen.
E) I think the second worst thing we can do is treat each other as
adversaries as if we need to get as many people on our side as possible
so our side can win.  NO, we're all in this together, and it will be
more productive to identify the hurdles and work through the solutions
in a cooperative manner than firing the guns through the smallest
perceived weaknesses.
F) We NEED the OS community -- not only Karl and U. of Ilmenau, but
ANYONE who cares enough to invest the effort to get the code and start
looking into it.
G) And, we still have a business to run -- we have customers who are
paying us money that puts food on our tables that we need to satisfy, so
a lot of stuff that has come up, I think, has to do with the fears (real
or perceived) of what happens to this part of life in order to support
the OS community, who may be (again, fear) just in it for a lark, and
not really seriously involved.  And I think these are not trivial
concerns, but I think there ARE answers, but they may not be easy ones.
H) It's also not just about Ingres -- we may have real contributions to
make to other OS projects (SVN or CVS for instance), Python, Ruby and
probably lots of others I don't know anything about.  So, we need to
start thinking "out of the box" as Steve Jobs likes to say and see that
we're (now) part of a much larger world of OS that we know not a lot
about (speaking personally too).
I) And finally (whew!), I don't think we can sort of "wait-and-see" what
management tells us to do.  We ALREADY have a mandate from the
highest-level to do this "OS thing" and it is up to US to figure out how
to make it work, and I think the longer we wait for the other guy to
come up with the right idea and then we'll get on board isn't going to
help us at all.  On the other hand, a lot of discussion just for the
sake of discussion because "his/her idea doesn't fit with mine" isn't
going to help either.

	And finally, for me, I think we need a serious effort at making
the entry cost for the Windows programmer (shudder to think) much lower
than it is.  I seriously struggled for quite a while just finding a
"real" set of build instructions with prerequisites so that I could
successfully build Ingres for Windows (and I have been programming on
Windows since Windows 3.0 .....) (And I DO have some community
contributions along these lines if I can only figure out how to get them
reviewed and submitted ;) ...)

	Just my 2 cents....

Roger Whitcomb | Architect, Engineering | Roger.Whitcomb at ingres.com |
Ingres | 500 Arguello Street | Suite 200 | Redwood City | CA | 94063 |
USA  +1 650-587-5596 | fax: +1 650-587-5550

-----Original Message-----
From: opensource-infrastructure-bounces at lists.ingres.com
[mailto:opensource-infrastructure-bounces at lists.ingres.com] On Behalf Of
Bodo Bergmann
Sent: Friday, June 06, 2008 1:26 PM
To: Discussions about the infrastructure needed to support a true
opensourcecommunity
Subject: RE: [os-infrastructure] State of the onion

Andrew,

I agree that WE (that includes David and myself) should enforce and
empower the community, play our role as know-how multiplier, etc.
And I think we do (also using community wiki/forums, development
sprints, etc.)

So I would truly welcome more community involvement - which would also
change roles from developers to mentors and peer reviewer, etc.

With my questions I just referred to your statement "If indeed 95% of
the work will always be Ingres employees, meaning we're open code, but
not open source."

But I still think the 5% external contribution is a good guess - each of
the contributions will also require quite some internal work (review,
QA, documentation, etc.) at least when taking over the code from a
community contribution into a main codeline, so the main workload (in
measurement of time) will remain internal.

-----Original Message-----
From: opensource-infrastructure-bounces at lists.ingres.com
[mailto:opensource-infrastructure-bounces at lists.ingres.com] On Behalf Of
Andrew Ross
Sent: Friday, June 06, 2008 9:57 PM
To: Discussions about the infrastructure needed to support a true
opensourcecommunity
Subject: RE: [os-infrastructure] State of the onion

Bodo,

It isn't how many contributions we've had. 

Incidentally, a handful from Marty Bowes, OGR driver support enabling
GIS applications to run on Ingres from Frank Warmerdam, a couple from a
team in China, a few contributions from interns in
Ottawa/Warwick/Ilmenau, an EXPLAIN command from Kai-Uwe Sattler from
Ilmenau U., a joint contribution Chris Dawe did with Alex Hanshaw, some
Google summer of code projects, etc. These are the ones I know about off
the cuff.

More importantly, how many people have WE brought into the community and
mentored to get them started. That's the key.

In terms of your comments about DBMS core developers. Two weeks ago I
sat in a room with over 200 PostgreSQL community members at pgcon. Not
all are core mind you, but not all need to be. It was the same thing at
osbootcamp6 for the PostGIS talk. Building a community around an open
source DBMS is entirely feasible.

If anything, we have a huge advantage in our dedicated employees and
resources.

Andrew 

_______________________________________________
opensource-infrastructure mailing list
opensource-infrastructure at lists.ingres.com
http://lists.ingres.com/mailman/listinfo/opensource-infrastructure
_______________________________________________
opensource-infrastructure mailing list
opensource-infrastructure at lists.ingres.com
http://lists.ingres.com/mailman/listinfo/opensource-infrastructure