[Ocaml-biz] The tactical future of OCaml in 1 year's time

[Brandon J. Van Every]

Tony Edgin wrote:
>
> I'm confused.  INRIA owns Ocaml.  We can't force a logo upon
> them.  They have
> the power to choose the logo and mascot, and we can only make
> suggestions.

Your confusion is that you view the law in theory, not in practice.  In
practice, we could try to create a faits accompli, get the OCaml
community to like it, then get INRIA to like it.  In practice, INRIA
might not be legally sharp or attentive enough to defend their
trademark.  If they don't defend it, it's fair game to poach it.

The PSF was well aware of this legal dimension and were vigilant about
asserting their legal responsibility to defend their mark against
dillution.  Unfortunately, this put them in a 'fortress' mode where they
couldn't bring themselves to accept outside help.  They were techies.
Although they knew they had to *prevent* infringing marks, they had no
desire to *create* a high quality mark of their own.  Instead they've
got the eyesore that Guido likes well enough, at http://www.python.org .
And so the world waits for Guido and the PSF to get their marketing shit
together.  It's really O'Reilly that provides any kind of marketing face
for Python, not the PSF.

We do not want to be put in the position of INRIA holding the keys to
the business kingdom.  They are an academic body.  At this time, they
have pretty well stated that they don't fundamentally care about
business.  Or if they do care, they're going to be secretive, closed
door, and Cathederal about it.  They aren't going to have open source
style public discussion of OCaml business strategies.  The 'bazaar'
model of business will never materialize under their tutelage, and that
includes efforts at logos and brand identities.

Asking permission is the wrong strategy.  The correct strategy is to
offend and be forgiven.

In the best case, we may not even generate offense.  The logo may be
liked and may be adopted by INRIA.  That requires quite a bit of
preparing the ground, however.  Not to mention quite a lot of
intellectual and artistic talent, and some luck, to create a logo
they'll swallow.

In the worst case, the OCaml business commmunity has to form a strong
enough identity that it can sustain its own efforts, and market itself
in its own way, with its own legally defended trademark and brand
identity.  If we reach that worst-case stage, we're talking lawyers,
incorporation, and a whole lot of things quite beyond the ken of us mere
techies right now.  It is quite a bit down the road and best left to the
future.  Let's see if we can get any real cooperation amongst commercial
OCaml developers first.

> I think possibly you had an extremely bad experience with
> PSF.  There is
> little to suggest that INRIA will be so hard to deal with.

The warning sign is that in the final analysis, INRIA views the R&D of
OCaml as its goal.  Not its business promotion.  That's what Xavier has
publically stated.  How amicable they really are to others doing the
job, and how much they're willing to cooperate on it, remains an
unknown.  It's a risk.  Contingencies should be expected and planned
for.

> A thing to keep in mind is that INRIA isn't in conflict with
> us.  At the very
> worst, they are neutral about about OCaml's commercial use.

Which is going to be a problem as the business community starts wanting
things from them.  INRIA doesn't have a problem with us *now*.  Because
right now, we're nothing.

> I'm guessing
> they are unofficially interested in it happening.

What we really want, for all sorts of things, is much better than a
guess.  We need a diplomatic liason to INRIA.  I don't know what person
can fulfil that role.  I believe that person is not here yet, and we
will have to find them / get them here somehow.  At present, someone
might take up the task of *finding* the liason.  It could take awhile.
I expect it can only move at the pace of serendipity.  We might hope
that by doing 'good things', INRIA smiles upon us sooner, but we know
little about what they consider 'good things'.

I'm quite sure I'm the anti-diplomat.  If you ever want to *break*
things with INRIA, send me right in!  :-)  If you can find my evil twin
(good twin?) somewhere in France, you'll have your man.

> What's cooler for
> programmer than have his product be a commercial success?

Many open source hobbyists are anti-business.  Don't underestimate their
sentiments.

> What's cooler for
> a university than getting name recognition in the business
> community?

Bluntly, recognition in an academic journal, and contributing to the
body of academic knowledge.  Academics aren't rewarded for business
strategies.  They are rewarded for publishing papers.  Matthias Blume of
SML/NJ was very up front about this.  At one point I detoured from OCaml
to SML/NJ, thinking it would be better at low-level C integration stuff.
Then he told me he only cares about his work contributing a few good
ideas to some future academic's research someday.  I detoured right back
when I saw the community was dead, and Matthias would *never* lift a
finger to make it otherwise.

> Okay,
> probably a lot of things, but research institutions have to prove the
> usefulness to the community at large or the government cuts
> their funding.

I'd look into INRIA's funding base before making that rationalization,
let alone advancing that argument to anyone at INRIA.  Such bullet
points must be handled *delicately*.  Furthermore, don't forget that
Xavier has tenure.  He's not in fear of anything.


Cheers,                     www.indiegamedesign.com
Brandon Van Every           Seattle, WA

Taking risk where others will not.