Subject: Re: small worlds and better than ransom
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: Sun, 07 Oct 2007 11:04:57 +0900

Thomas Lord writes:

 > > Unfortunately, your ethical position is that you wish to do so in ways
 > > that make it difficult for third parties to exploit your contribution
 > > economically.  So much for support from VCs ....
 > 
 > That's absolute nonsense.   Tell me: how has the open source nature
 > of the LAMP stack impeded the current generation of, for example,
 > "web 2.0" services?

It hasn't.  It is, however, a perennial complaint of developers like
you that those who benefit from the open source LAMP stack don't
contribute enough back to the initial, speculative developers like
you.  This is a crucial market failure because huge messy
big-ball-o-mud "solutions" like the LAMP stack generate equally huge
installed bases and impede development and deployment of elegant,
small, but equally powerful stacks like XQVM.  (Not a direct
competitor to LAMP, of course.)  Or was that a different Tom Lord who
posted complaints about Red Hat et al?

 > Right.   When I think in terms of my personal "long range plan (or vision)"
 > I think that, if 5 or 10 years from now I've had to sell equity in something
 > and personally take VC money then, probably I've blown it.
 > 
 > The simple problem is that taking VC money is a poor risk if you
 > can't possibly afford to lose -- and that would be my situation in
 > most imaginable scenarios.

Well, on the one hand VC is capital; they don't have grounds to sue
you for the money back.  On the other, unless you negotiate very
shrewdly, you won't own your own product, and (I've heard) VCs often
play dog-in-the-manger with products they fail to make money on.

 > The pre-purchase business is like selling tickets to a performance.

So you're back to "internet busking", which is where you are.  Dammit
Tom, you deserve better, just plain as a human being, and in
comparison to your contributions.

 > There is a venue there, of limited size, and with better and worse
 > seats.  In the course of the performance, a relation develops
 > between performer and audience and this relation colors and helps
 > to shape the performance.

 > The experience of the audient has some value -- the ticket prices it
 > supports.     A live performance has value when the audience becomes
 > engaged in its production -- an experience that can't be obtained
 > from mere recordings of a performance.

But this analogy is temporally backward.  The value in a live
performance is that you *were there*, it can easily be a spot
exchange.  The value in a relationship to an independent software
developer is something in the future, uncertain, and unlikely to bend
to the customer's will, any more than Pavorotti would have acceded to
requests for "Truckin'" for his encore just because people bought
tickets in advance.  (Nonetheless, I'd like to hear it!)

 > Are ticket prices royalties?  Do they interfere with software freedom?

Are you allowed to carry recording equipment into the venue?  In my
experience only for the Grateful Dead; most of my favorite indie
performers funded their guitar strings with CD and cassette sales
(yeah, I'm old enough to have bought cassettes from indies...).  Not
even the indies welcome recording equipment.

It worked for the Grateful Dead.  There aren't very many professional
indies who are very happy about the situation, though.

 > Isn't it a choice, really?   Many firms and many start-up folks
 > regularly mine the public open source world for innovations that
 > can be exploited to create new products or improve existing ones.
 > Pre-purchases and things like them offer an alternative approach.

Why pay for the milk when the cow is free?  These people generally
don't care about future development of the core app.  They have a
widget they want to build, they take your free version, hire a hacker
(with an axe, not one with a scalpel) to beat it into shape, combine
it with the hardware, and spend the lion's share on marketing.  If you
are willing to consult at the level of writing drivers for new widget
hardware and writing optimized assembly code for a particularly
performance-sensitive operation, yes, they'd love to prefund that and
get you committed to it (Ghostscript, Cygnus).  But they're not
interested in funding the progression from larch to tla to revc to XQVM.

Look, you've been there, with Canonical.  They clearly thought they
had a better use for your time than working on Arch, no?  They
considered revision control a problem under control, they could
delegate incremental improvement to people without your vision in that
area.  (I'm aware that there were a lot of other things going on, but
stated as bare facts that's correct, is it not?)

 > One thing to consider is what happens if not just me but
 > perhaps a few people actually succeed in starting a business like
 > this.   That will change the landscape a lot because many people
 > will try to emulate this.  That is, many entrepreneurial hackers
 > will follow suit and offer their work under similar terms.

Agreed, and I sincerely wish you good luck.  But the bottom line on
the back of the envelope I'm looking at says "you're gonna need luck
in round lots."