Subject: Re: So... Re: offering pre-purchases
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: Tue, 23 Oct 2007 06:03:16 +0900

Don Marti writes:

 > begin Stephen J. Turnbull quotation of Sun, Oct 21, 2007 at
 > 04:55:50PM +0900:
 > 
 > >  > VCs right now rely almost entirely on nepotism to find deals.  My
 > >  > claim is that they can build markets that will reliably produce
 > >  > deals -- but they actually need to help make those markets.
 > > 
 > > By which you mean they should be subsidizing you and anybody who looks
 > > like you, right?  Maybe ....
 > 
 > Or VCs could be trading away some innovation risks
 > that are dependencies for something they do own "IP"
 > on.  Invest in a movie subscription service, trade
 > contracts on the invention of a cheap 1TB hard drive.
 > Invest in an online document collaboration service,
 > trade on merge algorithms that work on XML.

I haven't *done* the math (it looks like an amusing exercise :), but
I'm pretty sure it doesn't work for either side of the market, not
for Tom as explained before, and not for the VCs.

That is, the amount of technological risk involved in the bets you
mention are second order compared to the business risks of the
mainline.  Think about it: as humans go, engineers are pretty
reliable.  They solve problems, that's what they do.  1TB storage
media *will* be in the range where it makes sense to attach them to a
$500 large screen HDTV in 5 years or so.  By contrast, consumers are
unreliable, they do whatever they damn please.  We don't know if
consumers ever will care to buy 1TB storage media, what with
innovations in networking.  Ditto XML merge algorithms; it's just not
a big enough part of the bottom line for VCs to worry much about the
risk that Tom will make it unnecessary for them to pay royalties on
merge algorithm patents the day after they sign a 5-year contract.

If you (and Tom) can't come up with more interesting hedges based on
technological risk than that, why should our cranially cramped friends
with the overstuffed wallets be expected to?