RE: [xml-dev] SGML complexity (was: RE: [xml-dev] Re: Recognizing...)

["Len Bullard" <cbullard@xxxxxxxxxx>]

Valid point, Mike.   The production of loss leaders is a perennial issue in
any business case.   I hazard to guess that it has to be a prominent
contributor to follow-on business or it can't be justified.  Hobbyists are
fine for keeping up list morale, but as we experienced with VRML, only value
transactions can sustain a long term ecosystem.  Egoboo isn't enough.  Even
if the market place is just a hot dog stand, free relish is purchased by a
few cents more for the hot dog and the bun.  A free browser needs a $299
editor and some sustainable defense against piracy.  It is a good thing that
I can make any XML application in a text editor with a batch parser if I
have the chops but to be productive, an application editor is better and
those don't need to be free.  Just affordable within a regime of value for
value.

On the other hand, the bigcos of the SGML world and the music world and so
on had a lot of middle-biters.  The execs living in those obscene megaplexes
on the sides of the hills outside San Jose aren't doing a lot more than the
average kid sitting home grinding our a parser.  Rewards became
disproportionate to contributions because as was said, selling software was
like printing money.   Anytime a system has unlimited input and no balancing
control, it goes to runaway feedback and eventually, something breaks.

So yes, it comes down to working out fair dinkum.  Reality sets in after the
bubble bursts and just before the next gold rush.

In the music market, the editor costs $299 for a solid kit and then one pays
between $29 and $50 for the VST-compliant plugins.  The kit will do the
necessaries and the plugins do the competitive-edgies.  It means some kids
do time at Burger-King for a summer to get started, but once they are
started, putting a self-sustaining act together is mainly initiative and
time spent on the phone.  Putting a megaplex on the hillside is a different
set of gear and contacts and a whole lot of luck.  

Earning a set of personal values that enables one to know what 'enough'
means is a lifelong gig.  The one thing I can say about my recent work
experience is that if enough people at the top haven't earned that yet, a
corporation and its customers are paddling toward a receding beach and not
noticing that giant wave rearing up behind them.

len


From: Michael Kay [mailto:mike@saxonica.com] 

> That's pretty much what I remember too.  Even the MS products 
> were free to download.  That was one of the aspects of the 
> SGML On The Web project.  It kneecapped the overpriced 
> software market.

Unfortunately, I think it also kneecapped the free software market to some
extent. It's noticeable how few of the XSLT 1.0 vendors are producing a
version 2.0 product, and in most cases I think the problem is that it's
difficult to produce a business case for developing free software. People do
a 1.0 out of blind enthusiasm and optimism, but a decision to do a 2.0 tends
to be more hard-headed.

The sort of six-figure sums that Omnimark used to quote were clearly out of
proportion to the value that most users were getting, and this severely held
back the SGML market. But I don't think that zero is the right price either.
For a healthy market, software should be like books or music - $25 to $50
should be the normal price. The pendulum has swung a bit too far in my view.

Michael Kay
http://www.saxonica.com/