You have to change the rules of the game and redefine winning.   It
sounds impossible but isn't.  Consultants matter but it helps if they
are influencing major procurements.  Even then, there is no guarantee
that what is procured will have a shelf life unless it is *obviously*
useful.

One of my favorite software vendors is Adobe.  They have a
non-controversial approach to software politics, are willing to try new
application domains (parallel experimentation), and their products do
something I need done without messing up products or their outputs (in
most cases).  They follow standards but they don't let them get in the
way of innovation. 

They aren't the most reliable date for a party but they do bring a nice
corsage and a safe car.

The dilemma is core technology costs and the frequency of buys (market
cycle).  I look at Sun and my conclusion is they take on too much
community effort and don't make enough *obviously* useful plugins.   I
have to wonder what the heck they use for market research.  You can't
make sales on 'we are the good guys' presentations.  You sell something
better when the customer is convincing themselves and the competitor has
worked themselves into the wrong corner and can't adapt fast enough.
OpenDoc is an example of almost getting it right but Microsoft in
classic borg fashion is adapting.

I wrote a blog response to some citations of Tim Bray on the Point of
Value at  http://lamammals.blogspot.com that takes up the topic I've
made in the past:  software is now in the same evolutionary path as the
music business.  To stay in it, you need hit software AND hit software
writers.   In that business, entrenched artists are dislodged
frequently, but being in the top 40 is almost as good as being number
one as long as you keep producing.

len


From: Michael Kay [mailto:mike@s...] 

> RelaxNG may be the greatest thing since sliced bread, but until it 
> reaches it's way into main stream, it will be a niche area.

That last sentence sounds like a tautology...

But it's like Windows vs Macintosh. Most people want to use what
everyone else is using, even if they hate it. Unless it's totally
useless, which XSD (like Windows) isn't. You can't dislodge an
entrenched leader just by producing a better product.