Semantic Web Takeoff and economics was RE: Asunto

On Wed, 2004-01-14 at 11:28, Victor Lindesay wrote:
> Thanks Joshua. As someone who is Microsoft platform through and through,
> one thing that's I've found a pain trying to do anything with RDF is the
> lack of tools for the Microsoft languages (VB and .Net). For example
> having to write my own .Net parser knowing that all sorts of great RDF
> kit is freely available for most of the other platforms and languages. I
> suppose I've been spoilt by the excellent Microsoft XML tools over the
> years from MSXML through to the .Net XML classes. Could you put in a
> word maybe Joshua ... :-)
> May I say something that might rattle some cages? 
> Without Microsoft RDF support, the Semantic Web will not happen.
I can see why you say that on one hand, however there are a couple of
counterpoints to that. I think that support in the browser and other
end-user commodity software would be great. At present msie has the
lions share, however they have basically stopped updating it. It's
beginning to show it's age in a few spots and they have never fixed a
few things like proper png support and a few other bad bugs. They
promise a new browser in longhorn(few years away) that uses c#. So
Semantic Web will take a few years if things follow MS's lackluster pace
of innovation. Early adopters will be obligated to go with other
solutions, that don't have MS's support. This offers great window of
opportunity for competitors if rdf or other technology has any real and
apparent compelling value-add to the end-users.

Other solutions like Open-Office, Linux, Mozilla, mono are not only
innovating faster at this point, they are gaining more take-up in
several end-user environments. Israel, Asia, IBM, and other large
parties have been in the news lately as vigorously promoting
alternatives to MS. Key MS products categories are gaining commodity
status, and many MS products do not offer much in way of compelling
competitive differentiation. This lack of differentiation is driving the
smart/wise shoppers to seek alternative sources even if the initial
costs might be slightly higher. In a lot of ways it reminds me of when
Coca-Cola in the 1930's "paid 40 cents a pound more to certain
manufacturers for certain ingredients to keep two sources of supply
open"[1] or how the military has guidelines to buy from multiple
suppliers. Rational self-interest -- Coca-Cola didn't want predatory
pricing from a monopoly sugar supplier. They had to fear the Havenmeyer
family and other groups, which wanted some more of Coke's profit margin
for themselves. Looking into alternatives for MS products is less costly
than the Coke example-- in fact has been known to produce drastic price
reductions from MS. Instead of being "MS through and through", at least
testing linux, moz, and mono and trying not to be gratuitously
incompatible can give you benefits; at worst it may give you great
negotiation leverage when you need it; at best using the alternatives
might significantly lower your TCO for widely deployed solutions. Think
bottom-line economic benefit.

MS is having lower uptake and upgrade levels then historical levels.
Recently they made noises to stop support of win98, however they had to
keep at least some support. 25% or so of people were still using win98
and MS was unable to offer compelling reasons for them to upgrade at the
price levels they are offering at. They also might be afraid if they
push to hard people will switch to a different platform and have to
compete against wine offering good enough support for older win api's.
They are losing ability to create uptake(upgrade trend-mill batteries
low) and thus field new technologies like rdf. They must offer
compelling reasons to upgrade, twist the vendor lockin angle more,
and/or lower costs. They also must do this while competing against lower
cost solutions that are having increasingly rapid innovation cycles.

The above is some of the reasons why I don't think MS has shoe-in for
creating the widely deployed solutions of the future. What does make for
a good success of a technology product? is one attempt to
find out.

Tim Bray lists nine factors which might plausibly be useful in
predicting the success of new technologies: Management Approval,
Standardization, Return on Investment, Compelling Idea, Investor
Support, Good Implementations, and Happy Programmers, and Technical
Elegance, 80/20 point. His list didn't really mention that cheap, widely
deployed creates it own wake at times. That is what I think you meant by
needing MS support or could it be vendor lockin? However your own RDF
lib and redland's which started
making c# binding since 0.9.13 using swig. Are both things which can be
used by many people... Even MS can use, and distribute them. MS would
prefer you used BSD licensing instead of Creative Common-- whats theirs
is theirs and whats yours is theirs;-).However even their "Windows
Services For Unix" includes a few GPLed pieces of software, so they
aren't that allergic to unamerican cancer really;-> BTW the Creative
Commons license is not recommended for software. I'd try LGPL/MPL mix or
just BSD depending on your goals. Adjust the licensing a little and
advertise on freshmeat, and you can help push the technology where it
really counts. The great thing then is that it doesn't matter if the
future is that MS stays on top, or people switch to more friendly
technology sources-- both can use your stuff--It's not bundled to
particular products or companies that way.

In Conclusion: Make your own day! You don't have to wait for MS to get
off it's hunches. The real push is can you supply the right goods at the
right price at the right volume.

[1] Albert H. Staton, Brief on Company Policies, 1939, MS10/51/4
Cited from _The_Essence_of_Capitalism_(2003) by Humphrey McQueen
40 cents a pound was big difference and a big deal back in 1930's.

Received on Wednesday, 14 January 2004 21:27:33 UTC