OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   RE: [xml-dev] The Best Technologies Don't Win

[ Lists Home | Date Index | Thread Index ]

From: Eric van der Vlist [mailto:vdv@dyomedea.com]  


>But since when do Web developers follow these standards? How many web
sites are (X)HTML conform? How many vendors support >CSS2? The situation
could seem to be slowly improving but more or less formal groups like
the WHAT WG could (for the best >or for the worse) still decrease the
influence of the W3C.

Good.  They follow earlier standards and are resistant to change.  The
model is holding.  The influence of the W3C is still strong on larger
entities (say DoD) but weakening in the smaller less connected bodies.

> > >That being said, I agree that W3C XML Schema is dominant in the 
> > >crowd
> > of "XML enterprise developers" but that's a
> > > relatively small community compared to the Web community.
> But one that buys platform tools.   I suspect that 'enterprisy' meme
> a rather calculated tactic by some developers who resent that they 
> aren't getting enough attention for their choices.  It is a cry of 
> pain but from a self-inflicted wound.  In fact, enterprise designs are

> thriving.  One wishes they didn't have to use so much bloatware, but 
> this is phase created by earlier decisions.  Again, swapping energy 
> for time.

>I don't know if we live in the same planet but we do not seem to surf
on the same Web. Mine is dominated by PHP 
>developers who do not care about enterprise applications memes...

Yours is dominated by customers who do.  You see a larger group that you
call 'the web' but it is weakly related.  When you have the time, take a
look at that paper I cited.  There is a model for what you want to
achieve but it requires time to work.

>They do vote because they are using other techniques to validate their
documents. They will vote again when schema 
>languages will really start to enter into the basic Web X.Y technology
stack. BTW, the fact that PHP has no decent 
>support for W3C XML Schema lets me hope that many of them might choose

There are schema technologies in those stacks but you choose to only
focus on the non-bigCo-vendor stacks.  I've no problem with that but you
aren't getting the point of why the best don't win: by the time they are
recognized, the money is elsewhere and the voting is effectively over.
That is what the wisdom of the masses achieves:  very slow inefficient

You say no one listened.  I say few understood.  The choice was
complexity and mediocrity:  a Hohmann transfer.  As long as speed to
market was more important than conservation of effort (time vs energy),
XSD was inevitable.  Relax is good but too late for this phase.  Can
that be changed?  Yes, as you say, by lots of little efforts but not for
some time.

>> You don't see it?  Hmm.  Maybe that is the point.  You can't.  You 
>> aren't a schema user.

>Are you kidding?

>I spend 80% of the time I do consulting dealing with schemas and 90% of
my customers belong to the enterprise crowd that >you describe and use
W3C XML Schema. 

That's what I wanted to see you say.  The majority of your contacts ARE
using XSD.  You know that situation can be improved.  You are making
noise in that direction.  Equate that with the small bits of feedback in
the low-energy transfer model.   Now what is the 'transfer highway'?
Where are there overlapping requirements that act like the cusps of the
manifolds so that a little effort sends them to the other orbit?  Every
time you get a customer to choose like that, you strengthen the pull of
that body by adding to its mass.

>I do much more day to day work with W3C XML Schema than with RELAX NG
and that's why I care. Maybe I shouldn't since my 
>customers pay me to deal with the horrors of W3C XML Schema but when I
see that I could save 75% of my time if they were >using RELAX NG I do

That's noble, Eric, but that isn't the web voting.  That's you. That's
local intelligence at work.  You are claiming to represent something
that you can't represent.  That is the same problem of the URI,
identity, on the web/off the web thinking.  There is no such thing.
There is only a dynamic communicating universe of differentiated
resources that may or may not agree with that URI.

>But these are different communities. My customers are not
representative of the Web and the Web has still to choose a
>schema language. 

You are losing with that classification.  Your customers are just as
representative as any other.  You are deciding that is is YOU who
represents the web when in effect, you are a local representative of
local users.  There is a lot of strength in locale if you understand it.
Any local space has overlaps with other spaces and at those overlaps,
the cusps in the manifold, it takes very little energy to influence a
selection.  But it does take a lot of time.

If they are wasting money on features they never use, they tend to want
to quit doing that.  If they are spending time managing work that has
little ROI, they tend to want to quit doing that.   Cost analyses do out
the cusps.  That is the whole point of low energy transfer:  targeting
while conserving energy.



News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS