Steve Newcomb wrote: (01)
> "Why
> does Tim Berners-Lee personally make all the decisions?" The correct
> answer, which amazingly few people fully appreciate, is that the W3C is
> a dictatorship because it *must* be in order to allow its members to
> engage in members-only conversations about how the Web will actually
> work. The W3C meets the definition of a "conspiracy in restraint of
> trade" ... (02)
I think that was more true in the 1990s than now. I agree that W3C is
not an "open standards process" for all of its trappings. But it is
becoming more and more of one, simply because there are too damn many
organizations that have influence on the actual development of the Web.
The 21st century reality is that Sir Tim and his privy council were
unable to deliver product at the end of the 20th century fast enough and
effective enough to maintain control. HTTP-cum-Java and XML Schema got
out of hand; PDF and OWL were not their creatures; Google was a power
they expected to own, but weren't smart enough to do as well. For
various reasons, they failed to control Web finance and transaction
security -- even though they foresaw the market, they didn't understand
it. Being Americans, they didn't anticipate the impact of the EU and
ebXML on Web commerce, and the corresponding rise of another power
(OASIS). They didn't invent SOA, because it wasn't in their joint
vision (and they had competing products in the area). And they did not
foresee the impact of open-source on commercial software development as
well as academic software. The Internet became technically accessible
much too fast for the cumbersome development processes of the software
giants, and with one exception, they had no history of building
effective high quality software rapidly. As a consequence, their vision
could be more rapidly realized, and distorted, by others. (03)
W3C is not the only game in town, and the privy council cannot really
control the organization. They can only stall or interfere so long
before the project acquires a competitor in another respectable
organization -- IETF, OASIS, UN/CEFACT, OMG, JCP, ... (04)
> Does the W3C's work to promote Tim's "Semantic Web" vision actually
> benefit the public? (05)
Yes. Its truly supportive work has produced many de facto consensus
standards that are widely implemented and 99% interoperable. (06)
> It probably does, but the real question is, would a
> different vision benefit the public more? I claim that it would. (07)
I'd have to see the vision [so to speak]. But the issue isn't the
vision -- it is the image projected by the implemented specifications,
and other de facto standards on the Web. (And frankly, I think Sir Tim
is already seeing Dorian Gray.) (08)
> I
> make an even more disquieting claim: that W3C's deliberate confounding
> of subject identification with web-resource identification amounts to
> nothing less than an attempt to seek rents on all kinds of ideas. (09)
I have some problems with this conflation myself, but after discussion
with Sir Tim and others, I think this is an instance of Heinlein's
Principle: Never attribute to malice what can be adequately explained
by stupidity. (010)
W3C had a vision for the URI, but they couldn't make it stick, because
they couldn't make DNS's big enough and fast enough. So the real
behaviors of URIs are dictated by the webserver developers, whose market
expands by allowing all kinds of usages, and by the app developers who
create those usages to make their product work. The URI is a hammer,
and it was the only tool in the chest (in order to displace its
predecessors), so it got used (and extended) for everything. (011)
The problem isn't what W3C is trying to do with URIs; it is what all the
apps actually do with URIs. And software engineers are notoriously
stupid about confusing the datum with the object it represents -- hell,
the "object-oriented" community used to claim that was the "right" idea. (012)
> If
> the W3C were really interested in public benefit, as opposed to private
> benefit, it would seek a next-generation Web that would intrinsically
> tend to amalgamate information around subjects of conversation, and that
> would encourage entities *other* than large aggregations of capital
> (entities like you and me, for instance) to share and leverage each
> other's insights, while preserving, honoring, and encouraging diversity,
> including *ontological* diversity (and meta-ontological diversity,
> etc.). (013)
So this is the alternative vision? I'm still trying to envision
"amalgamation of information around subjects of conversation", and
coming up confused. (014)
> The W3C is doing exactly the opposite of this. It's trying to
> build a vast AI system as quickly as possible, and it is well prepared
> to pay for it in the coin of civilization's diversity. For the public,
> it's a bad deal. It's not a recipe for world peace and mutual
> understanding. It's a recipe for economic and cultural dominance games
> and, with "luck", empire. (Whose luck? Not the public's.) (015)
Wow! The evil empire. (016)
> It's important to understand that Tim's knighthood recognizes him as a
> member of the economic establishment (a similar but lesser
> royalty-bestowed honor is the aptly-named "Order of the British
> Empire"). While the significance of much of what Tim says is as elusive
> as that of Alan Greenspan's economic forecasts during his tenure as Fed
> Chairman, this much is clear:
>
> (a) He wants us all to use a W3C-designed ontology language. Diversity,
> schmiversity. The W3C universes of discourse shall be THE universes of
> discourse on the Web. And if you're not interested in any of the
> W3C-blessed levels of support for logical inferencing, you have little
> to contribute to the "Semantic Web", no matter how much you know, or
> what you have to offer.
>
> (b) He wants all ideas to be identified via web addresses. (017)
In a word, rubbish. I don't see any evidence for this at all, and I
think it is a technical slander. Yes, the man has a vision, and yes, he
saw creating a dictatorship of sorts as the means of achieving that
vision in a coming situation of disorganized rapid growth. But he was
only partly successful, and now he is managing a large and
uncontrollable institution, and trying to use his aura to retain some
part of the intent and direct some part of that organization toward it. (018)
Trying to use URIs for non-resources avoids creating a 5-year ebXML-like
project to create a substitute. And that is Tim's chief objection. (019)
> Pat, despite what you say, a web address is just a web address, which is
> why I choose to call it a "web address". Any use of a web address as
> the address of a subject of conversation is (i) irrelevant to its use as
> a web address, unless the subject *is* the web address qua web address, (020)
This is confused. A URL is a web-address -- it locates a host and
perhaps a service. A URI is a "resource identifier" -- it identifies an
"object" of some kind that is accessible over the Internet. It isn't
necessarily the means of access; the standard only requires that there
be a server that can take it as an input and produce a means of access. (021)
> (ii) creates ambiguity at the very root of all expressions, ambiguity
> that cannot be resolved in a way that maximizes the value of any
> intellectual commons that uses web addresses as subject addresses, and (022)
The question is: What exactly is meant by "object that is accessible
over the Internet"? One can argue that anything you can buy from
Amazon.com is an "object that is accessible over the Internet". We
commonly conflate identifiers with things in various ways. When I order
a "book", I specify the identifiers for a class of physical objects,
which are printed copies of the intellectual property that has that
title, and what is sent to me is an instance of that class. How many
conflations are involved? (023)
And why is it a big concern that uses of URIs may represent similar
conflatiions? A URI is just a _term_, and agents will interpret that
term to refer to the objects that fit their functions. (024)
> (iii) to the degree that the confusion between web addresses and subject
> addresses is publicly accepted, to that same degree rents on
> publicly-owned ideas can be collected by the owners of the domain names
> that appear in the web addresses of those ideas. (025)
Well, now, there is a business model. If an organization provides the
server that provides the access to the object, it certainly incurs a
cost. Shouldn't it be able to recoup that cost by some business
strategy? Yes, unless it's function is to serve the public good by
providing that service using public funds, or altruistic endowments. (026)
If instead the organization provides no service, but acquires the domain
name from IANA and licenses its use in other identifiers for a fee, that
is a different business model. And the license agreement will be
whatever both parties are willing to sign. (027)
I don't see that as a vast conspiracy. I don't even see that as a
business model likely to be successful. But if it is, it will surely
have competitors. And in a certain sense, how is that really different
from the relationships between authors and publishing houses? (028)
> The "Semantic Web" creates a technological/legal means,
> at least for aggregations of capital large enough to hire teams of
> lawyers, whereby public conversations about things that matter to the
> public can be more privatized, and more big-media-influenced, and even
> less diverse, than they already are. (029)
Wow! This has 'non sequitur' written all over it. I don't see this
conclusion following from any part of the contorted reasoning above. (030)
> Pat, your claim that W3C is not a vendor consortium because it
> observably has so many academic members is unworthy of you. (031)
But it is a fact. W3C has too many influential non-commercial members,
specifically influential in the technical content of the
Recommendations, for it to be considered a "vendor organization". And
it is too big and powerful as a body to be effectively ruled by the
privy council anymore. That is just the reality of 2009. (032)
-Ed (033)
--
Edward J. Barkmeyer Email: edbark@xxxxxxxx
National Institute of Standards & Technology
Manufacturing Systems Integration Division
100 Bureau Drive, Stop 8263 Tel: +1 301-975-3528
Gaithersburg, MD 20899-8263 FAX: +1 301-975-4694 (034)
"The opinions expressed above do not reflect consensus of NIST,
and have not been reviewed by any Government authority." (035)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (036)
|