Well, there’s your problem,
Mike. 70,000 bifurcating minds will never do anything but quarrel among
themselves. You have to take a leadership role in this thing. Set
up an online poll and have them vote. RELAX in or RELAX out.
Since it can be said that 70,000 is a statistically significant representation
of the web, it’s a fair decision, you can send the results to Ozzie, and
you’re done.
;-)
I don’t think anything will derail
XML per se. OTOH, since Microsoft only has the one option, its customers
are possibly not going to ask for choices that aren’t on the list offered
to them. They believe MS pretty much as the web community once believed
the W3C with the difference being they pay for the software. Being
a former employee of an MS ThrallDom, I speak from experience. We
weren’t going to give up the first version of XSD from MS until you made
us do it.
XML ubiquity is as reliant on that set of
decisions across the web in general as it has ever been. I don’t
think Microsoft’s customers will change their minds one way or another.
If you are judged by that, you’re performance reviews must be a
hairy day but HR is weird everywhere these days since they’ve discovered
averaging, behaviorism, and that they only work as long as they defend company
interests over customer interest. As someone who thinks XML
is pretty much a done deal, XML 2006 isn’t interesting to me but it would
be interesting to see how many Microsoft customers are there and what are they
paying attention to.
As to the fear of resisting the temptation
to pull it all down and the malfeasance that might result, I’d like to
acquaint you with the Alabama State Constitution. It was written at
the turn of the last century and does not provide for home rule; thus, every
State-reserved decision has to be voted on by the citizens of every county in
the State even if the outcome only affects one part of one county.
As a result, over the last 100 years, it accrued thousands of amendments such
that no one knows what they all mean or even if they can be applied
properly. It is dragged out whenever a political agenda for some
candidate or party requires a legal means, and thus it means anything to anyone
at any time. You would think the citizens of our great State would
be ready to scrap it, but no, every time that is up for a vote they vote
against it because of the fear that the party in control at the time would
seize the opportunity to undo the aggregate good done by it. (I’m
not making this up. This is how a State in existence since 1819
consistently ranks 49th and 50th in almost every
comparison. “Thank God for Mississippi” is our mantra. How they manage to beat us at being
last is beyond me because I’m so scared by that idea, I won’t go
there even to visit the birthplace of Elvis.)
Sometimes keeping the customers happy and
keeping the powers that be away from the temptation to do bad things comes at a
high price for everyone else.
I’m teasing you a bit, but a) I doubt
there is much of a war and b) even if there is, apparently Microsoft
isn’t involved in it and c) the only authority is one that seems to count
for less as time passes. Relax is on its own but as the loser of this
round, if history is a guide, it wins the next one. Same as XML.
Oh well, back to hacking on this VRML
world. 80mb and counting….
len
-----Original Message-----
From: Michael Champion
[mailto:mc@x...]
Sent: Tuesday, November 28, 2006
11:44 AM
To: 'Len Bullard';
xml-dev@l...
Subject: RE: Victory has
been declared in the schema wars ...
It’s not FUD …
I don’t have any vested interest in XSD prevailing, and the 70,000 people
at Microsoft probably have 140,000 different opinions on the matter. It
is probably hyperbole J -- obviously RSS/Atom and ODF/OpenXML aren’t
going away just because their normative spec changes from one metalanguage to
another. But I think it could derail XML’s overall adoption as the
de facto standard for data interchange.
One reason is that
after having experienced the XML world through the eyes of our customers’
feedback, they tend to see XML and XSD as more or less inseparable. For
example, the #1 question we’ve gotten about LINQ to XML (formerly known
as XLinq) is “when will you have schema support?” (See Ralf’s
talk at XML 2006 for a progress report). That surprised and disappointed
me (as a “dochead” by instinct), but it’s the reality I have
to live in.
Second, the “cool
people” seem to be looking for alternatives to XML, or are treating it as
a boring part of the infrastructure that you don’t have to think hard
about. XML’s main selling point is its ubiquity; making people
think about it rather than just going with the flow is a prescription for
fragmentation. That might be a Good Thing in the long run (and a lot more
fun than trying to figure out why your XSD validator fails on somebody else’s
conformance test), but not good for the paying customers in the short run.
Think what you want about monopolies and noblesse oblige, “keep the
customers happy” is the rule by which I’m judged.
Finally, I really don’t
think that XSD is dramatically “worse” than the rest of the XML
corpus. The people who really specialize in and understand the spec
can work with it, and the WG is addressing the worst flaws in 1.1 DTDs
are what cause us the most day to day pain. Namespaces are probably second.
If you are going to repair the foundation, why not just pull down the whole
house and rebuild it “properly”? There are good reasons not
to do this, of course, but I’m not sure we’d be able to resist the
temptation.
From: Len
Bullard [mailto:cbullard@h...]
Sent: Tuesday, November 28, 2006
7:10 AM
To: 'Michael
Champion'; xml-dev@l...
Subject: RE: Victory has
been declared in the schema wars ...
MC: “…at this point it seems likely that XML itself will get buried in the
rubble if the XSD towers are pulled down”
Most of us glossed right past that
statement. Are you saying that resistance to XML itself is serious enough
that changing the internal schema machinery or providing support for an
alternative would be costly enough or just enough friction to get the world
wide web of users, buyers, sellers and developers to adopt another data
interchange syntax?
Is that FUD? Is that
speculation based on some other as yet undebated alternative such as JSON?
Is this a serious discussion inside Redmond or elsewhere?
len