size of XQuery developer community

[Michael Ludwig]

Hans-Juergen Rennau schrieb:
>
> of course you are right in that XSLT and XQuery are built on the XDM,
> not on XML as a serialization format, and you are right in that the
> XDM can be completely separated from the serialization format. But I
> prefer to view XML and the XDM as one organic whole, and exactly this
> is what I meant by the term "deep understanding": an understanding
> covering three _levels_ of XML: level #1: the lexical; level #2: the
> inforset (= structural model totally unconcerned with aspects of
> processing); level #3: the XDM extending the infoset in order to
> embrace the reality of processing = a global information model which
> preserves the structural achievements of the infoset and embeds it
> into a stunningly flexible definition of information value. Each level
> evolved from the previous one, I think, was it not so, historically?

Hallo Hans-Jürgen,

historically, I'm a newbie; but the other day on XML-Dev, one SGML/XML
veteran said that it was rather the work of "ladling committees" (had to
look up the word - seems similar to "anflanschen" in what it conveys)
ladling on patch after patch to XML.

RE: [xml-dev] Xml Revisited - Len Bullard
http://markmail.org/message/n5zscmypan4rzmaj

I really like this realistic view of things, but I also have a lot of
sympathy for idealistic interpretations.

As for the three levels you're referring to, I don't know enough about
level 2 - the XML InfoSet - in order to locate it in the scheme of
things.

What I've learnt (from Ken Holman) is that with XML, syntax came first,
and data models (plural) are an afterthought. Maybe a simple and obvious
thing to many, but - at least to me - crucial to understanding this
whole XML business. Like a lot of things, it only makes sense seen with
history in mind.

> This evolution is a wonderful achievement

Why not call it an Evolution? The process that has brought forth XML is
too visible in order for us to call it a Creation. And it works too well
to call it a Fabrication. (Although unfortunately, some fabrications
work pretty well, too.)

> and I am emphatically against "marketing" XQuery and XSLT as tree
> processing languages, as if you could come along with some tree and
> get it processed.

Well, you can. You just have to make sure your tree is represented in a
format amenable to an XML processing chain. Quite often, I start out
with documents like <Urmel/>, parse those into a DOM or XOM, and then
append data retrieved from various sources (HTTP, SQL, XQuery) to then
finally hand off the XML tree to XSLT and generate an HTML tree.

> What should be done is rather: encourage people to get familiar with
> the new reality which XML (via XDM) constitutes: a global information
> model of stunning power and flexibility, enabling XPath, XSLT, XQuery
> and what is to come.
>
> A poet (Joseph Brodsky) once wrote: it is not so that the poets should
> adopt the language of the masses; common people should adopt the
> language of literature.

But common people have, to my knowledge, never done so; which is fine.
And if you take a look around today, I fear you'll have to concede that
there is movement, but in the opposite sense of what Brodsky deemed
desirable.

> In the realm of IT, XDM/XPath/XSLT/XQuery are like literature, and I
> would not sell them for something else.

I agree there is beauty in it: the power and flexibility of the tree
navigation model.

Just consider that the computer has to perform some work to set up this
nice machinery; and while I find it often worth while doing so and
pretty convenient for me, I have to concede that it is not always
necessary.

-- 
Michael Ludwig