Re: [xml-dev] Re: Native XML Interfaces

[Liam R E Quin <liam@w3.org>]

On Sat, 2013-06-01 at 01:45 +0000, David Lee wrote:
> This is quite interesting.
> Not pointing fingers just Pontificating from a glass house:)
> 
> W3C doesn't want to be in the API world (dont blame them,  #fail #DOM #fail)

That's changed these days, but for sure DOM did a lot of damage to XML.

Standards bodies are not really needed until there are multiple
implementations and a spec is in demand by users. Vendors benefit from
incompatibilities until the users insist on standards.

We did well with XML because there were implementations of SGML, and XML
was originally SGML for the Web. SoftQuad and EBT both had Netscape
plugins to show SGML, but there was little or no interoperability
between them. We each chose a different proprietary subset of SGML.

One problem with inventing things at a standards body is that standards
bodies don't generally have much (or any) budget for promoting specs;
they assume an external market will do that.

JSONIQ is one of the more interesting things happening in the XQuery
space right now; if XProc 2 happens it might also be pretty exciting,
especially if we can get EXI to work with XDM instance streams so you
can distribute pipelines reasonably efficiently, no XML parsing needed.

My own fear is that practical things like "the size of an image in
pixels" are ignored [1] in favour of function currying, forgetting that
the big win spot for XML is people who don't consider themselves
"hard-core programmers". Higher order functions are useful; the ability
to be a back end for Web apps is essential.

Liam

[1] note - some implementations, including zorba, do have functions to
get image size, EXIF and/or IPTC metadata and other information out of
images, but it's not something the WG will pursue.

-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Ankh: irc.sorcery.net irc.gnome.org freenode/#xml