March 30, 2005

In follow-up to the DataDirect XQuery Study

I recently received an email from Larry Kim of DataDirect asking if I wouldnt mind making a post regarding his recent study covering the usage of XQuery amongst a study sample set of 550 developers. Larry is undoubtedly aware of my "history" of posts regarding XQuery so in doing so I get the impression that he feels confident about these numbers and is ready and able to answer the questions of the community in regards to this report. While I undoubtedly have my own opinions on the matter I think Elliotte Rusty Harold, Dimitre Novatchev, and Mike Champion have followed up the post to the list quite nicely so I plan to take a slightly different approach in this posting, instead offering up some advice to the XQuery community by way of some analogies. But first, a summary of the XQuery study (a press release style overview can be downloaded here):

"We found that 52% of the 550 software professionals surveyed were already using XQuery -- with another 33% saying that they plan to use XQuery in the next 12 months!"

and then the question:

"Should XSLT fans be concerned?"

Ummmmmm.... do you want me to be honest or cover it with some sugar to make it easier to swallow? Times up! No sugar for you! ;) The answer...

Ummmmmm.... No. In fact if anything this should be more worrysome from the XQuery crowd as if these numbers don't pan out or seem at all "channeled and chiseled" it will cause wonder as to why the need if XQuery is all that it claims itself to be.

Actuality instead of providing criticism of this post I think the bigger need exists in providing an opportunity for DataDirect to quantify their numbers
while at the same time looking to real world projects such as Bruce D'Arcus and his XBiblio/Citeproc (of which I have joined the dev-team for further development of his existing code-base into other realms, domains, and applications (which I believe are actually the same thing depending on what generation you subscribe your hacker "slang" to :) to better understand how the development world is
either using, plans to use, or at very least perceives the space in
which XQuery fits into. If XQuery can be successfully used in ways
that maybe haven't been considered then it seems this should be the
task of DataDirect or any other company planning to provide products
to fill the assumed market need. Code speaks louder than marketing
numbers as its the code that is more difficult to counter, as long as
the code is solid and difficult, if possible at all, to counter with
the equivalent or better XSLT 2.0-based solution.

Without a doubt theres a business in XQuery. The only concern I have
is that too much time and money will be spent 'hyping' the media and
not enough time feeding the development community code. Doing this (hyping)
has already cast a pretty dark shadow over XQuery as the attempts have
come in direct attack of XSLT 2.0 in which a community, nay, an army
of XSLT developers have already banded together to take up arms
against. If I can suggest one thing in moving forward using a common
analogy:

"Keep your friends close, and your enemies closer."

To me the "enemies" from the perspective of XQuery are perceived to
live in the xsl namespace. In reality XQuery has become its own worst
enemy and really should be keeping tabs on its own a$$ if this analogy
is to be taken literally. And it really should! It would be a shame
to see XQuery die before its even born as its a good and necessary
language that fits well into the space of querying XML data stores
using a syntax similar to that of SQL such that SQL developers and/or
developers coming from a heavy procedural background can simply make a
few syntactical adjustments to then add XQuery to their areas of
language expertise.

Whether or not the "better than XSLT" claims have any real substance
taking this stance has consistantly brought opposition and at this
stage of the game opposition is not something to be sought after.
There is a bottomline that now exists in the XQuery realm and,
borrowing from a slightly modified Jerry Maguire, "Show Me The Code!"
should become the XQuery mantra from now until Tim Berners-Lee extends
his status of "Sir" to annoint "Sir XPath 2.0, Sir XSLT 2.0, and Sir
XQuery 1.0", sending them on their way into the battle of providing
XML Transformation, Query, and Path-based Referencing services to
version 2.0 of the WorldWideWeb. If seen as a team, Sir XQuery will
live on to fight many-a-good battles and be seen as a hero among many.
If seen as competition Sir XPath 2.0, taking loyalty to the family
birthright, will simply step aside and let Sir XSLT 2.0 take things
into his own hands providing only sparse details as to what "really"
happened out their in the battlefield claiming "its just too hard to
talk about... He was such a good and brave Knight that Sir
XQu-er-sniff, sniff Sir XQu-ahwahhahhahaaaaahhhhhahaahhh" as he
projects his sorry by wailing into tears, unable to say the complete
name of his "good ol' chum", providing a performance greater than any
other Oscar winner from years past has ever performed. The story
continues... "No one really understood why Sir XSLT 2.0 was often
heard chanting "Boom Boom Chee. Boom Boom Chee, We Will, We Will, Rock
You! Chee, Boom Boom Chee" while Sir XPath 2.0 bursts into laughter
each and every time, as if almost on que, but most definitely right on
the path. (ok, that one was officially a little too cheezy :)

I think their is some opportunity here to really take advantage of
this situation an put a really positive spin back into the XQuery
namespace. I just really hope thats what happens and not the opposite
as:

"I've paid my dues
Time after time
I've done my sentence
But committed no crime
And bad mistakes
I've made a few
I've had my share of sand
Kicked in my face
But I've come through

And we mean to go on and on and on and on

We are the champions - my friends
And we'll keep on fighting
Till the end
We are the champions
We are the champions
No time for losers
'Cause we are the champions of the World"(:WideWeb:)

... will quickly become part of the theme music to "The Legend of
XQuery" as it is told to our future hacking sons and daughters to
scare them into the well formed ways of the code (scare tactics are
necessary to ensure no one ever strays from the true and real path to
coding happiness -- Hey, if M. Night Shyamalan can get away with an
entire movie based on this premise then M. David Peterson should at
least get to keep his one little sentence, right? :D Did I just
reference myself in Third person? Oh hell, thatll never happen again,
I promise!

Cheers :)

[1] >> This quote seems to either come from the same study, or in fact this statement is really the base in which the marketing department and/or public relations team did what they are paid to do and that is find the very best way they can spin what they are given into something with marketing "flash" that will ultimatelly sell more product without binding yourself to flat out lies to do it.

To be honest I don't think there is anything contained in any of this that suggests DataDirect has done anything beyond adding some positioning flash to help make their up and coming XQuery product more marketable. I for one can not point fingers as I just as much if not more am guilting of finding and using catch phrases and keywords with a bit more "zing" to the senses to help drive home a point which in this case seems to be that developers are interested in XQuery and in fact they are interested to the point that 50-some% of XML-specific developers have already started using it in one form or another (You can add me to that list as there are most certainly times I have chosen to use XQuery if not for the fact that in that particular case an XQuery solution seemed easier but quite possibly because I simply wanted to learn XQuery and the problem seemed well suited to aid in this department.) If I havent stated my true position on XQuery strong enough in the past then let me try again. I like XQuery. It solves a lot the verbosity issues of XSLT and brings another functional-like language to the XML development world and above all I am a functional language proponent first, XSLT filling my functional XML programming slot. But this list also contains LISP/Scheme from an old school fundamentals perspective and projects like F# and to some degree COmega from an "on my list of high instensity learning projects", although there are certainly others.

In the end I think DataDirect hasnt commited the party foul we may have first thought they did and even if it turns out that they have overstepped a bit there is always the "golden rule" that fixes all party fouls... at least till the next occurence: Get us all drunk at the next conference and take pictures to then later blackmail us with ;)

Theres also another way to let things slide a bit:

Build killer products... They have. I'm happy.