It's the start of a new year; what fruits will our computing orchard
serve up this season? This time last year the industry was excitedly
preparing us for how Web services would take over. Sun was gearing up
for their Sun ONE announcement in February after Microsoft had begun
filtering out information on what their .NET was really about. With
12 months now in the time bank, I can truly say I have not seen any
major change. Just a lot more people than usual, particularly vendor
companies, talking a great game.
A couple of months ago I introduced the notion that the whole
Web services revolution was a marketing gimmick, dreamed up to put a
new spin on old products. I purposely went a little over the top just
to test the water. I was interested in what you thought about the
whole Web services movement and I'm happy to report that you were not
backward in coming forward. I received a lot of e-mail and I
thoroughly enjoyed the dialogue with you over this. I can safely
report that approximately half of the responses agreed with what I
had to say; 35% thought I was missing the point and 15% believed I
was so far off base that I didn't have a clue what I was talking
about. Coincidentally, the majority of that 15% were sales and
marketing people, which I have to admit made me chortle somewhat.
My major difficulty with Web services, I believe, stems from
the name. I just hate the name. It says nothing and in my opinion
undervalues the whole movement. Web services is not a revolution but
more of an evolution - taking technologies and attempting to
coordinate them into one coherent solution as opposed to having
formats and standards fight for their market share. Take e-mail, for
example. Turn the clock back 20 years, or even as little as 10, and
you had e-mail systems that could not communicate with one another
without a specialized gateway that would translate the e-mail. With
the standardization of SMTP as the underlying mail transport
protocol, we can now e-mail the majority of the world without too
much worry. So, while I applaud the movement, I've spent the last 12
months searching for a much better name without knowing it. For me,
Having realized a couple of months ago that I was indeed on a
mission to find a new name for Web services, my standpoint became a
little different and I began trying to tease suggestions out of
people without them knowing what I was up to. I wanted an honest and
simple name, something that summed up the whole evolution without
using the words "Web" or "service."
I am proud to say I think I have found something that I
believe is better than what we are struggling with at the moment.
Sadly I can't take the credit; that would fall to Mel Stockwell from
IONA. In a lively e-mail exchange he uttered a phrase that I simply
had to pounce on: "middleware for the masses."
You may disagree with it, but I think Mel has stumbled onto
something big here and until I hear something better, I'll be using
this to describe solutions that are open, available, and flexible
(aka Web services).
* * *
Speaking of marketing...our "friends" at Microsoft are at it
again. In fact, this month they have incensed me so much that this
editorial is coming to you via StarOffice, not Word! (I have to
admit, I'm pleasantly surprised at how good version 6.0 is.) Late
last month I was sent a URL to a site that, at first glance, you
would be forgiven for thinking was a third-party developer's site,
but it was actually fully owned/operated by Microsoft
(www.gotdotnet.com/team/compare/). It's a bit of a shame, as that instantly puts us Java
developers on the defensive when browsing it, especially when you
read the opening gambit that they claim to have implemented the J2EE
Pet Store application using .NET technology and made it run 28 times
faster at only a fraction of the cost. Mmmmm...anyone that believes
that on face value is heading for trouble.
Being a little cynical I had to find out more. I spent a lot
of time looking through their documentation, their examples, their
white papers, and most of all their figures. Irrespective of their
conclusions, what annoys me is the fact that it's so obviously
weighted in Microsoft's favor. Had this been a third-party report,
then maybe that would have lent a little more credence to it, but, as
it is, you have to take it with a pinch of salt.
IBM wasn't too impressed with the findings either, it would
appear. They issued a counterreport detailing all the flaws in
Microsoft's claim. Microsoft had chosen to demonstrate their findings
on a number of different configurations, including Oracle, Sun, and
IBM. Although, as far as I'm aware, IBM is the only one to publicly
denounce the findings. Maybe Sun and Oracle are just too busy for a
When you read the IBM paper, "Setting the Record Straight"
/news/ibmnews/compreview4.jsp), it does indeed prove interesting
reading and on the whole offers a well-balanced argument to the
Microsoft claim that J2EE is 28 times slower than .NET. However,
that's not to say IBM's report is perfect - they make some
leaps-of-faith as well, although not as glaring as what Microsoft is
leading you to believe. I advise you to access the two reports, have
a read, and draw your own conclusions.
This was debated hotly on our mailing list (Straight
[email protected]), but sadly the majority of the posters are hard-core
Java developers, so we didn't have anyone with .NET experience to
argue against. This, I think, is the crux of the majority of debates:
NET developers are seriously thin on the ground. To that end,
Microsoft is very good at convincing nondevelopers that they are
developers and, therefore, I don't think we'll have a good debate for
a long time yet. Microsoft is good at empowering their community;
develop an MS Word Macro and you'll find yourself being called a "Web
services" developer! Since the word "developer" has a sexier ring
than most titles, a lot of people are happy to have this title
bestowed on them and begin to believe they are real developers.
Dangerous state of affairs.
This whole debate between .NET and J2EE could well be
academic. At the end of day, as one poster quite correctly pointed
out, we didn't consciously choose TCP/IP or HTTP. We chose them
because everyone else did. It was the beehive mentality with respect
to technology standards; we just swarmed to the most popular standard.
I believe the decision between .NET and J2EE will probably
come down to the largest user body. If that happens to be .NET, then
we'll start seeing a lot more crossover from the J2EE vendors to
ensure they don't lose their customer base. If it's J2EE, then
Microsoft will have to concede and start moving "C#" back to Java
again. (How does that saying go? "It will be a cold day in hell
before that happens.")
I look forward to what 2002 will bring to our Java lives. But
be warned, it's not going to be an easy year for us Java developers.
Our loyalties are going to be tested. We're going to be lured to the
"dark side" with tales of vast riches and connected, interoperable
systems. We have to resist.
We can't be assimilated to the Microsoft Borg. If we are,
then the computing industry will be a desperately bland and boring
place to work.
Alan Williamson is editor-in-chief of Java Developer's Journal.
During the day he holds the post of chief technical officer at n-ary
(consulting) Ltd, one of the first companies in the UK to specialize
in Java at the server side. Rumor has it
he welcomes all suggestions and comments!