Technological (non)predictions for 2007
The early part of any new year sees its fair share of prediction lists. You know the sort of thing : 'In 2007, X will go mainstream' or 'Y will finally grind technological opponent Z into the dust'.
For grins, I have taken a different tack. Here is a list of things that I predict will not happen in 2007.
Item: No consensus will emerge in the debate about how best to use Web technologies to integrate IT systems.
People will continue to use the entirely human - but sadly flawed - gambit of looking for a binary opposition and then waiting for one of them to 'win'. For example SOA versus REST. There are just too many disparate interpretations of what 'SOA' means for 'SOA versus
The punchline to this non-prediction will only work if you are familiar with the following surrealist joke[1]:
Q. How many surrealists does it take to change a light bulb?
A. Two, one to hold the giraffe and another to fill the wheel barrow with brightly colored machine tools.
Punchline: Some of the SOA/REST comparisons out there effectively read like this 'which is better - a giraffe or a wheel barrow full of brightly colored machine tools?'.
The only sane answer to the SOA/REST question (other than a slightly depressed giggle) is the sadly under-utilized answer: 'mu'[2].
Item: The thick versus thin application development debate will go on and on and on with no sign of resolution. Every time somebody steps up and says 'you will never be able to do X just using a web browser', up will pop some clever way of doing it. The question will rumble on and on until finally it ceases to become a sensible question. Browser plug-ins that today are optional will gradually become de-facto mandatory and thus the 'thin versus thick' thing will become blurred into irrelevance.
Flashback: I remember when graphics were optional on the web. I also remember when cookies were optional. Over time, both of these have become de-facto mandatory. Some will see this as creeping featurism[3], others as progress. Mileage always varies.
Item: No single Web syndication format will emerge to dominate all the others.
I have come across some potential users of Web syndication technologies who have decided to wait until a 'standard' emerges for syndication. Personally, I do not see this happening in the same way that I do not see the day when all cars will be 'standard' from the same manufacturer.
In all aspects of competitive endeavor from biological evolution to capitalism, there are branches in the road at various points in time. The branches have labels like: extinction, dominion, subservience, symbiosis and stable co-existence. For Web syndication, the time window for the first three has passed, the fourth was not an option which just leaves the fifth: stable co-existence.
If you buy into the notion that evolutionary forces have peaks and troughs thanks to random events, then perhaps battle for dominion will be joined again at some stage in the future. I don't see it happening in 2007, or 8 or 9...
Metaphor with an analogy thrown in: He who waits for the perfect wave, rarely gets to surf.
That's it. Three predictions in the form of (non)predictions. A message in an e-bottle to be re-read (at least by me) in years to come if Darwin spares me.
[1] http://diveintomark.org/archives/2004/06/18/favorite-jokes#comment-5977
[2] http://en.wikipedia.org/wiki/Mu_(negative)
[3] http://en.wikipedia.org/wiki/Creeping_featurism