From the lack of interest I’m seeing in my feed list, I would have to assume I’m not the only one less than enthused about Microsoft’s Azure. About the only person I know who has perked up and displayed real interest is Nicholas Carr, and some of his interest is most likely because he wrote a book on cloud computing.
That Azure is a competitive strike at Amazon is a given. What’s missing in Ozzie’s statement, though, is the fact that Amazon originally rolled out it’s cloud computing services as a way of maximizing under-utilized server farms during the company’s quiet times. I don’t know if this has changed, and Amazon is now farming clouds deliberately, but I hope not—the original idea was quite sound.
I am not surprised that Ray Ozzie would urge Microsoft into the clouds considering his background, first with Lotus, and then with Groove. Groove, especially, was cloud-based. Probably one of the more sophisticated cloud-based applications at the turn of the century.
But as far as I know, Groove never did catch on in any big way. I imagine when Ozzie was hired at Microsoft and Groove folded into the Microsoft family, the expertise this acquisition brought into the company played a big part in the implementation of Azure, but it still doesn’t compensate for the fact that Groove never caught on. Not in corporate America, which is Microsoft’s bread and butter.
Amazon can probably make do providing data storage and services for the startups, which seems to be its primary customer. I can’t see Microsoft doing the same, not the least of which, startup and Microsoft are not words that necessarily go together well. Microsoft has always been a corporate company, pricing its products accordingly, and in doing so, giving both Apple and Open Source room to breathe and expand. Apple sold to the artists and mavericks. (Can I use maverick still, or has that term been trademarked by the Republican Party?) Open source managed to capture all the folks who fell in-between, though when Apple released Mac OS X, there was some platform straddling.
Seriously, I have to ask: can you imagine Citigroup or Bank of America farming any of its applications out to the Azure platform? How about Chrysler, or Blue Cross? Oh, there might be some IT in these big companies that will want to experiment around, but I’ve not met a big company yet that didn’t want to control every last aspect of its data. Several industry types can’t do something like cloud computing for most of their data—they would be prohibited by laws built to safeguard private information.
So, Azure isn’t a move to entice the corporates to the cloud (can’t be, really can’t be). It’s seemingly a move to entice the smaller guy, something that Microsoft has not shown itself to be particularly adept at. For one thing, the Visual Studio 2008 application that developers can use to build to Azure is pricey. Microsoft still hasn’t learned that rule number one is you don’t charge the developers money to access the development tools, if you want the developers to drive business to your platform.
Oh sure, Microsoft puts out a baby version of its different developer applications, but I’ll bet you dollars to donuts, the baby version won’t interface with Azure. And SOAP? Seriously, who does SOAP anymore? I thought the whole SOAP/REST thing was decided, a long time ago. We are talking 2009, not 1999.
I must admit to being a skeptic of all the recent cloud fooflah, not the least of which we’ve all seen what happens when a cloud server like Amazon has problems, resulting in several different startups being without service for several hours. I can respect that cloud computing allows startups to get a leg up, but I have to wonder: is that enough for a long-term sustainable business? Is there really enough business for another player in the game?
Once I made the decision to quit writing for a living (there was a living involved?) and return to consulting and development, I looked around at all of the existing technologies and asked myself what I should spend time on, in order to sharpen my development skills. Perhaps I’m a relic of times past (yeah, all of five years, ancient times), but I decided to spend most of my time working with Drupal and one or two other CMS, REST, a little RDFa, a touch of SVG and other programmable graphics tools, and maybe a smidgen of this or that, whatever strikes my fancy, and that includes AIR and OpenLaszlo and some of the other web/desktop platforms. You can never go wrong becoming as proficient as possible with CSS, the markups, the data sources (SQL/RDF/XML/JSON), PHP (or Python or Perl, maybe Ruby, always C++), JavaScript, and REST.
One could say that what I described is all that’s necessary for cloud computing, but there’s a whole new game when you have to create pie slices of your applications and throw them into a black box. It takes no additional time to learn to do cloud computing, true. However, it takes additional time to learn to do cloud computing well. I’m not taking that time, for any cloud. Not Amazon’s. Not Google’s. Not Safesforce.com (Salesforce.com?) Certainly not Azure.
If I’m wrong in my assessment, I’ll watch the rest of you fly past me, like birds on the wind. If I’m right, though, and today’s cloud is the same as yesterday’s Web 2.0—more hype than reality—I’ll already be well grounded when this bubble pops.