Some time ago I wrote an article about Windows Azure. The reaction to the article was mixed - some agree with the “cloud” confusion and the lack of release dates, others (mostly Microsoft guys) asking for more time so the technologies will evolve; and last but not least people trying to explain how I don’t really understand Windows Azure and related technologies (to those, actually that was one of the points of the article - too much confusion).

So I find myself after six months since the release of Windows Azure looking at the same things as before. First, the “cloud” confusion. Everything running on the internet is now “cloud something”. I really appreciate Microsoft’s job of starting the real cloud race (let’s be honest, wasn’t a race until Microsoft came into play, it was more like a slow march), but I still believe they should take the lead and clear the sky of this everything-is-cloud-something confusion before it becomes a cloud jungle.

Second, although there are some advances in supporting other technologies in Windows Azure, like the PHP support released recently, we still don’t know the long term roadmap for supporting other non .NET technologies.

Third, I have to get back to the lack of information about when Windows Azure will be production ready and again, when are we going to have a Service Level Agreement (SLA). More and more companies (…)willing to move to a cloud solution are insisting on SLA. They need to ensure regulatory and legal compliance, integrity and security of their data in the cloud. How can anybody recommend a cloud solution where there’s no warranties of any sort, nor there’s a timeframe for having an SLA. Amazon already has for instance, a white-paper about creating HIPAA compliant medical data applications with Amazon Web Services. When will Microsoft follow? We don’t know, not any time son though. I know Microsoft has a roadmap for Windows Azure, I believe they should be more open and share with the public their long term vision.

Looking back at months of observing the development of Windows Azure and other cloud related technologies, there are a few things I noticed worth talking about. I’ll start wit the “cold RSS syndrome” where not too long after the initial release back in October 2008, when lots of Microsoft people were blogging about the new service, the RSS feeds of their blogs started getting colder and colder. This can be a sign they’re busy, or it’s just a sign they’re not that excited anymore. Even the official page of Windows Azure seems outdated.

Another trend emerging in the cloudosphere, is that every time Amazon or Google is making a move in this field, everybody looks at Microsoft and asking themselves “when Windows Azure will have this feature?”. It was the case with the recent release of Amazon’s MapReduce when the twittersphere and blogosphere were flooded with questions like “is there a Microsoft alternative to to this?”

Talking about alternatives, there’s a couple of services Microsoft must have for Windows Azure to be a complete solution and to dominate the cloud universe: a similar service to DevPay from Amazon, and a “Google for Domains” alternative. I hear this almost every day from developers and people getting into the cloud arena: the business model innovation is more important than technical innovation; how can I make money faster. Microsoft should realize that developers don’t look anymore just for cool tools to build cool services, they look for ways to quickly monetize on an investment they make and they will choose the cloud provider that will take them faster to the end-goal.

We’re in (almost) May, and there’s word from Microsoft they’ll announce the pricing of Windows Azure “this summer”. I really hope, they’re not only announcing the pricing, but actually announcing a release date (and a date for the SLA). I believe PDC in november will be too late to make the announcement, TechEd is too soon.

So, why Microsoft won’t release Windows Azure this year?

First reason is the reliability of the service. Since the beginning, Windows Azure has been plagued by server failures, service interruptions, buggy load balancers and security issues. Remember, Windows Azure is running in only one data center in North America, soon to run in another one, also in North America. Feeling safe already? I’d like to know what services from Microsoft are running fully on Windows Azure. Unless Microsoft is using Windows Azure in their commercial products, it is hard to believe any serious businesses will be built on this platform.

Second reason is the pricing model. I don’t believe they have it figured out. Period. It will take some time for them to come up with a straight forward, manageable pricing scheme. And remember, they also need to build good, reliable tools to record resource usage, no matter what the pricing model will be. This takes time and lots of testing (I hope), before playing with real money.

The third reason why Microsoft will not release Windows Azure this year is the scope of this endeavor. We know Microsoft is working on releasing the “on premises” Windows Azure solution together with the online version, which I think is a mistake. Also, they’re trying to match too many features to competitors’ features, instead of releasing a robust, basic, SLA backed cloud offering.

Microsoft does lots of things right: they are building a plethora of useful services, they are building the best development tools, they advertise, demo and evangelize their services like no other company, they opened the platform to other non-Microsoft programming technologies (PHP, Ruby) and they are releasing lots of source code. However, time is money, and in the real world people want to make money now, not waiting indefinitely for an SLA. So I do believe Microsoft should release something this year.

I’ll be in Los Angeles at TechEd in a couple of weeks, talking to people there about Windows Azure/Cloud Computing and blogging about it. You can also follow me on twitter @airimie, I’ll also tag #techedazure any interesting cloud/azure talk. Any suggestions or if you wanna talk to me at the event, send me an email.

Also, I’m still trying to find and showcase applications using Windows Azure. Unfortunately there aren’t many out there, so if you’re building an application based on Windows Azure and want to be showcased on this blog, please feel free to send me an email.