Keyvan Nayyeri

It’s PDC time, and unlike the last two conferences, I haven’t been very active in publishing the news and my opinions about it. PDC used to be one of my favorite technical events before, but to be honest, the last two conferences were very disappointing for me.

As you may have noticed, these days I’m more like a listener than speaker, and would like to read others’ content rather than writing my own blog posts on different topics, however, this great post by Frans made me write my own opinions on Windows Azure and cloud computing stuff that are being presented in this huge atmosphere by Microsoft because I believe that this whole thing is not a big deal to have such a noise around it, and somehow Microsoft is exaggerating about the products that it has provided in this area.

In PDC 2008 and 2009 Ray Ozzie has been focusing on Windows Azure and the nature of cloud computing as the future of software world and Microsoft development technologies, and has been stressing that they’re trying to provide an integrated set of tools and a unified model of software development based on cloud computing for Desktop, Mobile, and TV devices.

I have no comment on the latter assertion as I also admit that Microsoft is good at providing integrated solutions and it’s one of the main reasons for its success in this field, but I think that I’m not at the same line with Microsoft and some fellow community leaders on the importance of Azure at the moment. In essence, I believe that Microsoft has missed many things in between that had to come before Azure in the mainstream and has quickly jumped to something that is far (maybe impossible) from becoming the main concern of common software development scenarios. I try to develop this and explain my reasons in the below paragraphs.

I know PDC as the main Microsoft event for developers in which they try to introduce and showcase the newest products for software developers, and have a set of different sessions in various categories to teach developers some general information about the new technologies to get them started, and send them to the community and their companies in order to promote these products and spread the word. PDC 2003 and 2005 were great examples of this strategy and I can recall them as two great conferences that had a huge impact on the Microsoft community and follower companies for years.

In my opinion Bill Gates was a great leader for Microsoft and one of the key points of the long-time success of the company was the fact that he was able to predict the main requirements of the software world to invest in them. He knew when to enter the video gaming field by a console, and when to enter the mobile world with Windows Mobile. On the other hand, he also had some mistakes and one of them was the inability to recognize the importance of online search and advertisement, and they’re still unable to catch Google in the field!

PDC 2000 and 2001 focused on the .NET Framework and new generation of Visual Studio and in 1-2 years they became the mainstream because there was a huge demand for a cross-platform technology for software development. In PDC 2003 and 2005 they introduced enhancements in the .NET Framework and new technologies such as ASP.NET 2.0, ASP.NET AJAX (formerly known as Atlas), WCF, WPF, and WF, and you saw that they had their own demand in the past few years.

Here I’d stress that the reason that all these products could succeed was the real demand that was there for them among ordinary software developers and companies for common software projects. The set of classic controls and wizards in ASP.NET 2.0 or ease of AJAX development in ASP.NET AJAX were some stuff that were becoming the common demand for developers.

But PDC 2008 and the current one, 2009, are far from that point or at least it’s hard to believe for me that the main topic of these conferences, Windows Azure, can have that demand in the coming years. I think that Microsoft leadership has the wrong imagination of the current situation in the software world.

The first question that comes to mind is about the importance of cloud computing in the main software development world? Putting some discussions that support the assertion and are all made by evangelists aside, do you, as some ordinary software developers, believe that cloud computing is the main area of the software world in the next 3-5 years? What is it coming to solve? What is it coming to help with? Are these concerns the main concerns of all software developers, teams, and projects, especially yours? In my experience, the answer is a simple “No”!

The second question is, if cloud computing is really necessary, do we really need such a big and complex structure to respond to our needs? Why can’t we use another solution such as those provided by Amazon or Rackspace? Why should we go this far and put all this effort in Windows Azure? In other words, even if cloud computing is a major concern, what are the advantages for Azure over its smaller rivals that are simpler and easier to implement? Even if there are some software projects needing a cloud solution, I do think that most of them will be satisfied with very simpler solutions.

The other challenging point is about the precedence of cloud computing in comparison with other requirements in the software world. How can you run a software on cloud when a simple change and redeployment requires a recompilation? Yes, I know that Dynamic Extensibility has been a field of work in the .NET Framework but it seems that it has more priority and precedence over cloud computing while it hasn’t received it yet.

The last point is somehow related to the politics of software. Considering that Windows Azure is going to host enterprise large-scale applications, we all know that most of such applications fall in two categories: strategic applications (such as those written for banks, militaries, etc), and big commercial applications that would be comparatively less. How many governments and banks are ready to host such strategic applications on Microsoft data centers? I think that you know the answer yourself!

Besides, this Azure wave is something more like an imposed discussion by Microsoft evangelists to the community than something really demanded by people, themselves. There were noticeable demands for say, new features in ASP.NET 2.0 or ASP.NET AJAX, but I can’t see such a demand except those encouraged by Microsoft evangelists.

Having all this stuff said, to me it seems that in the current economic situation the best thing to work on is some ways to automate the Software Development process, accelerate the coding speed, and reduce the overall costs of a project for companies. I think that this could be a better area to invest in than cloud computing.

As a side-note, it’s very interesting for me that Microsoft conferences have become a place saturated with evangelists to spread a noise about something to advertise it. Microsoft wants to force people to accept its products and technologies rather than attracting them to use. At this point, some readers want to find me and kill me!

The bottom line is that they can waste much money on holding such relatively expensive conferences, giving fantasy keynotes with 3D glasses, and asking people to cheer them (the people who have paid to come and watch this fantasy), but the software market in the world is demanding something else, and if they can’t respond to these demands, they’re letting their rivals take the power. What I see is a huge group of community leaders and talented people who prefer to stay at home, and even don’t watch these keynotes and follow the stuff because they’re becoming so boring.

Windows Azure should be a great technology for a company like Microsoft to host many of its own services and upcoming integrations on it, but I doubt that it can be something appropriate for the main body of the software world.