PDC 2008: the Cloud, the Conference and the Data Mining

The PDC 2008 is finishing and there are some news for developers that involves BI too. The big news of this PDC is Windows Azure. Microsoft enters the hosting service market, but unlike other players, there is a new programming model “for the cloud”.

However, it has been disappointing the way Microsoft presented these new technologies to more than 6,000 developers attending PDC from all over the world. Ray Ozzie talked in two keynotes in two different days that seemed to be dedicate to non-so-technical press and analyst. This was the first PDC for Ray Ozzie and probably nobody explained him what kind of highly skilled developers the audience of PDC is. People who really want to get the big picture but also the inner details. In these keynotes, nobody really explained well what is the roadmap. Ok, some day every application will be a service in the cloud, but today we live in a different world. What are the intermediate steps in the years to come? What should be the transition from the present to the future? What is the vision of Microsoft in this migration?

I haven’t heard strong messages. I haven’t seen a clear roadmap. Like me, other attendees were frustrated. We had to understand what’s happening reading between the lines. Inferring information from breakout sessions without a precise guide. Not a good result, Microsoft. Believe me, these are the facts you should have emphasized during the keynotes.

·         WPF: it’s here to stay. I would have asked to attendees “how many of you still use Windows Forms?” – and then “Guys, I hope you have a strategy to migrate to WPF, we’re going there and we’re serious about it”. Visual Studio 2010 will have a WPF interface. Completely pluggable. This is an important message that goes beyond the Visual Studio extensibility. It’s a signal to the market. WPF is ready for big games. But only 2 minutes were dedicated to this.

·         Oslo: this is the most revolutionary thing for developers that will change the way they work, probably more than the transition from COM to .NET. And, guess what? No mention of Oslo in two keynotes. But, wait, there was a 90 minutes keynote from Don Box and Chris Sells. And what happened? They had to shrink their presentation to 60 minutes, talked about Windows Azure and didn’t mentioned Oslo. This has been premeditated. Why they did so? I would like to know the answer. The result: a lot of developers (may be the majority) are still wondering what is the purpose of “M”. It’s wonderful, but nobody state a clear message about its positioning in the long term strategy.

·         .NET 4.0: sessions around had good information about it, but even in this case, more information, strategy and future directions during keynotes would have been appreciated.

·         PDC 2009: yes, there will be another PDC on 17-20 November, 2009. Still in Los Angeles (I would prefer Las Vegas, anyway). But this “short” distance means that something that Microsoft is disclosing is still in an early stage. Otherwise it means that there are other new things coming soon. May be both. However, this could be not a good signal to the market – more like “wait another year before making strong investment on anything”. At least, to me it sounds like this.

So, this is my feedback about PDC 2008. Now, what’s coming for BI? There are several services including the integration of SQL Data Services as a data source to Reporting Services and an incubation project about Data Mining provided “by the cloud”. The first becomes interesting once some public shared data would become available on the cloud (imagine updated demographic data to be compared with internal sales data). The second is really interesting because it opens the doors to a wider use of Data Mining tools. But today we can only play with it, because there are no information about the cost of such a service whenever it will go in production. However, the message is that the cloud is relevant for BI too.