The keynote speech delivered by the ever-enthusiastic Jason Zander, as usual, promised and delivered peeks into forthcoming Microsoft products. In this case, pre-alpha versions of both Visual Studio 2010 and the much awaited Windows 7.
What I saw of Windows 7 didn’t appear too different from Vista but its early days yet and the goodies people were there to see were in VS 2010. The coolest new feature, is support for multi-touch devices. This technology is way-sexy and I’ve been salivating over these devices since I first saw them demonstrated on youtube.
One of the new features was the ability for solutions to self-document. The IDE could generate activity diagrams and a sort of drill-down class diagram with weighted dependency links showing relative dependency coupling by the thickness of the lines. Be nice, if you export it as XAML for online application documentation via Silverlight or something. Hint hint Microsoft. All in all, pretty neat though.
What was abundantly clear, is that automated testing and software quality were high on the radar for VS 2010. A neat test tool runs application testing within a complete environmental capture session, tracking an application’s entire runtime experience. So that when faults occur, the application can be retroactively debugged from the telemetry recorded within the test environment. It’s a sort of “offline” debugging. The idea is to bypass the “it works in my machine…” brigade by making the event always reproducible.
I was thinking of having a tee-shirt printed with that. Anyway, also on offer were enhanced refactoring tools. To be fair, Visual Studio has always been behind Eclipse in out-of-the-box refactoring features. This effort seems to close the gap a bit. Lots of new features, long overdue, for the C++ weenies, including Lambda Expression support and apparently SharePoint development is easier too, wahoo.
Only two seminars were available today. I chose a seminar that turned out to be on .NET application custom installation and deployment. Something I’ve spent a lot of time doing. But, in fairness, there was the odd interesting nugget there, including some NGen demystification and clarity over deploying assemblies to the GAC. Apparently 3.5 SP1 has some interesting features affecting start-up performance of strongly named assemblies (Spring.Net users take note!).
My other choice turned out to be a very cool description of the, relatively new, dynamic languages supported by the framework (Iron Python and Iron Ruby). This was really well delivered by Curt Hagenlocher, a developer on the Iron Python team, who stood in for Dave Remy. Being a static, strong typing fan myself, (sort of like a Big Endian for the Dean Swift fans) I’m extremely dubious of the many “features” touted by the dynamic scripting proponents. That being said, there are times when dynamic language features are insidiously attractive. The ability to define class signatures dynamically and duck typing, (walks like a duck etc…) all have their uses, I’m sad to say. I feel positively heretical after admitting that. Strong static typing saved many a program and the attractive concepts in dynamic scripting are a satanic temptation to the weak willed.
More to come from Tech Ed EMEA 2008…
0 Responses to “Tech Ed EMEA 2008 Day One”