Arjun Bijanki

Arjun Bijanki Arjun Bijanki

Niner since 2008


  • Arjun Bijanki: Making Sense of VC Intellisense

    evildictaitor wrote:

    Yep. Parsing C# is a walk-in-the-park compared to parsing C++. Just look at the respective size of standard for that (C++, C#). Don't forget that pasing a C# file might involve parsing say 5000 lines at most, whereas an expanded C++ file with the #includes can often be over 50,000 lines of code - windows.h is a good example of this.

    C# also reflects-in lots of it's information, whereas C++ has no such ability, so C# can get away with a tiny intellisense store, whereas C++ has an absolutely gigantuan database with a hundred different things reading and writing to it per second, which means it breaks.

    Microsoft last I heard was 100% committed to it's C++ customers, and for them to be crippling their own product in order to convert people to C#, which is a fundamentally different type of application for political reasons seems frankly ridiculous.

    I don't have exact values, but there are a huge number of developers using VS for C++.  They're the entire focus of the team I work on, so we're definitely committed to them.

    Like I mention in the video, the next release for VS will offer a vastly improved Intellisense, Browsing and Navigation experience for these developers.  We realize this is painful today, we released some perf work, and we're going to do it much better in the future.

    Also, the VC++ libraries feature pack for VS 2008 is in Beta right now.  Give that a try!

    There's a lot of cool C++ stuff on the horizon. Cool