Wrapping up 2017’s extended CLR Week-and-a-half with some links to other CLR trivia


No original content today. Just links to some CLR trivia.

This concludes CLR week-and-half 2017. Sorry for wasting your time.

Comments (16)
  1. DWalker07 says:

    CLR week is not a waste of time.

    1. xcomcmdr says:

      I wonder who said it was ?

    2. I agree. They are educational, despite not being systematic or goal-oriented. It’s far better writing about your own dreams after all.

  2. pm100 says:

    All my work (well 95%) is CLR. These posts are always interesting. In particular you bring the perspective of where the CLR abstraction rubber meets the win32 road. I would welcome more

  3. Antonio Rodríguez says:

    Raymond linking to StackOverflow! What has the world come to!?

    1. pc says:

      Raymond (at least I assume it’s the same one) participates regularly on StackOverflow: https://stackexchange.com/users/485903/raymond-chen?tab=activity

      And Raymond, thank you very much for “wasting your time” constantly blogging for almost a decade-and-a-half and counting! It’s always a joy to read, even though I do little-to-no Win32 or CLR programming on a regular basis.

    2. Well, at least the first link is to a question answered by Jon Skeet.

      1. And the other by Raymond Chen himself!

    3. xcomcmdr says:

      StackOverflow is good.
      You just have to know how to use it.

    4. Antonio Rodríguez says:

      Well, it seems my sarcasm has passed under the radar :-( . Yes, StackOverflow is sometimes useful. And yes, Raymond seems to have an account there *and* use it. What I find amusing is that StackOverflow usually is a place where beginner programmers (and those not worth their salaries) come for a quick, copy-and-paste solution to their problems, while Raymond’s The Old New Thing takes the opposite approach, centering on the how and the why of things.

  4. Yukkuri says:

    If it is a waste of someone’s time they can just stop reading :V.

  5. Joshua Schaeffer says:

    Why do inherited .NET COM interfaces have to specify the entire vtable down to IUnknown or IDispatch rather than just the new methods that are added to the base interface like in C++?

  6. Neil says:

    But that begs the question, why is D the suffix for double, since C managed without it…

    1. Jonathan Gilbert says:

      Because a numeric literal can be ambiguous. “5” parses as an int, a float, a double or a decimal. The compiler will use the “lightest” of these types, so “5” is an int. If you want it to be interpreted as floating-point instead, “5f” is a float, “5d” is a double, “5m” is a decimal. Before writing this, I did some testing, and I had thought that e.g. “3.14” would parse to float because it doesn’t require too many digits of precision, but my testing proved me wrong; while any integer literal will be an int by default, any literal with a decimal point will be a double by default.

      1. Neil says:

        Sure, but any int can be cast into a double anyway, so it only matters if you’re doing manual overload resolution, at which point you should probably be explicitly casting anyway.

  7. BZ says:

    “M was chosen simply as the next best letter in decimal”. By that point, M is meaningless. It’s just a random letter. It’s not like a menu keyboard shortcut where the letter can be underlined.

Comments are closed.

Skip to main content