Racing at WLAC

Had a great time racing at WLAC on Sunday. Great family groove, I had some good runs, and I managed to walk away completely unscathed, but I'm a very stiff 42-year-old this week. I finished 13^th out of 16 in the slalom, but a lot of open racers didn't take timed runs, and after all of my excuses are properly accounted for, I'm quite comfortable in pretending a mid-pack finish. And I'm sticking to that story.

But seriously - it was fun to race with the LCB's (Tiger, Josh, WeS, Lynn, Hamm), and I'll definitely be back for the GS/Tight Slalom on July 20.

On Wine, Anna Nicole Smith, and Multiple Inheritance

Hi.

My Name is Mickey.

And I use Multiple Inheritance.

Now I know what some of you are saying, I hear it all the time:

• MI isn't worth the trouble
• MI is confusing
• MI complicates debugging and maintenance
• Horrified: What about the Diamond of Death?

These are certainly issues with the MI implementation found C++, as the MI in C++ had to deal with many legacy issues. But these are not problems with MI in OO languages in the general case. MI is quite useful in Eiffel - to the degree that no meaningful Eiffel system avoids multiple inheritance. Eiffel has several constructions that can be used by a class designer to accurately convey how any name ambiguities are handled.

My point of view is that avoiding MI because of C++ horror stories is a bit like avoiding all wine because of Anna Nicole Smith.

Now, although I would really like the C# spec to avoid MI as implemented in C++, I would really like to see non-CLS support for Eiffel-like MI in the CLI (if not the C# spec), so that languages such as Eiffel can more easily map to the runtime.