by Michael S. Kaplan, published on 2011/03/05 07:01 -05:00, original URI: http://blogs.msdn.com/b/michkap/archive/2011/03/05/10137218.aspx
I've said it before: the Turkic folks have it right.
I mean, if we had these four letters in our alphabet:
then we'd case things the way that they do (dots with dots, dotless with dotless).
We really would.
Now the notion that we have now learned our lesson about the Turkish I and that we could do the right thing by default for Turkish is one that I hear from time to time.
I heard it over 10 years ago when I and many others agreed that the behavior could finally be 'fixed" in Windows XP. The major wininet cache bug that came out of the change was a really impressive tip of an iceberg (since tips are not generally expected to have as much influence on their own).
A bunch of people heard it before that when they created a default behavior for .Net that made .Net 1.0 an unusable security nightmare on Turkish and Azeri that took a division wide "everyone review code" break to produce a .Net 1.1 that didn't suck the same way.
Now there haven't been any extreme cases since then, but even these days at least six times a year I have someone come to me and bemoan some particular aspect of the native or managed behavior and how they don't understand why developers couldn't just solve the problem the way their customers did.
And then of course they explain their solution, which either doesn't really fix the bug because they misunderstood the problem, or it causes another problem. Or multiple other problems.
And then I realize that I am absolutely nowhere if my goal is truly to have all of this be intutive or at least understandable.
Probably just as well that goal didn't go in my commitments....
On a less serious but much more fun note, I'm heading out to Comicon today, and I will be there for Venice is Sinking tonight. And try to forget about all of this until next week.
bob on 3 Apr 2011 4:00 AM:
i dont no ??