In article <20011209110610.GB949 / nagual.toltech.dk>, "Leo Laursen"
<1858323m001 / mail1.stofanet.dk> wrote:


> Is the typo "sychronize" in your code? I corrected it in my test.

This is a mistake I've made about five time now ... I'm a slow learner
:-).

It's exacerbated by the fact that the default is for threads that throw
an exception to not abort, which means when you make a typo like that,
you don't see any output to tell you.  You just see what looks like the
thread's not doing anything ... when, in fact, it has presumably died.

I'm not suggesting that should be changed, just saying that it's
something to drum into your brain (well, mine, at least!), to remember to
do, at least during testing.  Ie, turn on Thread.abort_on_exception.

I'm surprised that someone hasn't jumped in and explained why the same
code that seems to run on Linux doesn't run on Windows.  I've not
experienced it myself, because I'm purely Linux here, but I know of at
least one other person who's had the problem.

My theory is that it's "operator error", because it seems like something
that would be given some pretty serious testing before releasing new
versions.