Blair,
I had another occurance of that 4 byte corruption problem. This time, it
showed up in a context menu - the command was mangled. I'm not sure if it
was relevant, but, I had just mutated the owning view (from a
multi-selection list box to a multi-sel enhanced list view). Correcting it
and resaving seems to have fixed it. The unit test that first brought my
attention to it also failed, and as before, it failed in D4 as well as beta
2.
I'm starting to suspect Novell as the culprit. I've been disconnecting the
machine from the network (to reduce exposure to hackers), and maybe that's
been causing a problem?? Anyway, _something_ isn't right. The unit test's
failing in D4 and a general lack of supporting evidence weaken my other
suspicions. Just in case though, I'll mention that I was momentarily
wondering whether #binaryStoreBytes and/or something about write streams
over byte arrays was involved. Put another way, I've pondered whether the
particular way of calling a primitive is causing problems that wouldn't show
up if the collection's initial size or the increments by which it grows in
size aren't just right..
As before, rebooting the machine "fixed" the unit test.
Have a good one,
Bill
--
Wilhelm K. Schwab, Ph.D.
[hidden email]
Locked
