[Bitkeeper-users] runaway "bk sfiles ..." after ^C of bk push (Win32 bk 4.3.1)

[Dave Hart]

I ran into a problem with bk 4.3.1 on Windows yesterday apparently
triggered by interrupting a "bk push" before it had gone very far (2
or 3 lines of output).  I noticed a while later that my keyboard
repeat rate was sluggish and variable, and discovered a never-ending
series of "bk sfiles" processes.  There was only ever one at a time
running, but each seemed to spawn the next, fouling my primitive
attempts to kill them.  After manually closing all my open programs,
it took about 5 minutes to log off my 2GHz 2GB laptop.  When logging
off, Windows eventually sets a flag that causes new processes to fail
to initialize.  Given these processes were each lasting seconds, I was
surprised at the delay.

This is not a proper bug report, of course, and I haven't spent time
trying to reproduce it, as I have several more pressing matters at the
moment.  Consider this a heads-up and a solicitation for guesses,
clues, or reports of seeing something similar.

Cheers,
Dave Hart
ntp.org developer