I can wind up with data sets which can take quite a while to load, sort,
modify, etc. Not having to restart because some null pointer popped an
exception is good for debugging efficiency.
Yeh. This sounds like the point. You work with all of this
monster big data, big program VLSI design stuff. So anything
that makes any of it more interactive is a "good thing."
Is this the driver?
Ayup.
Also, I sometimes wind up with systems which are network state
sensitive. This means that I may get one-off conditions which will be
hard to reproduce again.
Ugly. Sop you can use the patch and run to test against the
same rare occurrence?
Or to mark the code path as needing recovery code. I can immediately
add the recovery code, execute it and see if it worked before the
network state goes away.
-a
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
