Valdis wrote:
And the exact gotcha involved, from 'nman rcvstore':
BUGS
If you use the "Unseen-Sequence" profile entry, rcvstore
could try to update the context while another nmh process is
also trying to do so. This can cause the context to become
corrupted. To avoid this, do not use rcvstore if you use the
"Unseen-Sequence" profile entry.
I removed that notice from the man page after 1.5 went out.
However, I shouldn't have. Ken added locking to the context
and sequences files nearly a decade ago. That addresses
file mangling during write, but not possible interaction with
other nmh processes.
What it doesn't mention is that the resulting train wreck
can mangle other sequences in the .mh_seq file as well.
There shouldn't be concurrent writes of context and
sequences files. So that kind of mangling shouldn't happen
any more.
Also, should we change that text to say "do not use
rcvstore without an external locking mechanism'? It's
perfectly save to use it from procmail with unseen
sequences as long as procmail does the locking for you, so
there's no need to scare off users.
I'll restore the warning and add mention of an external
locking mechanism if Unseen-Sequence is used. However, I
wouldn't call it perfectly safe, given the possible
interaction with other nmh processes that it mentions.
David
_______________________________________________
Nmh-workers mailing list
Nmh-workers(_at_)nongnu(_dot_)org
https://lists.nongnu.org/mailman/listinfo/nmh-workers