Re: [bug-mailutils] [PATCHES] various collected patches for comsatd

[Damon Harper]

hi again -

On 29 Aug, Sergey Poznyakoff wrote:

> I have implemented first part of the proposed changes, roughly
> corresponding to mailutils-1.0-comsatd-echo-enhancements-1.patch and
> mailutils-1.0-comsatd-fixes1-2.patch. The changes are:

nice, glad they were useful.

i realised that to work around the atime problem, i can always apply a
workaround by setting the atime in the past via exec in my .biffrc
once the `exec-without-tty' fix from my other email is applied.  not
entirely elegant, but then neither is the atime patch.  definitely an
issue to be aware of though!

i'm still not sure why the old biff+comsat code doesn't suffer from a
similar problem - i guess i could trace through the mailutils libs and
find out exactly where the atime is adjusted but soon i'll be back at
work without a lot of spare time. ;)

for the sigchld issue, could you maybe at least apply your suggested
workaround of not setting it to SIG_IGN in comsat.c?  that should
solve the issue of messages from perl scripts.  i still wonder if it
might be worth doing more elegant sigchld handling, but just removing
the line that sets it to SIG_IGN would be enough i think.

> I'll try to figure out some elegant way of passing extra info to
> expand_escape. It should not take long, I'll let you know about
> the progress. Thanks again for your patches.

sounds good - or i would definitely consider the global variable
option as a simpler workaround - it doesn't seem very likely that
there's going to be much more to add that expand_escape needs to be
aware of; right now it's already getting more or less maximal
information for what it needs to know about.  that said, i might eat
those words next time i think up some new escape that should be
supported. ;)

-damon