[sent to mh-workers(_at_)ics(_dot_)uci(_dot_)edu and
nmh-workers(_at_)math(_dot_)gatech(_dot_)edu]
The only reason MH needs a patch to run on Linux is because
it mucks about inside the FILE structures, something I consider a
MAJOR kludge.
... Over the years I have compiled MH for at least 5
different unix versions and I have always had problems. Usually,
they have to do with differences in the FILE structure. There may
have been a time when the speed of direct manipulation was important,
but the cost in portability is too high.
We need to replace any MH code that mucks about inside FILE. I
would be happy to help in this.
I think that's one thing that the nmh-workers project was going to work
on. I haven't heard anything from nmh-workers since December. Richard
or anyone, what's up? Have you got any code changes that could be merged
into the next UCI version (after the about-to-be-released MH 6.8.4)?
I'd guess that John Romine doesn't have the time to rewrite MH internals.
He may even think that it's a bad idea. John, if someone contributed a
patch to change the FILE stuff, any chance that you'd use it? What kind
of testing would we need to do to be sure it didn't cause any problems
on platforms where MH works now? (BTW, it'd be a pretty major change;
I wonder if it would mean MH 6.9.)
--Jerry