A big advantage of nmh's code stability is that it provides
a natural place to stop. Should we consider starting with a
clean slate? An extensible architecture and modern language
would make enhancements and maintenance much easier. And
safer, e.g., nmh still uses mktemp because replacement is
just too painful.
Here is my $0.02 on that topic.
I have no objection to anyone doing that. If people want to do that, hey,
more power to them! It's not like I have it in my power to stop a group
of people from doing that, even if I wanted to.
But ... I'm not going to participate in that myself.
I can't speak for anyone else, but my undergraduate days when I could
spend weeks programming on a project just for fun are long gone. I've
got a full-time job, a mortgage, a kid ... and I barely have enough time
to balance my checkbook, much less do programming for _fun_. When
I work on nmh, I have to give up something else, like watching TV or
a movie with my family, exercise, or sleep (for the record, all of my
significant programming work on nmh has been as part of larger projects
at work).
When I think about the amount of work involved in starting a nmh
replacement from scratch ... well, it's just exhausting thinking about
it. And I have to be blunt here ... while there are plenty of people
who have good _ideas_ about nmh, what we seem to lack is an abundance
of people who have the time and/or knowledge to _implement_ those ideas.
Sure, we can talk about Grand New Ideas for New NEW MH until we're
blue in the face ... but who's going to write that code?
Me, I know that while it may be a hack, it will take less _time_
to change nmh to make it do what I want ... and time is what's really
in short supply for me.
And for the people who think nmh has reached the pinnacle of software
development ... well, let me ask you: Why do you care what happens
with nmh? You can still download MH 6.8.3, and every version of nmh
ever released. If you think nmh 1.3 is perfect, well, if someone adds
IMAP support, a native Java interpreter, or three copies of Emacs to
nmh ... so what? The perfect nmh version 1.3 (or whatever) will still
exist for people to use; I don't see anyone suggesting that we purge
the Internet of all of the old ancient copies of nmh.
Of course, it won't be easy to reach consensus on what
should and should not be in it, what language, whether or
not to use middleware, etc. But going through that process
should give the gsoc student "more exposure to real-world
software development scenarios."
A giant pile of ancient, poorly-documented legacy code? A cranky and
brittle userbase? Sounds like the most real-world software development
scenario imaginable! :-)
--Ken
_______________________________________________
Nmh-workers mailing list
Nmh-workers(_at_)nongnu(_dot_)org
http://lists.nongnu.org/mailman/listinfo/nmh-workers