>>>>> "NH" == Nick Halloway <snowe@rain.org> writes:
NH> So, I wonder, is the extra complication of the Majordomo source code
NH> necessary for my application?
That depends on who's going to be using your application and just how you
expect it to be used.
NH> Are some people not able to set the Reply-To: line in their message
NH> headers?
Yes. Not a small number, either.
NH> Would I get more "please do this by hand for me" requests if I used the
NH> simple Perl script with formail?
That depends on how well you document it. The 1.94.x code makes things
harder for the user than they need to be. The 2.0 code is vastly more
complex code-wise (since is saves state in a database) but it's also
modularized so you could just grab code wholesale. The 2.0 method is much
easier for the end user to deal with.
In between is the patch by OXymoron that's available from the patch page or
the majordomo-users archives which is where the 2.0 style of token
acceptance came from. I'd think that stealing code from there would be
your best bet.
NH> Do some people not send mail from accounts where they receive mail?
Many, many people. Majordomo 2.0 solves this problem with an alias
database so that users can specify additional addresses that are considered
to be the same as the one they're sending mail from.
NH> Would Majordomo also work OK if you did subscriptions only from the
NH> header?
Majordomo doesn't process any commands in the headers. It looks there for
the user's address and (in 2.0) for the token that's being accepted or
rejected.
- J<
Follow-Ups:
References:
|
|
