>>>>> "BH" == Bill Houle <Bill.Houle@SanDiegoCA.NCR.COM> writes:
BH> In the 1.9x code, Majordomo has 3 "subsystems" identified in the config
BH> file: majordomo, resend, and digest.
I've always taken those to indicate logical groupings, and extended that
quite a bit for 2.0 so that you can restrict access to certain groups and
retrieve all variables in a group. A web interface could use them to
present a hierarchical view of config variables. My group assignments are
probably wrong in some places, though.
[internalizing -approval]
BH> This would be trivial to implement in 1.94, but it would involve a
BH> significant process change when people try to upgrade. Is this
BH> something we would want to do?
I think we should let it lie; the current confusion makes it easier to
define a consistent behavior that fits into a scheme that we want to use
for far future versions.
BH> As for 2.x, as long as (a) users still saw a consistent application of
BH> std alias names, and (b) there were no hardcoded assumptions about
BH> where to direct -approval and moderator type requests, then anything
BH> goes.
Only the defaults are hardcoded. (I guess that's what makes them
defaults.) If users want something really interesting they can always go
through the access control stuff to get it. (My goal is to eliminate all
of the code hacks required to do interesting things.)
BH> Preferably, one should be able to set the owner/moderator/etc addresses
BH> from within the config file itself, which I believe you have already
BH> provided for...
Well, owner- and -owner (I still haven't figured out which it's supposed to
be) has to be an alias so we can get bounces there. I have yet to write
the forwarder for -owner, but there will eventually be code there to look
for bounces and to forward to the real list owner, Right now it points to a
real address. (Said forwarder is on the TODO list.) It may be prudent to
have that forwarder also sit at -approval and -moderator (or whatever we
may decide) to give people enough mnemonic names to send to.
- J<
Follow-Ups:
References:
|
|
