It seems that there are several ways to deal with this, all rough hacks.
If the list manager has system-level control, one could just disable the
config command -- this would mean that the config files would have to be
edited (horrors!), but the crontab solution also requires this. Another
option is to make the config files read-only.
However, another possibility occurs to me. Could one set up two
majordomo aliases, which would run two different majordomo setups? One
of these would have a hidden alias and would run the full majordomo,
while the other would use the standard majordomo alias but would run a
bowdlerised version with dangerous commands removed -- one could disable
config and newconfig, who, and any other commands one wished in this
version, while the head honcho would know the secret alias and could use
all majordomo commands at will.
To be specific, at present time the alias is normally something like
majordomo: "|/usr/majordomo/wrapper majordomo"
and my proposal would replace this with the pair of aliases
majordomo: "|/usr/majordomo/wrapper majordomo-x"
honcho: "|/usr/majordomo/wrapper majordomo"
so that mail sent to majordomo would be processed by a program called
majordomo-x with limited commands, while the person in charge and his
trusted cohorts would send their commands to "honcho" and would have
access to all majordomo commands.
I haven't thought this through in its entirety, which involves looking
at all the supplementary files and the role of majordomo.cf, but it
seems like a possibility that might help with this and related problems
that frequently appear on the list.
Of course another possibility is just to rename the commands without
telling anyone!
Bill Silvert
>From some recent postings:
> One other technique is to periodically re-write the .config files
>with a crontab function to over-write any changes they might make.
>I have some people that need to be able to approve postings and
>subscriptions to their mailing lists. I don't want them to be able to
>change the config files, however, because they are not technically adept
>and they'll just break something.
--
Bill Silvert, Habitat Ecology Section, Bedford Institute of Oceanography,
P. O. Box 1006, Dartmouth, Nova Scotia, CANADA B2Y 4A2, Tel. (902)426-1577
|
|
