Great Circle Associates Majordomo-Users
(September 2000)
 

Indexed By Date: [Previous] [Next] Indexed By Thread: [Previous] [Next]

Subject: Re: uns?bscribe_policy
From: "Joe R. Jah" <jjah @ cloud . ccsf . cc . ca . us>
Date: Sat, 16 Sep 2000 20:57:47 -0700 (PDT)
To: Michele Tomkin <michele @ uclink4 . berkeley . edu>
Cc: majordomo-users @ GreatCircle . COM
In-reply-to: <200009151747.e8FHl6f02397@uclink4.berkeley.edu>

On Fri, 15 Sep 2000, Michele Tomkin wrote:

> Date: Fri, 15 Sep 2000 10:47:06 -0700 (PDT)
> From: Michele Tomkin <michele@uclink4.berkeley.edu>
> To: majordomo-users@GreatCircle.COM
> Subject: uns?bscribe_policy 
> 
> I find myself confused on how the "uns?bscribe_policy" configuration
> option works in Majordomo 1.94.5.  In particular, I wish to know if
> there is a way to require owner approval when a list member sends an
> uns?bscribe request.
> 
> My experience is that irregardless of how I set this option, a user
> can sucessfully uns?bscribe themself from a list.  
> 
> I have looked at the majordomo code and it seems that the uns?bscribe_policy
> only pertains when the requester does not match the subscriber address.
> However, the comments in the configuration file indicate otherwise:
> 
>  	# unsubscribe_policy   [enum] (open) <majordomo> /open;closed;auto;op
>         # One of three values: open, closed, auto; plus an optional
>         # modifier: '+confirm'.  Open allows people to unsubscribe
>         # themselves from the list. Auto allows anybody to unsubscribe
>         # anybody to the list without maintainer approval. The existence of
>         # the file <listname>.auto is the same as specifying the value
> **      # auto.  Closed requires maintainer approval for all unsubscribe
>         # requests to the list. In addition to the keyword, if the file
>         # <listname>.closed exists, it is the same as specifying the value
>         # closed. Adding '+confirm', ie, 'auto+confirm', will cause
>         # majordomo to send a reply back to the subscriber if the request
>         # didn't come from the subscriber. The reply includes a
>         # authentication number which must be sent back in with another
>         # subscribe command.  The value of this keyword overrides the value
>         # supplied by any existent files.
> 
> 
> The comments in "majordomo" describe a different situation though:
> 
> 	# Check to see if this request is approved, if the unsub policy is
>         # auto, or if the subscriber is the person making the request (even
>         # on a closed list, folks can unsubscribe themselves without the
>         # owner's approval).
> ....
> 
> 	# Either the request is approved, or the subscriber is the
>         # requester, so drop them from the list

This is an old philosophical story.  I believe One or more developers of
Majordomo believed that a mailing list with closed uns?bscribe_policy is
like holding prisoners; hence, the comments in "majordomo".  This is in
direct contradiction to what the developers provided: a choice of closed
uns?bscribe_policy. 

Developers' schizophrenia somehow pervaded the code;)  You can set the
policy to closed, but setting uns?bscribe_policy to closed has the exact
same effect as setting it to auto, more open than open;(  I call that a
bug;  some may call it a philosophical feature;) 

There are many legitimate circumstances in which a closed
uns?bscribe_policy is not only acceptable, but it may be necessary. 
Besides, there are legal ways to pursue spammers who may keep captive
audiences in their mailing lists. 

Any way if you wish to have the choice of true closed uns?bscribe_policy
in your Majordomo 1.94.5, you should apply the following patch:

        ftp://ftp.ccsf.org/majordomo-patches/1.94.5/majordomo.5

Thanks to your observation I modified it today to have the patch remove
that legacy comment, because it would no longer be true if the patch is
applied.  The old patch is now in: 

        ftp://ftp.ccsf.org/majordomo-patches/1.94.5/0ld/majordomo.4

P.S. Squishing that bug is only a side benefit of the patch; the main
     purpose of the patch is a more robust and user friendly confirmation
     procedure.  Read the patch before applying it.

Regards,

Joe
-- 
     _/   _/_/_/       _/              ____________    __o
     _/   _/   _/      _/         ______________     _-\<,_
 _/  _/   _/_/_/   _/  _/                     ......(_)/ (_)
  _/_/ oe _/   _/.  _/_/ ah        jjah@cloud.ccsf.cc.ca.us




References:
Indexed By Date Previous: Re: uns?bscribe_policy
From: Dan Liston <dliston@netscape.com>
Next: Re: error when sending in HTML
From: Richard Welty <rwelty@suespammers.org>
Indexed By Thread Previous: Re: uns?bscribe_policy
From: Dan Liston <dliston@netscape.com>
Next: error when sending in HTML
From: "Douglas Munoz" <dmunoz@meyersgroup.com>