Great Circle Associates Majordomo-Users
(April 1999)
 

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

Subject: Re: Approved: password
From: Jeffrey Goldberg <J . Goldberg @ Cranfield . ac . uk>
Organization: Cranfield University Computer Centre
Date: Tue, 20 Apr 1999 07:18:48 +0100 (BST)
To: "Jeremy H. Griffith" <jeremy @ omsys . com>
Cc: Majordomo-Users @ GreatCircle . COM, vlb @ cfcl . com, barr @ cis . ohio-state . edu, tibbs @ math . uh . edu
In-reply-to: <3792719f.2497070387@smtp.omsys.com>
Reply-to: Jeffrey Goldberg <J . Goldberg @ Cranfield . ac . uk>

On Mon, 19 Apr 1999, Jeremy H. Griffith wrote:

> >On Sun, 18 Apr 1999, Jeremy H. Griffith wrote:
> >
> >> 2. Delete all headers from the original message *except* To:, From:, Subject:,
> >> and Date:,
> >
> >I tend to like to leave in other headers as well.
> 
> Fine, as long as you don't mind having them appear as message text
> rather than as headers, and as long as none of them have continuation
> lines.

Huh?  When I do what I've suggested, I don't get the headers in the text.
I haven't really paid attention to continuation lines, but since almost
all Received: headers have continuation lines, I'm sure that I've had no
problem with these either.
 
> No, [the info-for-list-owners doc] does not describe it adequately; if
> it did, there would be few questions here about it.

Fair point.  I should also add that as "majordomo-owner" I always train
new list managers personally when setting up new lists, and I do step
through an example of manual approval with them (often to see that it
works with their email client).

> The reason they all failed was that the Received headers, and a few others,
> were more than a line long, and the mail client inserted returns (regardless
> of the Word Wrap setting).  The first continuation line therefore ended the
> header section as far as MD was concerned.  See any mention of this above?

The docs were written prior to the popular use of obnoxious mail clients.
Especially those that are not "what you see is what you send".  Such mail
clients are a very large headache for mailing list management.

> One of the respondents on this thread implied that in such a case, I should
> just use a different mail client.

Lets not get into that, ok?  Whether you "should" or not, majordomo and
majordomo docs will have to deal with such clients.  I think we can agree
on that point.

> The reason people have trouble with this is not because MD is broken, it
> is because the docs are broken.

You have convinced me.  I never read those docs carefully simply because
I "knew" how MD probably behaved.

There are a number of well known problems with the majordomo docs.  They
are fragmented in unpredictable ways.  Mostly they do not address the four
different kinds of people that they need to clearly.

   (1) The manager of the MTA the system goes on
   (2) The "majordomo-owner"
   (3) The list managers
   (4) The list subscribers

Often 1-3 are the same person, but that is not necessarily the case.
Person 1 will be able to make a pretty good guess at how MD must work, so
a cursory look at the approved header information will be enough to get
them to do the right thing if acting as person 3 (and as in my case fail
to notice all of the problems you pointed out in those instructions).

Anyway, as far as I know, the MD developers are very much looking for
people to help with a rewrite of the documentation and the documentation
for version 2.  I really wish I had some time to help.  Keep in mind that
just a few years ago, most of the people acting as (1) and (2) (and even
3), were far more clued in to how mail transport worked then currently.
The fragmented and limited documentation was closer to being adequate
then.  As someone's sig file said "The Internet used to be a bunch of
smart people sitting in front of dumb terminals.  Now ..."

So yes, the documentation needs an overhaul.  I think that we all know
that.  Care to help?

-j
-- 
Jeffrey Goldberg                +44 (0)1234 750 111 x 2826
 Cranfield Computer Centre      FAX         751 814
 J.Goldberg@Cranfield.ac.uk     http://WWW.Cranfield.ac.uk/public/cc/cc047/
Relativism is the triumph of authority over truth, convention over justice.



References:
Indexed By Date Previous: Re: Approved: password
From: Tim Miller <Tmiller@nethawk.com>
Next: Re: Problemas con lista.
From: Jeffrey Goldberg <J.Goldberg@Cranfield.ac.uk>
Indexed By Thread Previous: Re: Approved: password
From: jeremy@omsys.com (Jeremy H. Griffith)
Next: Re: Approved: password
From: Vicki Brown <vlb@cfcl.com>

Google
 
Search Internet Search www.greatcircle.com