how to arrange a list with 'affiliate' members
I am trying to configure a list or group of lists to exhibit the following behaviour, but have been having trouble:
I run a list called 'discuss', which is a discussion list for members of our organisation. We would also like to invite a handful of people to be 'affiliate' members, but who would not have quite the same experience of the list.
We want these 'affiliate members' to be able to post to 'discuss', and any replies to their posts to be distributed to them. In addition we would like regular members to be able to insert a keyword to indicate that the conversation they are starting is to include these 'affiliate members'. But for all other list traffic, they would excluded.
The most obvious mechanism would be to subscribe the affilate members as full members of 'discuss' and use Topics - restricting them to receive only posts with [affiliate] in the subject. However, I can see no way of preventing them lifting this restriction from themselves, and in any case they would be obliged to remember to use the keyword if they wanted any replies.
The only other solution I could think of was to have another list - say 'open', of which 'discuss' was a member. This would allow control of who was an affiliate member, and would take care of inbound posts to 'discuss'. However, to allow replies back to the affiliate members would involve either:
*Having the 'discuss's reply-to UN-munged (which I am against because we have 100% non-technical people, and no replies would ever reach any list at all if they had to remember to hit 'reply to all'!)
or
*Having 'open' being a member of 'discuss' (in addition to 'discuss' being a member of 'open'). Presumably this is what is technically known as a BAD THING!
I wouldn't be at all surprised if I have overlooked something very obvious - in any case all suggestions would be very gratefully received!
Thanks
Do you Yahoo!?
Win a $20,000 Career Makeover at Yahoo! HotJobs
http://hotjobs.sweepstakes.yahoo.com/careermakeover
"Barnaby" == Barnaby Scott <barnabydscott@yahoo.com> writes:
Barnaby> I am trying to configure a list or group of lists to
Barnaby> exhibit the following behaviour, but have been having
Barnaby> trouble:
Barnaby> I run a list called 'discuss', which is a discussion list
Barnaby> for members of our organisation. We would also like to
Barnaby> invite a handful of people to be 'affiliate' members, but
Barnaby> who would not have quite the same experience of the list.
Barnaby> We want these 'affiliate members' to be able to post to
Barnaby> 'discuss', and any replies to their posts to be
Barnaby> distributed to them. In addition we would like regular
Barnaby> members to be able to insert a keyword to indicate that
Barnaby> the conversation they are starting is to include these
Barnaby> 'affiliate members'. But for all other list traffic, they
Barnaby> would excluded.
Why not just have two separate lists, and automatically subscribe the "discuss" members to the the "affiliate" list? No keyword needed, just address your post correctly. Reply-To will not go to an "unsafe" place (unless a discuss member decides to move a thread from the "affiliates" list to the "discuss" list, and forgets to change the address---but I don't see why this is more likely than forgetting to remove the [affiliate] tag).
Since the discuss list is apparently closed (ie, membership requires moderator approval), this imposes a slight burden on the moderator (I don't think vanilla Mailman provides a facility where subscribing to one list subscribes you to a second list automatically), but otherwise is just what the doctor ordered AFAICS.
Barnaby> The only other solution I could think of was to have
Barnaby> another list - say 'open', of which 'discuss' was a
Barnaby> member. This would allow control of who was an affiliate
Barnaby> member, and would take care of inbound posts to
Barnaby> 'discuss'. However, to allow replies back to the
Barnaby> affiliate members would involve either:
Barnaby> *Having the 'discuss's reply-to UN-munged
Yup, in this case Reply-To Munging Is Unquestionably Harmful.
Barnaby> (which I am against because we have 100% non-technical
Barnaby> people, and no replies would ever reach any list at all
Barnaby> if they had to remember to hit 'reply to all'!)
It's a shame that you and list admins who think like you didn't start lobbying their members and the vendors to fix their broken MUAs ten years ago, but that's no reason not to start now---it's not the last time this kind of case will arise.
N.B. Stop using "non-technical" users as an excuse. If they're typing replies, they have sufficiently well-developed muscle memory to handle this, too. The problem is that the users, quite rightly IMO, resist using reply-to-all because it pollutes their screen with unwanted junk addresses, and annoys fellow list members with duplicate posts. They "know there's a better way," and they are right---but their MUAs don't offer it.
So fix the damn MUAs, and everybody's happy. It's technically trivial: add a reply-to-list function which looks at List-Post first, and if it exists uses its value (only), otherwise acts like reply-to-sender. As default for _both_ list folders and personal mail, this works over 95% of the time for me, and I participate in a lot of lists where I want to make private replies, or include non-list-members as CCs. For 95% of users, I bet making reply-to-list default would work 99% of the time, and they'd quickly learn to use reply-to-sender and even reply-to-all correctly for the exceptions. Somewhat surprisingly to me, I do not make the mistake of posting private replies to a list (except when Reply-To is being munged, and I've fixed my reply-to- sender function to ignore Reply-To if it points to a list). YMMV, but this really looks plausible to me, and ought to be tried.
The only problem is getting lots of people who use broken MUAs to write to their vendors. My list members use sane MUAs, I can't help with this. ;-)
-- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.
Thanks for your reply. This was a solution I considered, but as you point out, members of one list do not have an automatic right to post to a 'sub-list'. As the moderator, I really do not want the job of subscribing/unsubscribing people to the extra list.
As for the Reply-To munging, you make some very valid points, and personally I am, in principle, on the non-munging side of the debate. However you seriously underestimate the technophobia of my members! These people are confused enough - many cannot even grasp the idea that you use 'Reply' to reply to *any* kind of email and, conversely, our archive is awash with replies that are actually new threads (presumably because they cannot operate their address book and therefore reply to an old post to save typing the list's address - despite then having to delete all the original message!) With the exception of myself and perhaps 2 other members, I believe they all actively hate everything to do with computers and only tolerate them because of what they can do. Not surprising really, when you learn that we are all hand-makers of furniture - so by definition have turned our backs on much that is technical.
So despite my natural inclination, I really do think there is a place for Reply-To munging, and with clientele like mine I'm afraid I really can't budge on it!
In desperation I am now working on a script which acts as a gate between the two lists. An address which aliases to this script will be subscribed to both lists, though only to one topic in 'discuss'. My task is to make sure that any one message can pass through the gate to the other list only once so that no loop could ever be established (even in the event of X-BeenThere headers being lost). Any tips on this project would of course be welcome. The principle seems pretty straightforward, but making it fool-proof is a bit harder.
--- "Stephen J. Turnbull" <stephen@xemacs.org> wrote:
"Barnaby" == Barnaby Scott <barnabydscott@yahoo.com> writes:
Barnaby> I am trying to configure a list or
group of lists to Barnaby> exhibit the following behaviour, but have been having Barnaby> trouble:
Barnaby> I run a list called 'discuss', which is
a discussion list Barnaby> for members of our organisation. We would also like to Barnaby> invite a handful of people to be 'affiliate' members, but Barnaby> who would not have quite the same experience of the list.
Barnaby> We want these 'affiliate members' to be
able to post to Barnaby> 'discuss', and any replies to their posts to be Barnaby> distributed to them. In addition we would like regular Barnaby> members to be able to insert a keyword to indicate that Barnaby> the conversation they are starting is to include these Barnaby> 'affiliate members'. But for all other list traffic, they Barnaby> would excluded.
Why not just have two separate lists, and automatically subscribe the "discuss" members to the the "affiliate" list? No keyword needed, just address your post correctly. Reply-To will not go to an "unsafe" place (unless a discuss member decides to move a thread from the "affiliates" list to the "discuss" list, and forgets to change the address---but I don't see why this is more likely than forgetting to remove the [affiliate] tag).
Since the discuss list is apparently closed (ie, membership requires moderator approval), this imposes a slight burden on the moderator (I don't think vanilla Mailman provides a facility where subscribing to one list subscribes you to a second list automatically), but otherwise is just what the doctor ordered AFAICS.
Barnaby> The only other solution I could think
of was to have Barnaby> another list - say 'open', of which 'discuss' was a Barnaby> member. This would allow control of who was an affiliate Barnaby> member, and would take care of inbound posts to Barnaby> 'discuss'. However, to allow replies back to the Barnaby> affiliate members would involve either:
Barnaby> *Having the 'discuss's reply-to
UN-munged
Yup, in this case Reply-To Munging Is Unquestionably Harmful.
Barnaby> (which I am against because we have
100% non-technical Barnaby> people, and no replies would ever reach any list at all Barnaby> if they had to remember to hit 'reply to all'!)
It's a shame that you and list admins who think like you didn't start lobbying their members and the vendors to fix their broken MUAs ten years ago, but that's no reason not to start now---it's not the last time this kind of case will arise.
N.B. Stop using "non-technical" users as an excuse. If they're typing replies, they have sufficiently well-developed muscle memory to handle this, too. The problem is that the users, quite rightly IMO, resist using reply-to-all because it pollutes their screen with unwanted junk addresses, and annoys fellow list members with duplicate posts. They "know there's a better way," and they are right---but their MUAs don't offer it.
So fix the damn MUAs, and everybody's happy. It's technically trivial: add a reply-to-list function which looks at List-Post first, and if it exists uses its value (only), otherwise acts like reply-to-sender. As default for _both_ list folders and personal mail, this works over 95% of the time for me, and I participate in a lot of lists where I want to make private replies, or include non-list-members as CCs. For 95% of users, I bet making reply-to-list default would work 99% of the time, and they'd quickly learn to use reply-to-sender and even reply-to-all correctly for the exceptions. Somewhat surprisingly to me, I do not make the mistake of posting private replies to a list (except when Reply-To is being munged, and I've fixed my reply-to- sender function to ignore Reply-To if it points to a list). YMMV, but this really looks plausible to me, and ought to be tried.
The only problem is getting lots of people who use broken MUAs to write to their vendors. My list members use sane MUAs, I can't help with this. ;-)
-- Institute of Policy and Planning Sciences
http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.
Do you Yahoo!?
Win a $20,000 Career Makeover at Yahoo! HotJobs
http://hotjobs.sweepstakes.yahoo.com/careermakeover
"Barnaby" == Barnaby Scott <barnabydscott@yahoo.com> writes:
Barnaby> However you seriously underestimate the technophobia of
Barnaby> my members!
Perhaps. I still think if they were given a chance to work with a system that was crafted to work for them, instead of against them, they would learn to use the technology quickly.
Barnaby> Not surprising really, when you learn that we are all
Barnaby> hand-makers of furniture - so by definition have turned
Barnaby> our backs on much that is technical.
So? Perhaps you don't know the etymology of the term "hacker". It's a self-deprecating analogy of one's programming to "making furniture by hacking at lumber with an axe." I don't think we hackers and you furniture makers are so different.
Still, you face the problem that by and large the software your members can get their hands on is going to be pretty rude.
Maybe you would be better off with a weblog or wiki-style website, which gives more control in some ways, and less in others. Certainly the setup cost would be higher than a mailman list, though.
Barnaby> My task is to make sure that any one message can pass
Barnaby> through the gate to the other list only once so that no
Barnaby> loop could ever be established (even in the event of
Barnaby> X-BeenThere headers being lost).
Well, anything that can strip the X-BeenThere headers can make any other tag disappear, too. The only foolproof method is to pass one message in each direction, then shut down forever.
However, one thing that is almost certain to appear in every message is a Message-ID, and unless you face an actively hostile agency, it probably will be preserved. So you could keep a database of message IDs that have been seen. If a message should arrive that doesn't have one, add one and record that. This is nowhere near foolproof, but in combination with the X-BeenThere strategy should do pretty well.
-- Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Ask not how you can "do" free software business; ask what your business can "do for" free software.
participants (2)
-
Barnaby Scott
-
Stephen J. Turnbull