At 7:49 PM -0400 9/5/96, John Orthoefer wrote:
>And once we add the diplomacy server we will be where Listserv is. A
>program which has totaly outgrown what it was designed to do.
Yes, but is what I designed it to do 4 years ago what it needs to do today?
>Majordomo is desgined to add/delete names to/from a list via RFC 822
>compliant mail.
Yes, that was a reasonable design then; is it today?
>Yes I implemented a CGI for Majordomo. But it works the right way
>(tm) it works via the one channel majordomo uses mail messages.
>
>lets not try and lose perspective here guys. This program was written
>because listserv is TOO complex. Brent wanted to get listserv running
>spent a day trying to get it to work and couldn't and said well I
>wanted to learn this perl stuff anyway. (at least this is what he
>told me at Usenix a few years ago.)
Yup, that's the way it was. I'd volunteered to manage 17 new mailing lists
for SAGE during its foundation, couldn't get listserv installed/configured
right, and decided I could do what I wanted myself in perl faster than I
could get listserv working. And I still think that was a correct decision
on my part; the questionable decision on my part was releasing the damn
thing! :-)
All that was 4 years ago, though.
If somebody posed the same problem to me today, would I solve it the same
way? No, probably not. The Web was in its infancy 4 years ago. The
number of lists and number of subscribers to each list is way up over 4
years ago. MIME is much more commonplace than it was 4 years ago. Bounce
handling is a much bigger problem than I recognized 4 years ago. Etc, etc,
etc.
If I were to start from scratch today, I'd probably start with a WWW
interface for user-level interactions (subscribing/unsubscribing), and
another WWW interface for list management (config settings, approvals,
etc.). I'd probably do the email interface a close second, but possibly
only for the user-level interactions; I might decide it's
better/simpler/easier just to do all the list-owner interactions through
the Web interface. And so forth.
The real question is, to what extent do folks want to keep pushing and
stretching and kneading and reshaping the existing Majordomo code base, and
to what extent do they want to start from scratch on something new that's
designed for today's environment and uses today's tools? There are strong
arguments for both, and there's nothing that says you can't do both.
-Brent
----------------------+----------------------------+------------------------
Brent Chapman | Great Circle Associates | 1057 West Dana Street
Brent@GreatCircle.COM | http://www.greatcircle.com | Mountain View, CA 94041
----------------------+----------------------------+------------------------
Internet Tutorials from the Experts!
Follow-Ups:
References:
|
|
