[122] in Back_Bay_LISA
Minutes of meeting of 12/2/92
daemon@ATHENA.MIT.EDU (Steve Harris)
Thu Jan 14 23:33:47 1993
From: etnibsd!vsh@uunet.UU.NET (Steve Harris)
To: bblisa@inset.com (Back Bay LISA Maillist)
Date: Thu, 14 Jan 93 20:31:38 EST
Apologies for taking so long getting this out. Blame it on the holidays :-)
MINUTES OF THE BACK-BAY LISA MEETING, 2 DECEMBER 1992
Attendees (from sign-up sheet):
. Doug Mildram mildram@xylogics.com
. John Heasley heasl@chpe.org (cannot read email name)
. Steve Harris vsh%etnibsd@uunet.uu.net
. Eric Conrad econrad@service2.omnet.com
. Dave Curado davec@ima.isc.com
. Ed Pereymer eperey@chipcom.com
. Margaret Metcalf mar@bbn.com
. Tom Heft heft@husc.harvard.edu
. Ed Anselmo anselmo@nic.near.net
. Douglas Alan nessus@mit.edu
. Dave Cartier dsc@epoch.com
. John Rouillard rouilj@cs.umb.edu
. Re'my Evard remy@ccs.northeastern.edu
. Rob Taylor rst@ileaf.com
. Stuart Freedman stuart@ileaf.com
. Roger Hale roger@sst.ll.mit.edu
Sorry if I misspelled your name or email address.
This was the meeting that almost wasn't.
When I arrived, several people were attaching meeting notices (which
somebody had printed up -- thanks whoever you are) to the doors and walls of
MIT Building E51.
We had all become accustomed with meeting in room 140, so when we found it
already occupied, we had to scramble to find some other place to meet. An
empty classroom nextdoor was found and appropriated -- it even had a
projector!
The meeting notice signs were modified to indicate the alternate room, and
attached sideways (or upside down) so the arrows would point in the right
direction.
By around 7:15, about 15 people had found their way into the room, and the
meeting got under way.
The meeting evolved into two parts, an administrivia discussion and a
presentation by Doug Mildram on his work with the UK Sendmail kit. After
which there was general free-form discussion.
ADMINISTRIVIA DISCUSSION
Tom Heft began by raising the question of how the meeting screwup had
occurred ("how many people got the email announcing the cancellation of the
meeting?"), from which he moved to a discussion of the organizational
structure of bblisa, and raised the question if the current structure was
viable.
The following is my recollection of the points Tom made:
In brief, bblisa is the result of a group of people who decided to try
to set something up. This group of seven (or so) people met several
times, came up with a general outline of how they envisioned bblisa, and
organized a few meetings.
From the beginning, the idea was that the group would be entirely
run by volunteers, that there would be an administrative group to plan
meetings, and that participants/attendees would volunteer to assist in
the running of the meetings and of bblisa itself.
It was also envisioned that, over time, different people would
participate in the admin group.
Tom was the only admin-group member at the meeting. Other members
were either occupied with other things, or out of the country, or unable
to be reached (email would bounce, or disappear into a black hole).
The screwup(s) regarding the meeting forced Tom to wonder if the loose
structure which has worked, more-or-less, up to now, was adequate for
long term viability of bblisa.
There was some discussion on this question. Points that I recall
(summarized in MY words) include:
Well, it's the week after Thanksgiving, and Christmas is coming up, and
it's not totally surprising that people may be hard to reach.
The system has worked okay up till now, let's not throw the baby out
with the bath water.
Perhaps if the bblisa membership were made more aware of the need for
volunteer assistance, meeting organization would not fall on the
shoulders of the same people all the time, and a bit more stability
could be achieved.
It was decided that a request for volunteers would be made, and that the
admin-group would attempt to standardize the procedures to be followed in
coordinating a meeting, so that volunteers could be put to good use as
meeting coordinators.
Followup: Summary of bblisa-admin meeting on 1/11/93
A meeting of several of the bblisa-admin maillist members was held
at John Harvard's restaurant in Harvard Square. The beers were
fairly good (I liked the stout best), but were served too cold.
A complete summary will be posted to the bblisa maillist shortly.
In brief, bblisa needs the support of its members to prosper. The
admin group will do advance preparation, solicit ideas for talks,
plan speakers, etc.
Each talk will need a coordinator, or "shepherd", to make sure all
the sheep arrive at the right pasture at the right time (who came up
with the name "shepherd" anyway? :-). The next couple talks will be
shepherded by current admin-group members.
We are looking for volunteers to help out with shepherding of future
meetings. If you are willing to devote a few hours over the course
of a month, please send email to: bblisa-admin@inset.com. If you
can, indicate which month(s) you would be able to shepherd.
SENDMAIL DISCUSSION
I'm not totally clear on this, but I think both Doug Alan and Doug Mildram
were going to speak. Anyway, Doug Alan begged off (lack of sleep), and Doug
Mildram gave us an overview of his experiences with and modifications to the
UK sendmail kit.
The UK sendmail kit is a set of tools (scripts, m4 code, etc.) for
generating sendmail configuration files. Until Doug's talk, I had never
heard of this kit.
Doug built upon the basic toolkit to make it easier to use, but, IMHO,
judging from the hierarchy of files Doug described, it's still takes a lot
of work to master.
Doug's kit is in the file ukcf.tar.Z. To ftp the kit:
. ftp 132.245.33.7
. cd pub
. binary
. get ukcf.tar.Z
(that's what my notes say, since I'm not on the internet, I cannot verify
that this works).
If you were at the meeting, you can decide for yourself how useful the UK
Sendmail kit, with Doug's enhancements, would be to your installation.
I hope we can schedule a more introductory meeting on sendmail.cf files. I
gather Doug Alan's talk was going to be more along this line; he has offered
to give the talk at a later date, I suggest we schedule it sooner rather
than later.
After Doug's presentation, there was some discussion about some of the fine
points of different versions of sendmail. Among those mentioned were
various revisions of Berkeley sendmail, Sun sendmail, and IDA sendmail.
The topic of "fuzzy name matching" generated some heated debate; proponents
of IDA sendmail said it worked well, Jeff (Kellem?) said there were problems
handling names with a middle initial.
There was some discussion of specific rewriting rulesets.
Doug Alan drew the (famous) rewriting set semantics diagram:
. /-> 0 -> resolved address
. /
. / /-> 1 -> S -\
. 3 -> D < > 4 -> msg
. \-> 2 -> R -/
(I've recently had to diddle the sendmail.cf files on our client machines,
so I am now intimately familiar with this diagram; at the time of the talk,
I couldn't have reproduced the diagram for love or money!)
There was some discussion of alternative mail senders. Jeff Kellem
mentioned that somebody was working on a public domain implementation of
"upas", the mail delivery agent in Plan 9. (But he also said not to hold
your breath!)
Somebody else mentioned mmdf, but (as I recall), most everybody agreed that
it required far too much effort to master and thus was suitable only for
environments where its additional capabilities were needed.
Other packages mentioned include "king james sendmail" and "zmailer".
There was some discussion about maintaining a nameserver database (my notes
are pretty vague here), apparently IDA sendmail has hooks for such a
capability.
There followed a discussion of modems -- somebody asked for recommendations
for reliable, inexpensive modems that could be given to employees to take
home (or into the field).
Somebody said to avoid Zoom modems, but later somebody said the Zoom modems
worked well for them as long as you did not attempt to connect Zoom to
Zoom!
Jeff Kellem said that Software Tool and Die (world.std.com) leased ATT
Paradyme modems, for dial-in only, in rack-mount configuration; leasing
makes sense given the rate at which communication technologies evolve.
Other modems mentioned include:
. Xycel (sp?) -- highly recommended by somebody
. US Robotics -- several people liked them
. Telebit (good but expensive)
. Practical Peripherals
. "Power Cubes" (are my notes correct here??)
. Telebit Cellblazer, for mobile computing
CONCLUSIONS
Now that I'm playing with sendmail, my notes make a little more sense to me;
I'm sure I would have gotten more out of the sendmail discussion now than I
did at the time. I still have a lot to learn.
Let's reschedule Doug Alan's talk as soon as possible. I will volunteer to
shepherd that one as I (now) have a vested interest in seeing it come to
pass.
I won't be at the Jan. meeting (I'll be at USENIX in San Diego), so I'm
looking forward to the summary of the meeting. Hope you take better notes
than I do :-)
--
.. Steve Harris - Eaton Corp. - Beverly, MA - vsh%etnibsd@uunet.uu.net
--
Mail for the `bblisa' list goes to `bblisa@inset.com'.
ONLY mailing list requests go to `bblisa-request@inset.com'!
bblisa administration messages go to `bblisa-admin@inset.com'.
