Volume 17 Number 81
                       Produced: Mon Jan  9  0:24:13 1995


Subjects Discussed In This Issue: 

Mail-Jewish, Future directions
         [Avi Feldblum]


----------------------------------------------------------------------

From: Avi Feldblum <feldblum>
Date: Mon, 9 Jan 1995 00:20:11 -0500
Subject: Mail-Jewish, Future directions

Thanks to everyone who sent me in suggestions. I've read carefully
through them all, and have tried to distill them into a series of
options below. The final format that we end up with may be some
synthesis of what is listed below. This is an updated call for comments
both from those who did not have respond yet, as well as from those who
did send things in to see if you think I have caught the basic ideas you
submitted in one of the choices below.

Several people indicated that they did not favor simply voting for one
choice, as there may be a few that they would be happy with and some
other set that they definitly do not want. The best way I know to handle
this type of situation is a technique I have heard called
"multi-voting". In a multi-voting situation, each person voting gets a
number of votes that they may cast, say 5 for an example. They can use
their votes in any way they wish, 1 vote each on 5 different choices, 5
votes on their favorite, 2 on their favorite and 1 on each of their
three runner-up choices. One then adds up the votes for each option, and
then drops the half of the options that got the lowest votes. One can
then discuss the remaining half, revote (often with a lower number of
votes) and repeat to get a final few. You then have a "regular" single
vote on the final few to pick the chosen one. Right now we have about 15
choices, so I would go for 8 or 10 votes in the first round

Several of the options below make reference to an editorail or advisory
board. Several people have sent in messages volunteering to be on such a
board. I have kept all those messages, even if I have not responded to
you on your email to me. Until the direction is defined, so that we know
what the board will do, I have not followed up on the board issue. Once
we have defined the direction, we can then figure out how to populate
this board and what exactly it will do.

I think that there really are two parallel but related issues here. One
is what volume do people want the list to aim for, the second is what
structural changes if any (very likely driven by the opinion to the
first question) should we implement on the list. The multi-voting is
applicable only to the second portion, the first can be a fairly
straightforward vote (I know they are not strictly uncoupled, there may
be options underwhich you want limited volume and other options below
where you might accept unlimited volume).

NOTE: THIS IS NOT YET A CALL FOR VOTES, THIS IS A CALL FOR REVIEW AND
COMMENTS. 

00) What is your view of the mail-jewish volume you would like to see? I
will express the size in current digest units, where each digest is
about 300-320 lines, which after subtracting the Table of Contents and
trailer portions is about 250 lines.
	a)  12 issues per week
	b)  18 issues per week
	c)  24 issues per week
	d)  32 issues per week
	e)  unlimited

Multi-vote portion starts here:

1) Short postings less than N lines (N is around 10 say) get highest
priority and go out within 1-2 days

2) All submissions that meet the halakhic and flame-free requirements
get posted to mail-jewish. No limits on the size of postings, no limits
on the number of issues per day. What this means basically is that the
current process for mail-jewish is continued, except that there is no
attempt to curb the growing volume. Moderation would be the same as it
is currently. The goal would be to try and turn around all articles
within 3-5 days, with the majority in less than 3 days.

3) A restriction of N posting (N about 5-8) per week for an individual
will be the only restriction added to 2) above.
	perl programming assistance will be solicitated to help
implement this solution

4) No limits on the size or number of postings by any individual member,
provided they meet the halakhic requirements. An editorial board will be
set up to monitor the postings, edit them and group them together by
subject for publication, the number of issues per day depending on the
diversity of the subject matter.
	a) This will probably reduce the average turn around time from
1-3 days to 3-7 days
	b) This will require a way to choose high level subject areas

5) mail-jewish should be limited to XX issues per week. To try and
acheive this postings will be prioritized by size. Most postings greater
that a certain size will have a summary of the article in the main
mail-jewish and the article will be placed directly in the archive area
for people who want to read it to retrieve it. If the author or
moderator feel that there is particular general interest for the full
list, the article will be forwarded to a editorial board for review. Two
limits may be defined. At the lower limit, a simple majority of the
board is required to post to the list, at the higher limit, 2/3 or 3/4
majority will be required. A new list will also be created, which will
get all the archived articles, so for those that want to receive
everything, they will get it.
	5a) Posting size to go to archive/board
		i)    50/100
		ii)   75/150
		iii) 100/200
		iv)  150/300

6) Mail-jewish regular issues should be limited to XX issues per
week. To try and acheive this postings will be prioritized by
size. Postings greater that a certain size will be grouped by specific
subject and archived into special issues.  A list of such special issues
will then appear in each regular issue, to allow quick reference to
ongoing discussion without cluttering mailboxes with an unreadable
volume of material.
	(6a vote same as 5a)

7) mail-jewish should be limited to XX issues per week. To try and
achieve this, maximum weekly or monthly limits will be placed on all
submitters. No more than YY lines will be accepted from any single user,
the value YY will be chosen/modified to achieve the weekly XX issue
goal.
	Note: this will require a perl programmer to volunteer some
time to help implement this solution.

8) mail-jewish should be limited to XX issues per week. To try and
achieve this more a rigorous editorial policy will be implemented [what
should this policy be?]. An editorial board will be convened to help
implement this policy.
	The editorial board will not rewrite articles, they will
help vote yes or no on articles and will help communicate this
information to the authors.
	This is likely to increase turnaround time on many articles
from 1-3 days to 3-7 days.

9) mail-jewish should be subdivided into 2 or more lists. [What is the
criteria for this subdivision?]
	I do not know how to implement this subdivision. If this is
the direction chosen, I will lean heavily on those chosing this
direction to help define how it should be divided.
	One suggestion: to (a)divide the list by broad topic and
(b)post the listings certain days of the week. For example: Sunday and
Tuesday would be halchik issues (questions, answers, comments, etc.);
Monday and Wednesday would be philosophical issues
(orthodoxy,politics,belief in creation theories, etc.) Thursday would be
your travel and kashrut issue and Motzei shabbat, "misc.".  The idea
here is that (a)it would give people more time to respond to postings;
(b)give you more time to edit and organize within the list (c) allow
readers to chose topic areas of the most interest if they had limited
time during the week to keep up.

10) mail-jewish issues will be divided into a "general" section and
topic-specific sections. Submissions will be held for a period to see if
enough submissions arrive on the given topic to make a dedicated issue
on that topic.
	this will reduce the average turn-around time from the 1-3
days currently to 3-7 days.

11) Each letter would be given YY of time (say a fortnight) for replies
to accumulate.  You would edit these into a single issue of m-j, which
would therefore be confined to lots of people's views on one subject.
Meanwhile, the same would be happening with lots of other subjects.
Then time YY later, the replies to the first replies would be edited
into another issue on the same subject, etc.  Every so often, a
miscellaneous issue would have to come out too to deal with new topics.
The result would be that the total volume of m-j would stay the same,
but each issue would be narrower in its list of topics.  I have noticed
that you already manage to do this to some extent, grouping letters on
one topic, but I suggest that a more extreme version of this could
constitute a sixth option.  It would have the advantage that a given
issue would be easier to skim for less-interested parties, and
conversely all the letters on a given topic would be easier to collect
by more-interested parties by printing a small number of issues.  The
disadvantage, presumably, would be yet more work for you!
	11a) Time to choose:
		i)	1 week
		ii)	2 weeks
		iii)	3 weeks

12) mail-jewish should be subdivided into 2 lists, a "main" list and an
"overflow" list. The moderator keeps the rate of posting to the "main"
list at the desired level by sending any messages judged more suited to
the overflow list, to the overflow list.  To reduce moderator workload,
the overflow list could simply reflect each message received to its
members, one message at a time, as it is received.

13) Move the focus of the list to a moderated newsgroup. Note: the
mailing list is aready a moderated newsgroup under the israel.*
hierarchy. This set of groups is being carried by uunet and propagated
to many of the Internet Service Providers (if your provider says that
they offer 8000+ or 9000+ newsgroups we are already there probably. I
know digex carries them).

14) Stop using the current digest format and send each submission out as
an individual mail message. This will allow threaded newsreaders to
bring together all common topics more easily if reading in a
newsgroup. It will allow people to see the Subject more easilly in the
Subject part of their mailreader, if reading it as a mailing
list. People not reading it in newsgroups who do not want to see many
messages in their mailer can use the listproc digest mode to get all the
days postings as one big bundle.
	a) As this will greatly increase the number messages sent out,
to keep bounced mail from overwhelming me, any addresses that generates
an error will be automatically deleted by the listproc software.
	b) As far as I can see at first look, this means that volume
and issue numbers get replaced by a date notation
	c) The archiveing would be done by daily logs.

15) Mail-Jewish should be limited to 4 or 5 issues per day, with one
   day's postings having no effect on another day's.  Currently, this
   would leave things as they are now.  If the volume increases a lot,
   however, it will keep things capped to what I'd consider a
   reasonable volume.

   Overruns (when there are too many messages in a day to go out at
   once) would be handled depending on the size of the overrun:

   1) In the case of a small overrun (say, 100 lines past volume #5)
      the last-received messages can be tacked on to the next day's issues.

   2) In the case of a large overrun (say, more than 100 lines), then
      the largest articles would be queued up for later posting.  When
      there is a day whose volume is low enough so the large article
      can fit in, it will be posted.  If an article sits in the queue
      for more than a week, it would be archived, with a pointer to it
      posted to the list.  Of course, before this process begins, the
      author should be notified, since he/she may want to shorten it
      for more immediate submission.

   3) In the case of a large overrun, where no messages are very long
      (something I don't think is very common), then I would treat it
      like the small-overrun situation, and put the last-received
      messages into the next day's queue.  This would also happen if
      the list remains in an overrun state after large messages have
      been queued by case (2)

   If, by the end of the week, it's clear that the system is getting
   bogged down with queued messages (meaning we've been in overrun
   case (3) for an entire week), then there would be one of four
   choices:
	- Put out extra issues to clear the backlog
	- Archive the messages (make the digests but only mail out
	  pointers to their location in the archive)
	- Discard the messages
	- Queue them for the next week

16) In my opinion, the high volume of posts on many topics tends to outlast
its popular interest.  That is, there is a point after which certain threads
become a semi-private discussion (o.k. argument) between a relatively small
set of posters.  My suggestion is to have a "mail-jewish-continued" list.
Once a thread warrants - either because of a high posting volume, or a low
general interest level - that thread will be switched from MJ to MJC.
MJC could be subject to more relaxed poster volume restrictions (or none
at all), and each MJC distribution should be composed of posts relating
to a single "continued" thread.

The advantages of this are:
a) the relaxed (or non-existant) poster restrictions on MJC would
probably satisfy those annoyed by any restrictions on MJ.
b) readers could opt to subscribe/unsubscribe to MJC as needed to allow
them to follow a particular interesting thread.
c) the topic-specific mailings of MJC would make it easy for a reader to
follow their threads of interest without weeding through uninteresting
(to them) threads.
d) a single over-flow list (i.e. MJC) would seem more workable and generate
less administrative overhead than trying thread specific secondary lists.
(A thought which had originally occurred to me.)

One hitch would be handling the likely "spin-off" threads that occur on MJC.
Posters could always submit shorter general interest spin-off posts back
to MJ.

Some general good points:

Participants should run their submission in spell-checker.

Moderator should establish a citation code for uniformity.
	(I not sure what this means, but I suspect that I would not be
sure how to do this and even if done, I worry about the added overhead)

 Participants should be encouraged to cite their quotes, as it is very
frustrating to a reader to have only a vague idea of the sources.

Moderator should not accept items which he believes are not in conformity
with halacha. These submissions should be send back with the following remark
"editor/moderator believes your submission is not in conformity with halacha.
If you believe that it is, please cite a support for your position from an
acceptable halachic source". I do  not believe editor/moderator could know
all the halachic sources to be able to judge himself, and therefore should ask
for "support" when item is in doubt.
	(That is largely what I think I'm currently doing, but probably
not as structured in asking for sources)

Moderator could declare, as a result of changing times, a special event,
or public demand, to have a special issue dealing with a specific topic. The
moderator could change the size of submission for such an issue.

I think that a summary of the various opinions on a topic (e.g., Sephardic
vs. Ashkenazic pronunciation or legal fiction) could be and should be written
for the members, after it was exhausted . Such a summary will bring the
various opinion expressed, cite them (e.g.,MJ17#88). Moderator should assign
this summary to volunteers from among MJyers. These summaries should be done
in a more"correct" way (i.e., English. citation, structure).

-- 
Avi Feldblum
mail-jewish Moderator
<mljewish@...> or feldblum@cnj.digex.net

----------------------------------------------------------------------


End of Volume 17 Issue 81