Following the discussion which has taken place this Summer about the
AMAST newsletter and mailing list, a few conclusions have been drawn,
and are submitted to your attention. Here they are:

(1) To start the regular series of the AMAST newsletter on Monday 10
    October. The sample issue n. 5, which is being distributed today,
    will thus close the `sample volume'.
 
(2) To publish a `Call for Editors' in the last sample issue (n. 5).
 
(3) To call the newsletter `AMAST Links'.
 
(4) To continue the AMAST list as a moderated mailing list. This means:
    together with the newsletter, distribution of the digest of e-mail
    messages to the list will continue (with the same frequency, that
    is: once every two weeks). The digest will have a slightly
    different format than that which I've been assembling so far. While
    still filtering out messages that prove irrelevant to the list --
    such as subscription requests, personal replies to the newsletter
    (viz. actually meant for me), etc. -- I shall put the other,
    relevant messages together into the digest, but also including an
    appropriate part of the respective e-mail headers. This will allow
    the processing of the digest by the usual e-mail handling programs,
    after a little editing, that is after removal of:
    (i)  the header of the message which conveys the digest itself, and
    (ii) the one-character prefix `>' which the mail system adds at the       
         beginning of every line that starts with the word `From'.
    The digest will also be available by anonymous ftp on the AMAST
    repository in Twente, and on the WWW (when retrieved by either of
    these means, the aforementioned little editing won't be necessary).
    In view of the new format, and to keep a consistent numbering with
    that of the newsletter, the digest numbering will restart from 1
    with the next digest -- more precisely, from: `Year 01, Number 01'.

(5) To introduce a `Views' section, starting from AMAST Links Vol. 01,
    N. 01 (1994). The new section is meant to host a discussion about
    AMAST, its goals, its ways of pursuing them, and anything else that
    proves relevant to this subject.

The 5th AMAST digest is enclosed. This collects the comments received
between 12.09.94 and 26.09.94 on the proposal for a new, permanent
AMAST list. As usual, the comments are presented in the chronological
order of their arrival. I look forward to receiving further comments;
if you have some, please send them to the following address:

                               amast@cs.utwente.nl
 
They will be included in the next digest, which is scheduled to be
distributed on Monday 10 October 1994.
 
Best regards,
             Pippo
                          ______________________
                          STATISTICS -- 26.09.94
                          ~~~~~~~~~~~~~~~~~~~~~~

The current statistics about subscriptions and no. of people
contributing comments on the AMAST mailing list are as follows:
 
(A) 51 contributed comments (8 twice, 3 thrice),
 
(B) 295 subscribed, of which:
    (B1) 257 to full newsletter and digest of comments,
    (B2)   4 to the full newsletter only -- no digest of comments thus,
    (B3)  24 with the ToC-only option and no digest of comments,
    (B4)  10 with the ToC-only option, and digest of comments as well.
 
(C) 5 info-only subscribers are temporarily put in the digest+ToC-only
    list.
 
(D) 25 unsubscribed explicitly.
 

[beginning of 5th AMAST Digest] ========================================
____________

Ildiko Sain:

The sample copies of the AMAST newsletter look very nice, and I find the
contents of them most useful.  
____________

Yamine Ait-Ameur:

I am satisfied with the amast newsletter format. 
____________

Peter Thiemann:

I think the format of the Newsletter is very good now.
You might consider splitting the list in two for a
further reduction of bandwidth:
list1:
	the newsletter as it is now, for people w/o WWW access
list2:
	just the TOC of the newsletter, perhaps with URLs added for
	the items or even right away in HTML (which amounts to sending
	a copy of .../SIAN04.html with absolute URLs) so that people
	can store that in their local html-area.
____________

Arie van Deursen:

The letter looks very good and informative.

Also, would it be a good idea to make www page entitled
"People working on Algebraic Methodology and Software Technology",
containing pointers to home pages of people involved in AMAST?
____________

D. Paul Benjamin:

I find the newsletter to be very useful, as many of its announcements
are not available to me in any other way.
____________

Maurice Nivat:

The latest 4th edition of AMAST news convinced me that it is a useful
thing but I wonder how you will manage to keep it going:it seems to me
a tremendous amount of work
____________

Martin Wirsing:

I find your new arrangement of the AMAST news very good and fully  
support your proposal!
____________

David Powers:

I am indeed interested in the set up of the list - we are
involved in a similar exercise with ACM SIGART and ACL SIGNLL - see e.g. 
URL: http://sigart.acm.org/ 
 
MINOR SUGGESTION:
 
By the way, I find it helpful to ensure that URL appear like this,
I have my URL viewer aliased to URL: and can just triple click and
paste a line like that to bring it up. (Most of those below are close
to this format - what is most helpful too is if they are placed
right at the beginning (one is) for each of the variants (ToC/whole/dir)
people might need.
 
I'd also like to ask what software you are using to generate your
html.  (I'm experimenting with various postprocessors for frame and latex
- at SIGART we prefer framemaker format submission yet get mainly latex.)
 
I also have some other suggestions which I will make below.

> [...]
> It's time to adopt a new communication style.
 
Moreover, charging is often on packets received (and the Australian
AARnet is moving that way - the British JANET has been charged that
way from the beginning I believe, similarly with the German net, etc.)
 
KEY SUGGESTION:
 
For this reason, I would strongly recommend that you include also html
of just the contents and for each of the individual items, just as you
do for .txt files.  (In fact you could hide the .txt file from your
URL directory processor - they are primarily for ftp users.)
 
This means that you get much faster response - it takes some time
for one of your sample newsletters to appear, and then I only looked
at some of the items.  Moreover some viewers (and xmosaic with the button two
cloning click) will reread the file - we only have a 128kb link to
the outside world (for the entire campus) so this is slow.  I also
went back and forward to/from the directory a few times, into another
sample, etc.
 
You'd obviously want to include next, previous and back buttons in
each of the subfiles.  Generally, the secret of effective Webbing is
to supply information in optimal sized chunks, which I would estimate
as 3-4 pages/screens.  One day viewers may even be able to preload on
a speculative basis (like caches and branch predicting pipelines)...
 
I'll look forward to seeing how AMAST develops.
____________

Teodor Rus:

I looked at all comments you had from various people and at the first
three editions of the AMAST Links and I think  that there could be just
one conclusion: the amasters (to use a term coined by somebody) want
a moderated  mailing list. You have already implemented it and the
first AMAST Links show that this could be a success. There are two
major concerns in the comments you received: (1) the messages to
be short but complete and (2) the access to be moderated.
Both of them are addressed and I have nothing more to add but this:
Please keep things simple, so simple that I can understand them.
____________

Graham A Stephen:

I found the sample copy of the newsletter very useful.
 
I have one, very minor, suggestion regarding the layout.  Perhaps
you could restrict the line width to a bit less than 80 chars (60
maybe) to help with readability.

____________

Carolyn Talcott:

I think a news letter for AMAST matters would be good.
My preference would be to limit conference announcements and such to
pointers to ftp/www sources.   My experience with the ETACS electronic
news letter is that there are never any conference announcements that
I haven't already seen (usually several times).  Including full announcements
just makes it hard to find the interesting bits.  
 
For those with access to WWW maybe a WWW page with a link to the current
news letter (and recent ones) would be good.  Reading it via a web browser
would make it easy to follow links of interest.  One could then simple
send a short announcement each time a new edition appears.  Having a TOC
in the email announcement is useful.
____________