Re: PHYS-L Needs New Owner and New Home!!

[Doug Craigen <dcc@cyberspc.mb.ca>]

Dick Smith wrote:
> Here's what the new manager needs and can expect:
>
> 1. Access to a listserving program and hopefully some savvy systems
> person to run and maintain it.  Phys-L currently operates from a
....
> This idea means that I don't really read the 50-200 daily messages, but
> rather only keep an eye on which folders seem to be filling up.
....
> Phys-L responsibilities wouldn't be a major increment.  That person
> probably doesn't take extended vacations away from his/her email, so
> error messages don't pile up for more than a few days.
>
> I suggest we discuss this matter publicly on the list, rather than using
> private email with me.  One topic I think we ought to explore is whether
> to move from a mailing list format to a web-based format, and to
> consider the advantages and disadvantages of each.  And as part of that
> discussion I hope we can identify the person who can make it happen.

Dick has provided us with some idea of just how much work he has been
doing for all of us, and I for one would like to express my thanks to
him.

Those of us on PHYS-L have been very happy with it.  On the one hand,
this may seem a rather trivial statement - if we weren't happy here, we
wouldn't be here.  But to have a 600 member discussion group with a long
history (I'm not even sure how long - I joined in 1992 and it was
already very well established) is no small feat.

When you ask contented people about making changes, the expected
response is "as little as possible" - and so far the response has borne
this out.  However, finding another volunteer willing to put in this
much work, and another institution willing to host it, may not be easy. 
We may not be able to just carry on the same format from a different
site with a different administrator.  Perhaps Dick would like to fill us
in on whether his private email has shown any nibbles to date.

I any event, Dick has requested some public debate on the issues of how
to proceed with PHYS-L in the future, and this is a natural time to do
so.  So I will begin by throwing some ideas into the ring, and I'm sure
there will be responses.

****

PHYS-L was built around email listservs.  This is an old, but tested and
true technology.  Perhaps the greatest benefits to this technology are
the ones mentioned a few times already: 
1) when reading the list from home we only need to log in for a minute,
download the day's messagages and log out 
2) we can make use of the categorizing etc capabilities of Pegasus,
Euroda, or other mail readers

USENET groups have too low a S/N ratio, and web based discussion groups
have neither of these benefits.

What are some disadvantages of the listserv?
1)  50-200 errors bouncing back daily for someone to sort out - in my
books that is a definite disadvantage
2)  time taken downloading and then discarding subject threads which I
am not interested in
3)  special considerations have to come up when I go on vacation from my
email, lest PHYS-L fill up the mail queue and get my ISP nasty
4)  the > >>> > stuff can hinder readability, especially given the
different places that various emailers do the word wrapping.
5)  its not obvious how to subscribe or unsubscribe
6)  not many people know about it or can easily find out about it

Actually, 5 & 6 could be seen as plusses or minuses.  They are probably
a large part of why the S/N ratio is good.  People generally join PHYS-L
because of a recommendation from a member.  The question is what level
of eliteness (is there such a word?) is desired.

Web based discussions can eliminate disadvantages 1-6, which is probably
why some people favour them.  They are however in their infancy, and
personally, I haven't joined any of them.  The web does offer a lot of
scope - given some development.  For example, one could bundle messages
together in various ways (by date, by subject) so that people could
still download what they want (and only what they want) in bulk and then
read it off line.  Cookies and/or membership names could be used to
prevent you receiving the same messages over and over in different
packages.  In fact, you could be have the server filter out any
undesired classes of messages (e.g. you wouldn't have messages from Doug
Craigen listed if you chose not to).

The really big disadvantage to this as far as I know, is that it isn't
off the shelf available.  Can anybody see applying for something like an
NSF grant to develop such a thing?

And while I'm talking money, how much subsidy is the host institution
putting up with the machine, technicians and bandwidth?  I'm not sure. 
But to put it into some sort of context, my order-of-magnitude figure
(based on Dick's traffic numbers) is that if I were to run a list the
size of PHYS-L as a part of my business, my ISP would probably charge me
another $2000 per year.   -   What I'm driving at is this, is it likely
that some amount of money would have to offered to a new host?  If so,
how would it be raised?  Pass the hat, sponsorship (e.g. AAPT, or
perhaps corporate), membership fees, others?

******

Just my $0.02 worth.  Perhaps it will generate some of that debate which
this group is so good at.