|
|
|
|
|
|
|
|
NetMonth, January 1989
******** **************************************************
* * *
* * The independent guide to BITNET *
* * *
* * January, 1989 *
* * *
* * Volume 3, Number 6 - 7 *
******** * *
* *
*** * *
* * * * *
* * * * *
* * * * **********
* ** * ********** ***********
* ********** ***********
* * ********** ***********
* * ********** ***************
****** * ********** ************
* * ********** ****** ******
* * ******** ********** **
* **** *************
******** * ** ***********
* * ** *******************
* * *******************************
* * **************************** **
* * ********************************
* * *************************** **
******** * ** *********************
* ** *************** **
*** * ** *****************
* * * ** ****************** **
* * * ** *************************
* * * ** ********************** **
*** * ** ***************************
* ** *********************** **
****** * ** ***************************
* * ** **************************
* ******** ** *************************
* ***************** ********************************
**** ****************************************************
****************************************************
* ****************************************************
* ****************************************************
****** ****************************************************
* ****************************************************
* ****************************************************
****************************************************
******** ****************************************************
* ****************************************************
* ****************************************************
* ****************************************************
**** **************************************************
1
* * ****** ******* * * ***** * * ******* * *
** * * * ** ** * * ** * * * *
* * * * * * * * * * * * * * * * *
* * * ***** * * * * * * * * * * *******
* * * * * * * * * * * * * * *
* ** * * * * * * * ** * * *
* * ****** * * * ***** * * * * *
* * * *
*********************** *******************************
Christopher Condon Editor CONDON @ YALEVM
Timothy Stephen Associate Editor STEPHEN @ RPICICGE
Craig White Associate Editor CWHITE @ UA1VM
June Genis Contributing Editor GA.JRG @ STANFORD
David Hibler Contributing Editor ENGL0333 @ UNLVM
Henry Mensch Contributing Editor HENRY @ MITVMA
Deba Patnaik Contributing Editor DEBA @ UMDC
Gerry Santoro Contributing Editor GMS @ PSUVM
Marc Shannon Helpdesk Editor HELPDESK @ DRYCAS
Glen Overby Technical Assistant NCOVERBY @ NDSUVAX
Gary Moss Point of View MOSS @ YALEVM
********************* Contents - Issue 28 *********************
*********
* *** * EDITORIAL PAGE____________________________________
* *** *
* *** * Bitnotes ....................................... 1
*** *** The Human Factor ............................... 4
* *** * Flames To: ..................................... 8
* *** * The Way BITNET Moves .......................... 10
* *** *
*********
*********
* *** * FEATURES__________________________________________
* *** *
* **** * JBH Online .................................... 13
* ***** * Announcing CHESERVE ........................... 14
* ****** *
* *** *** *
* *** ****
*********
*********
* * DEPARTMENTS_______________________________________
* *****
* *** * Headlines ..................................... 15
* *** * New Mailing Lists ............................. 17
* *** * Feedback ...................................... 19
***** * NetMonth Policies ............................. 20
* *
*********
*********************** Distribution: 3622 *********************
1
Page 1
*********
* *** * Bitnotes
* *** *
* *** * by Christopher Condon
*** ***
* *** * Yale University
* *** *
* *** * CONDON@YALEVM
*********
"Terminology of the Damned"
The tall man in the italian suit steps into the brightly lit
television studio. He smiles, hiding his upper lip with a
dark, bushy moustache. The reflected glare from his teeth is
blinding, if only for a moment. He is at ease here, he is
comfortable here. This is his environment.
His three companions on the stage are not so fortunate. While
they are obviously somewhat smarter than Him, they are not
nearly so polished, so confident. Especially not here, with
millions of people watching.
Well, they are only taping now, but people WILL be watching...
They wonder why they have come to this. A day in the
spotlight? A MENSA membership? Money?
Our host's smile fades slightly, imperceptibly as he glances at
the cards in his hand. One of the contestants speaks, and it
begins...
"I'll take BITNET SERVERS for five hundred."
He smiles and reads the answer. "People access this kind of
server when they are looking for someone--"
A bell rings.
"What is a list server?"
He frowns at the contestant. "In some cases, yes, but not
quite what we wanted. I'll finish the answer. People access
this kind of server when they are looking for someone's network
address."
A brief pause and a bell sounds again. Another contestant:
"What is a name server?"
1
Page 2
Out host glances at the judges and they shake their heads.
"I'm sorry," he says. The contestant looks very confused.
A buzzer sounds, and he explains. "The terminology was changed
recently and the term *name server* is no longer accurate. The
correct response would have been *user directory server*..."
*****
Computing, like many disciplines, has it's own language. Like
any language it has its own dialects, accents, and
characteristic inflections. This is especially true of
computing, where the technology moves so quickly that new words
are invented every day so we can avoid using new terms like
"that new thigamajig from Apple." If anything, we can confuse
the people who were starting to understand what we were saying.
BITNET has it's own dialect in the language of computing. We
have terms like "file server" and "list server" and "electronic
magazine." While these terms are often used in other areas of
computing, here they have a different (if subtly so) meaning.
I don't know if anyone is sure from where the term "name
server" came. It is one of those little pieces of our young
culture that no one bothered to document. Certainly it is
accepted, or else no one would understand what significance the
names NAMESERV@UNCAMULT and NAMESERV@DREW have.
Alas, the Internet has grown up with a very similar term of its
own, the "name service". Unfortunately, this type of service
does something totally different than the "name servers" to
which we are used. As a matter of fact, they have been reading
the literature that comes out of BITNET, confused by the
similarity between the terms, and thinking us complete boobs
for saying that a "name service" does THIS, when we were really
talking about "name servers" all along.
Makes sense? Of course not...
Now, the Internet term "name service" has little relation to
the function it performs. However, it has been suggested that
this confusing Internet/BITNET terminology situation is the
perfect opportunity to replace "name server" with something a
little more descriptive. Hence the phrase "user directory
server." It doesn't exactly roll off the tounge, but I didn't
make it up, so don't blame me. It appears now in BITNET
USERHELP and BITNET SERVERS (as well as the BITLIB online help
product).
1
Page 3
Come to think of it, "list server" is a little vague, too. I
mean, it could be a list of anything. Perhaps "mailing list
server" would be more sensible, or "the Eric Thomas All-Purpose
Server", or...
Virtually,
Chris Condon@YALEVM
1
Page 4
*********
* *** * The Human Factor
* *** *
* *** * by T. D. Stephen
*** ***
* *** * Rensselaer Polytechnic Institute
* *** *
* *** * STEPHEN@RPICICGE
*********
The pace of technological change being as dizzying as it is,
most computer users, however reluctantly, have grown accustomed
to watching their applications and hardware become obsolete
overnight. The thirty five hundred dollar IBM-PC compatible
that I sweated out of my department in 1985 lost 70% of its
resale value and was technologically passe a year after I took
it out of the box. Similarly, the technology used to provide
BITNET is already outmoded and has been outmoded for quite a
long time. Much better wide-area networking protocols than
BITNET's -- protocols that provide dynamic routing of messages,
wider bandwidths and faster transmission between machines --
have been in existence for some years. BITNET continues to
rely on its present technology for the same reason that I
continue to rely on my outdated office PC: keeping up with the
state of the art is expensive and if what is already in place
continues to provide a minimally acceptable level of service,
it's all the more difficult to justify the cost of upgrading.
Nevertheless, it is worth considering what a future, upgraded
BITNET might offer. Articles appearing during the past year in
the EDUCOM Bulletin and the Chronicle of Higher Education have
provided a glimpse. Imagine a network uniting researchers,
educational users from every level (primary and secondary
schools, community colleges, four-year and graduate
institutions), corporate centers, government agencies,
libraries, publishers, supercomputer centers, and professional
associations.
Like today's BITNET, this BITNET of tomorrow would provide
electronic mail and file transmission; however, it would do so
differently. One of the more noticeable differences would be
the type of terminal that you might use to connect up. Imagine
a mouse-driven desktop workstation with a large (17 inch or so)
high-resolution screen. The operating system for this machine
provides a Macintosh-style graphic-oriented user interface to
network functions (pull down menus, icons, layered windows).
If you've been following the press descriptions of the NeXT
computer, picture one of those.
1
Page 5
To use a network resource like Comserve, the user opens a
window on a directory of network services, moves a pointer to
the entry for Comserve and taps a button. A second or two
later, a window opens to a terminal session with Comserve.
Although mail, files and messages will still be sent between
individuals, the primary mode of interaction with network
servers will be through directly established terminal sessions.
No more commands sent in mail messages, no more one-line "tell"
messages hopping from node to node; dynamic routing, high speed
links and wide bandwidth mean that it will be practical to have
a terminal session with a server in the same sense that one has
a terminal session on one's university mainframe. In this way,
the net will function more in the fashion of today's PC
bulletin boards where users directly operate a program running
on a remote machine.
Getting a file from a server would be more convenient and far
faster than current methods: point to the file's name in a list
displayed in a sub-window by the server, tap a button and the
server transfers a copy of the file to your workstation.
According to estimates, transmission speeds of 10 to 100
million bits per second are possible (BITNET's present
transmission speed is 9,600 bytes per second between most
nodes) with the potential for speeds to reach far higher, so
there would be no appreciable delays. Transmission would be
from point to point, as it is in the Internet FTP procedure
today. This means no more waiting days for poorly monitored
intermediate computers to come back on line and transmit your
messages, no more waiting days or weeks for large files to make
it through clogged queues; in fact, no more queues!
The range of network services available to users would be
expanded considerably. Library materials -- papers and
journals certainly -- would be available from library servers
located at various points on the net. Professional
associations and other organizations would provide services to
users over the net. The availability of Internet-style
"Telnet" sessions (i.e., using the network as a vehicle for
establishing a session on a remote computer) would allow new
dimensions in cooperative scholarship: imagine being able to
collaborate with co-authors at distant sites by working on a
draft of a document simultaneously.
Ira Fuchs, Erich Bloch and others have described some of these
potentials in the Summer/Fall 1988 issue of EDUCOM's Bulletin
and I encourage individuals interested in the future of
academic networking to take a look at this set of articles.
The potential of the national network they describe is
tremendous in its implications for enhancing educational
excellence and for improving the quality of new knowledge
1
Page 6
produced within research centers. The major stumbling block of
course is money. The cost of such a network will be
staggering. According to a recent wire service item, Senator
Albert Gore introduced legislation last October that would
create a 400 million dollar fund to cover a portion of this net
-- presumably the backbone. Four hundred million is not a
great deal of money, but keep in mind that this is only a
rather small part of the total cost. Consider that NeXT-style
workstations will be at least five times as expensive as the
terminals or PCs most of us use to communicate over BITNET
today. Since many university departments cannot presently
afford to provide professors with simple ASCII terminals, the
cost of computing hardware will need to fall substantially
before many will be able to participate fully in tomorrow's
network.
While legislators deliberate and we wait for the costs of
workstations to come down, we should be doing whatever we can
to educate university, public and government officials about
the potential such a network can provide. It is going to be
difficult to sell the network using the standard arguments
about how it will increase productivity or sharpen America's
competitive edge in world markets. Unsubstantiated claims of
this sort were shoveled at the public by manufacturers of
personal computers during the early portion of this decade and,
having weathered that barrage, many Americans have lost their
innocence. Furthermore, many would rank BITNET's
interconnection with international networks (JANET, EARN, etc.)
as one of its greatest values, so it could be difficult to
explain how a network that allows researchers and scholars from
other countries to participate is going to create an edge for
the US. Joining these hurdles is the credibility problem
created by the recent media circus over network viruses. In a
focus segment, the McNeil/Lehrer Newshour depicted Arpanet as a
vulnerable network jockeyed by a group of exuberant Stanford
undergrads who appeared to sustain themselves solely on lockers
of candy bars.
Clearly, however, a strong base of grass roots support can
help. We need to let legislators and campus administrators
know that we want them to get behind academic networking. In
particular, we need to extend campus involvement in networking
as widely as possible. Those who haven't used networks are
usually impressed once they've seen things work and have a
sense of the possibilities. Each newcomer represents a
potential network advocate (obversely, each student or
professor who is denied access represents a lost opportunity
for support), so it is to our advantage to increase
participation in networking throughout the academy.
1
Page 7
Based on our experience with Comserve, I think that there is no
question but that disciplinary services can function as
significant magnets for involving new people in academic
networking. In fact, I invite anyone who has any doubts about
the promise of networking to take a look at what has happened
with Comserve over the past two and a half years. Comserve is
a service that provides educational and research materials,
information about graduate programs, course syllabi and so on.
It provides a method for users to search for bibliographic
references and it provides a 'white pages' of network addresses
of students and professionals in the communication studies
discipline. In addition, users can subscribe to about 20
online conferences covering divisions of communication
scholarship.
Comserve processed its 100,000th command around the first of
the year and is now managing 3,000 subscriptions to its suite
of conferences. It has been used by people from over 6,200
network addresses who represent 19 countries and 5 other
networks and it gains approximately 100 new users each week.
Now if Comserve was a system that dispensed nude photographs, I
wouldn't find these figures surprising. However, Comserve is
an academically oriented service and, as such, delivers some
relatively dry material. Nevertheless, this type of
disciplinary resource has proven sufficiently valuable to
attract quite a number of professors and students to the net,
and these users hail from a social science/humanities
discipline that is far from the forefront of computer literacy.
Clearly, one way to increase grass roots support for academic
networking would be to set up services similar to Comserve that
cater to other academic disciplines.
Today's BITNET was built from the voluntary labors of many
people and their efforts deserve our gratitude. Now there is
an opportunity for each of us to contribute to tomorrow's
BITNET. If this visionary network is to become a reality, we
need to get behind Fuchs, Bloch and the others and promote
academic networking to our legislative representatives and to
our campus administrators. Those in BITNET's administration
could do much to forward the effort by helping academic
disciplines create and sustain services similar to Comserve.
1
Page 8
*********
* *** * Flames To:
* *** *
* *** * by Craig White
*** ***
* *** * University of Alabama
* *** *
* *** * CHWITE@UA1VM
*********
Hello all,
I hope everyone's holidays were good and that this year will be
your best ever. It seems ironic that the season of good cheer
can be so trying because of unattended computers and the like.
A couple of times this month I have anxiously awaited files
that took so long to arrive. I have received many pieces of
mail regarding the Shapiro book. Several people wrote to tell
me how the book could be ordered. So for all of you who are
interested, I gratefully pass on the following from Michel S.
Perdreau who is a librarian at the Southern Campus of Ohio
university. "It may (?) still be available from Rand Corp.
1700 Main Street, P.O. Box 2138, Santa Monica, CA 90406-2138,
for $4.00 (US). It is also available as an ERIC document
(ED2690033) from most university libraries."
This months flame is not really a flame. It is a question
that, judging from my mail, seems to be important to many
people. The question is this: "How do I go about retrieving
files from the trickle servers of the SIMTEL20 archives?"
First off let me explain what these trickle servers are. On
the Internet there is a machine, SIMTEL20. The administrators
of this machine have seen fit to provide several directories of
public domain software, mainly of the IBM PC varieties but not
exclusively. This software is available via ANONYMOUS FTP.
For those of you who are unfamiliar with FTP (File Transfer
Protocol), it is a method of transferring files between two
hosts. It is the primary way to move data around on the
Internet. Because FTP works in a disk to disk fashion as
opposed to SENDFILE which stores files in your reader, you must
have a userid to log into the system you want to transfer files
from. This is where the anonymous part comes in. ANONYMOUS is
a special userid that anyone can use for the purpose of
transferring files.
Because most BITNET hosts do not have FTP access to the
Internet, the trickle servers were invented. When you request
a file, if the server that you are dealing with has the file,
1
Page 9
it sends it to you. If it does not have the file, it requests
it from the next server up. Each time a server gets a file it
keeps a copy for the next person who may ask for it. This way
requested files "trickle" down to the servers with each request
going only as far up the chain of servers as is needed.
In my opinion this is a very efficient way for servers to
operate. It keeps network traffic down and allows server to
evolve depending upon what files the users in a particular area
are interested in. In addition, the servers keep track of how
much they send to each node and there are limits on this.
Unfortunately, this can create a problem. The problem that
many people are encountering is that when they try to get a
file they are informed that the limit for their node has been
exceeded for that day and that they should try the next day. I
have talked with people who say that they ordered files right
after midnight and got the message then too. I have been told
of a similar situation that occurs with LISTSERVs that retrieve
files from SIMTEL20. It probably has something to do with the
different time zones.
There is a slightly different twist on this scenario with
LISTSERVs. You order a package of programs and at some point
during the retrieval of the package you are informed that you
have execed the limits for the day and you must try again
tomorrow. Many times it's very difficult to figure out what
files have and haven't been shipped. This can make it
extremely frustrating for users to get the software they want,
and in the case of new networkers, leaves them with a negative
impression of networks. Many people who have taken the time to
write to me about this are genuinely upset and have things to
say like "What good is a network if you can't get what you
need?" and other such statements often accompanied by crude
expletives.
Winding down, I think that the problems caused by this feature
of servers are serious, but with network congestion as it is,
it seems to be effective in reducing network traffic. I would
like some feedback on this problem and I will continue to
follow this for all of us. Meanwhile, I hope you will get a
copy of the Shapiro book and read. This is what it's card
catalog entry looks like here:
Author: Shapiro, Norman Zalmon, 1932-
Title: Toward an ethics and etiquette for electronic mail
Norman Z. Shapiro, Robert H. Anderson.
Publication: Santa Monica, CA : Rand, 1985
Material: vii, 35 p. ; 23 cm.
As always send your questions, comments and flames to
CWHITE@UA1VM.
1
Page 10
*********
* *** * The Way BITNET Moves
* *** *
* *** * by Eric Thomas
*** ***
* *** * Centre Europeen de la Recherche Nucleaire
* *** *
* *** * ERIC@LEPICS
*********
* Note: This editorial was extracted with permission from
correspondence between the author and the Editor. It is in
response to the Bitnotes column appearing in the November 1988
NetMonth.
Funny that I got the issue of Netmonth where you're mentioning
user directory standardization right after I finished the test
of the first version of the LISTSERV UDD...
Anyway, I take your point about the disorganized way in which
changes are made to BITNET. I would fully agree with you if
BITNET were a corporate or self-supporting network. But it's
not, and I therefore have to disagree.
The discussion groups you are mentioning are, quite often
(albeit not always) nothing but a group of people who have
self-appointed themselves as being responsible for Deciding
what "The Way to
|