A guide to establishing mainframe computer-mediated communication between a host and a remote site.
The growth of CMC at universities has also made possible the interconnection of local CMC networks at groups of educational, industrial and governmental institutions. Connected CMC networks can be thought of as high-speed phone lines that allow text and data to transfer almost immediately from one mainframe to another. Scholars have utilized CMC networks to undertake cooperative research ventures with others not only on their campus but also at universities in other cities and countries. A common transaction involves an instructor who composes a text, sends that text or data to someone else, that person edits it and then sends it back to the original author.
Once a researcher physically leaves the immediate vicinity of their home university, however, accessing that university's mainframe system for CMC may become troublesome. Often the portable hardware that is compatible with the home mainframe may be hard to find and/or expensive to use.
This article reports the findings of a case study that tested the viability of establishing CMC between a "home" site, The Pennsylvania State University (PSU) in the U.S., and a "remote" site, Sussex University in England. The study focuses on the use of a JANETBITNETI gateway. Our team consisted of a communication researcher (who went to Sussex) and a computer-systems analyst (who remained at PSU). The primary purpose of the article is to provide the reader with a practical guide for establishing and coordinating CMC between a remote location (the place where the scholar will travel) and a home location (the place where the scholar resides at and normally conducts research).
Remote Hardware Configuration
In deciding which is the best configuration for accessing the home mainframe, the main objective is usually to find the cheapest and least burdensome way to continue e-mail transactions. At least three options for hardware configurations may be used at a remote location to access the mainframe at home, subject to the availability of the necessary equipment:
(1) A computer terminal or microcomputer connected to a modem can communicate directly and asynchronously with the home mainframe. However, the telephone charges for long-distance calls can be prohibitively expensive, especially between the U.S. and Great Britain, so we did not pursue this option.
(2) A terminal or microcomputer with modem can be used to access a CMC network that is near the remote location. This network can then be used at the remote site to log on to the home mainframe. This feature is available with some networks, such as ARPANET, but was not available on the PSU mainframe at the time of our case study.
(3) A terminal or microcomputer with modem can be used to access a user account on a remote mainframe, which has a CMC-network link to the home mainframe's CMC network. This is the configuration option used in our study.
If options 2 or 3 are being considered by the travelling researcher, two essential criteria should govern the decision. The first is to find a mainframe that is both close to the remote location and which accesses a CMC network that is capable of communicating with the home mainframe's CMC network through some combination of inter-mainframe networks (BITNET, UUCP, EARN, JANET, etc.). In our case, we had only to locate one remote mainframe because the communication researcher was staying in one region of England. Other researchers, however, may have to identify multiple remote mainframes if they move around frequently.
A second criterion is to find a remote mainframe, if possible, that has an operating system and CMC software similar to those of the home mainframe, as well as consultant support for CMC usage. This is important because when one learns a new (or different) technology, there is always a period in which there is no real productivity except in terms of learning the system.
Access to the Remote Mainframe
Once the remote mainframe site is selected, the next step is to gain access to a temporary CMC-user account on the remote mainframe for e-mail purposes. Luckily, the chances of being granted such an account with a modest amount of disk space are pretty good within academic environments. The best approach is to contact the system manager at the remote location. Addresses for such personnel may be found by checking online lists of mainframes that connect to CMC networks, or by posting inquiries to a shared BBS such as the UUCP NETNEWS system. If all else fails, we recommend trying to send an e-mail message of inquiry to the target system's address; there is a good chance that someone will see it and relay it to the system manager.
Once a contact is made, the researcher should request a temporary computer account. Adding one user to a system on a temporary basis is not usually a problem. The ultimate decision often depends on pragmatic matters such as the availability of system capacity (disk space) and how the paperwork for setting up the temporary account is handled.
Once a user account on the remote mainframe has been acquired, a researcher can travel to the remote location and begin using the remote system immediately. Before the researcher leaves, however, he or she should make plans to have any e-mail that arrives at his or her host mainframe address forwarded to his or her remote mainframe address, either manually or through an "autoforward" mechanism if it is available.
The JANET-BITHET Connection
Two e-mail terms that we refer to in this section are note" and "file." A note looks like an inter-office memo and is normally used for short messages. It is not usually stored in the memory of a mainframe. However, a file is normally used for papers, chapters and other, more extended messages, and it is usually stored on the mainframe system for subsequent editing or referencing. Sometimes, because of compatibility problems in CMC software, only notes can be sent between two mainframes. In such cases, files can be embedded within notes for transference.
In order to facilitate the JANET-BITNET connection for our case study, we pursued several steps. This process took very little effort. First we simply generated a standard cover letter in an electronic note that described who we were and why we wanted to facilitate CMC between PSU and a British BITNET site. Then we scanned the BITNET DIRECTORY file on PSU's mainframe. Once we obtained a list of BITNET users in Britain we sent our e-mail cover letter to about ten of them asking for help in obtaining a computer account with access to a CMC network in southern England, somewhere near to where the speech communication researcher would be staying.
Within a couple of days we received a note from the system manager at the Rutherford Appleton Lab (RAL) in north England. He had seen our original note and passed it on to the system manager of a Prime computer at Sussex University in southeast England, who would grant us a free account on the Prime mainframe (which accessed the JANET CMC network). Our next step, he said, should be to contact Sussex University's Prime system manager directly by e-mail.
We did so immediately, sending an e-mail note to Sussex thanking the manager for granting us an account on his mainframe. To further legitimize our request we sent a cover letter via regular, postal mail confirming our research intentions. In the interim, the manager in Sussex responded (by e-mail note) that he could teach the speech communication researcher how to use the Prime CMC system once the researcher arrived in England.
The correspondence that took place between the RAL manager, the manager at Sussex University and our team all occurred within the space of about three days. The only difficulty we experienced was figuring out the correct mail headers with which to address our notes.
Some Problems Were Encountered
The first problem we encountered was that Sussex University's Prime connection to JANET did not offer "file transfer support." Files sent from PSU's IBM mainframe on BITNET could be received by the Prime via JANET, but were automatically converted to incoming notes by the Prime system. This meant also that the Prime did not support outgoing files. Consequently, note transfer was the only option for sending e-mail from Sussex University's Prime to PSU's IBM.
A second difficulty arose because the Prime could not communicate "interactively" with the IBM. Interactive communication is when a message can be sent to a person on a network, and if they are logged on, the message will appear on their screen within a couple of seconds. This type of communication is available between most BITNET-connected CMC systems. However, only asynchronous notes from the Prime could be sent through the gateway. Once the notes were sent from Sussex they were queued" to pass through the RAL gateway to BITNET. Moreover, the longer the note, the longer it took to pass through the gateway. We were able to exchange about one note each per half-hour at the quickest, though sometimes it was an overnight activity because of the length of the queue.
A third obstacle was that Sussex' Prime did not support mail-tracing. Thus, the ability to track the path of a note sent from Sussex through different nodes to its final destination was not available. Similarly, at the PSU end it was not possible to monitor the path of a note past the RAL gateway. In short, JANET proved to be the leg of the e-mail journey between our two sites that defied a path trace. As a result, neither of us was able to monitor the progress of e-mail in route as a means of estimating the other's response time.
A fourth problem involved responding to notes at the Sussex site, in that no "reply" feature was available. On the PSU IBM mainframe, typing the command REPLY while accessing a note sent from another user automatically generates a header that addresses the reply note back to the person who sent the original note. This function saves quite a bit of typing, and is especially useful in sending mail across networks where the headers needed to designate the targeted receiver's address are quite long. Unfortunately, the Prime required headers to be typed in full each time a note was generated.
A final disadvantage resulted from the short time period (three months) of our case study. This was not enough time to allow the speech communication researcher to learn many of the features available on Emacs, the Prime's word-processing program, which in turn made it difficult to edit files.
Distinct Advantages Emerged
Despite these five problems, however, there were four distinct advantages in communicating electronically across the Atlantic through the JANET-BITNET gateway. First was simply the time we saved. If we had used regular postal mail to send data and text from Pennsylvania to England it would have taken an average of about seven days to arrive. But, in one example of how much time CMC saved us, a 30-page document was sent from PSU's mainframe to Sussex' Prime in about five hours.
A second advantage of our JANET-BITNET connection was ease of data transportability. That is, the speech communication researcher did not need to physically carry research documents to England. All notes and files were sent electronically to Sussex' Prime before the researcher left the U.S.
A third advantage of our JANET-BITNET connection was cost efficiency. Basically, we were able to "talk" to each other through e-mail notes at no cost (other than physically travelling to and from the remote site), whereas transmitting the same information by telephone conversation or regular mail would have been expensive. We viewed this advantage as ultimately more important than the others in using the JANET-BITNET gateway.
A final advantage to our connection was that the Prime mainframe was relatively "user friendly." For example, once one was logged on to the system, a simple command of MAIL allowed access to incoming notes. Commands used to execute functions on the Prime in most cases were represented by pneumonic words denoting the meaning of desired function. Also, since the researcher had already known how to use e-mail on BITNET at PSU, learning how to use e-mail on Sussex' Prime was relatively easy since many of the same principles applied.
Caveats and Precautions
In the end, we evaluated the experience very positively. We firmly believe that despite the difficulties encountered in facilitating the communication link, the overall gains made in continuing research away from home made the efforts worth it. Thus we would not hesitate to recommend that scholars planning to travel from their home university consider pursuing similar CMC links if they need to access their home mainframe accounts while they are away.
However, we also would like to propose this hypothesis about potential difficulties in establishing and coordinating inter-network CMC links: That communication between two different networks such as JANET and BITNET will be reduced to the least common denominator. In essence, communication is done by functions that are common to both systems. In our case, we could not send files or communicate interactively, and were instead forced to use asynchronous notes, a CMC function common to both systems. Other users will likely face similar circumstances.
Finally, we suggest that three precautions be exercised before undertaking an
international or intranational mainframe computer connection between two or more CMC networks. First, plan ahead as far as possible. In our situation we were able to secure a consent in a few days, but it can easily take a few weeks to complete all of the paperwork necessary to justify usage of an account by a non-resident researcher.
Secondly, we recommend that all remote contacts be clearly established before leaving the home university. This includes contacting support personnel to assist the travelling researcher in learning the remote system. Once the researcher arrives, there is usually an immediate need for help in getting set up on the remote system.
Lastly, we strongly suggest that the travelling researcher develop contingency plans should the established computer communication connection fail--temporarily or permanently. We would like to reassert, however, that if preparatory measures are undertaken such as those recommended in this article, the travelling scholar should be able to use CMC to continue his or her research with very little trouble while he or she is away from home.
Roben McKenzie is an assistant professor of communication studies at California State University-San Bernardino, in San Bernardino, Calif.
Gerald M. Santoro is an assistant professor and the manager of Microcomputer & Workstation Support Services at the Center for Academic Computing, The Pennsylvania State University, University Park, Pa.
1. JANET stands for joint academic network," and BITNET stands for because it's time network."
|Printer friendly Cite/link Email Feedback|
|Author:||McKenzie, Robert; Santoro, Gerald M.|
|Publication:||T H E Journal (Technological Horizons In Education)|
|Date:||May 1, 1991|
|Previous Article:||TNT: an ambitious network from the 'Big Apple' state.|
|Next Article:||Electronic distance learning: positives outweigh negatives.|