Date: Tue, 07 Jun 1994 13:24:25 +0000 (U)
From: Eric Morgan <eric_morgan@LIBRARY.LIB.NCSU.EDU>
Subject: Mr. Serials

At the recent NASIG conference I presented my implementation of an electronic
serials acquisitions process. I have been encouraged to post the text of my
presentation for this list's "consumption". Thus, you can now either press the
delete key to send this message into the ether or you can read about Mr.
Serials, a prototypical electronic serials acquisitions process.

Enjoy.

Eric Lease Morgan, Systems Librarian
NCSU Libraries
Box 7111, Room 2316-b
Raleigh, NC 27695-7111
(919) 515-6182
eric_morgan@ncsu.edu
http://dewey.lib.ncsu.edu/staff/morgan/morgan.html

Mr. Serials, an electronic serials acquisitions librarian
by Eric Lease Morgan
http://dewey.lib.ncsu.edu/staff/morgan/mr-serials-at-NASIG.html

Introduction/Abstract

Mr. Serials is the code name for an electronic serials acquisitions process
being applied at the North Carolina State University Libraries. By employing
various technologies (LISTSERV, FTP, gopher, WAIS, WWW, and a bit of
programming), Mr. Serials has been able to systematically subscribe, collect,
organize, archive, and disseminate electronic serials quickly and easily. This
text outlines the Mr. Serials process.

Selection

As I see it, Mr. Serials has five issues do address concerning electronic
serials:

  1. selection,
  2. storage,
  3. organization,
  4. access, and
  5. acquisitions.

His first task is to select and subscribe to the ejournals. He took it upon
himself to collect only a small segment of the population, specifically the
titles dealing with library and information science. These titles presently
include:

  * ACQNET
  * ALANEWS
  * ALAWON
  * ALCTS Network News
  * Citations for Serial Literature
  * Current Cites
  * HOTT -- Hot Off The Tree
  * Information Retrieval List Digest
  * INFOSYS
  * Issues in Science and Technology Librarianship
  * LC Cataloging Newsline
  * LIBRES
  * LITA Newsletter
  * MC Journal
  * Network News
  * Newsletter for Serial Pricing Issues
  * Public Access Computer Systems Review

This part of the process was easy. Mr. Serials just used Charles Bailey's list
of ejournals (ftp://dewey.lib.ncsu.edu/pub/stacks/guides/bailey-library.txt) as
a guide and followed the instructions listed there.

FTP For Storage and Rudimentary Access

He decided to solve his storage problems using anonymous FTP. This way anybody
can get the articles if they have access to the Internet. FTP also provides a
mechanism for archiving the materials. It is important to note he created a
single directory for each title. Within that directory are all the issues and
articles from that title, each with a unique but identifying name. For example,
every file from Citations for Serial Literature  (CSL) is designated to begin
with "csl". Since CSL is issued in volume and number, each designation is
followed with something like "-v1n22". Therefore a complete designation is
"csl-v1n22". Issues from Public Access Computer Systems News  looks like
"pacsn-v1n03". IR List Digest , which comes out sequentially, is saved as
"irld-001". Take note of the leading zeros in the preceding examples. They make
it possible to accurately sort the files by volume and number, and consequently
date. Also, keep in mind, the serial codes he has created are arbitrary. As
long as he is consistent and there is a method to his madness, then it doesn't
matter what he calls them.

Gopher or WWW for Enhanced Access and Organization

FTP is strong on storage, but weak on access. This is where a gopher or World
Wide Web (WWW) server come in. The gopher and WWW protocols excel at
maintaining pointers to Internet resources. Since anonymous FTP sites are
Internet resources, and since Mr. Serials has both a Gopher and WWW server in
place, he created gopher and WWW pointers to his ejournals. This solved a
number of problems. First, the gopher link files provide the capability to list
files in a more readable fashion. Second, the user no longer has to use FTP
commands to access the collection.

Below is a fictitious, but correctly formatted gopher link file:

  Name=ALAWON v1n35 (May 5, 1992)
  Type=0
  Port=70
  Path=ftp:dewey.lib.ncsu.edu@pub/stacks/alawon/alawon-v1n35
  Host=dewey.lib.ncsu.edu

When the patron uses the gopher they see "ALAWON v1n35 (May 5, 1992)." By
selecting the item the gopher server transparently opens up an anonymous FTP
connection to dewey.lib.ncsu.edu, changes to the /pub/stacks/alawon directory,
retrieves the file "alawon-v1n35" as a text (ASCII) file, closes the
connection, and displays the file to the user. From there the patron can read,
save, or print the file.

Access via the WWW works in a similar manner. WWW-compatible text files
(hypertext markup language documents) are created listing each of the titles.
These documents are organized according to volume, number and date. These files
are then put in the WWW server. When the user uses the WWW server, the issues
are displayed. By selecting one of the issues, your WWW browser FTP's the
selected file from dewey.lib.ncsu.edu and displays it on the user's screen.

WAIS For Keyword Access

Mr. Serials is now using FTP primarily for storage and rudimentary access. With
gopher and WWW he ehnanced access and organized the information. Unfortunately
FTP, gopher and WWW are only useful if the patron knows exactly which issue or
article they want or if they want to browse the collection. WAIS now enters the
picture allowing the patron to search the collection via keywords. Since Mr.
Serials owns all the articles associated with any particular ejournal, and
since he has a WAIS server in place, then it is a simple matter of creating a
WAIS index for each of his titles and putting these indexes in the appropriate
directory of the gopher or WWW server. To accomplish this, everyday at 2:30 AM,
the entire collection of electronic serials owned by the North Carolina State
University Libraries gets reindexed. Thus the index is never more than one day
old.

Automated Acquisitions With AC

Finally, after all this planning, Mr. Serials has a place to put his ejournals
once he collected them. He began by manually retrieving the files with the
LISTSERV "get" command. Once retrieved, he manually edited the files, saved the
edited versions in their appropriate FTP directory, wrote gopher link files,
and updated the WAIS indexes.

Needless to say, this got old quickly! From the comp.infosystems.gopher USENET
newsgroup he had heard a lot about a programing language called "perl". He
retrieved the perl archive and wrote his very first perl script. This script,
which he now calls "ac" automates much of the process listed above. ac:

  1. reads an email message,
  2. removes the header,
  3. prepends the name of the file to the message,
  4. saves the file in an FTP directory,
  5. creates and saves a Gopher link file or WWW HTML document,
  6. deletes the original email message, and
  7. "cooks until done" (repeat)

Now, all I have to do is wait for mail to arrive, save the message as a file,
and use ac to make it available to the world.

For more information about ac use your World Wide Web browser to point to
http://dewey.lib.ncsu.edu/staff/morgan/ac.html.

To see the fruits of Mr. Serial's labors, point your World Wide Web browser to
http://dewey.lib.ncsu.edu/disciplines/library.html#Journals.

Summary and Future Considerations

Looking back, the most labor intensive aspect of this ejournal collection
process has been collecting all the back issues. This problem should be behind
us except if we decide to collect other electronic serials.

It was tempting to begin the acquisitions process without a structure for
storage and access, but now that the structure is in place things are going
smoothly.

A problem with Mr. Serials and ac is the inconsistant manner in which
electronic serials are formatted. As long as any single title is consistantly
formatted, then ac works just fine, but as soon as the formatting changes, even
by one word or blank line, then ac must be reconfigured for that title. This
reconfiguration process is not difficult, just tedious. The problem could be
eliminated if each of the titles supported a standardized header such as the
one described in chapter 5, "The TEI Header"
(ftp://ftp.ifi.uio.no/pub/SGML/TEI/P3HD.DOC) of TEI P3: Guidelines for
Electronic Text Encoding and Interchange
(http://www.lysator.liu.se/runeberg/teip3files.html).

Throught this process an unexpected benefit became apparent. Since Mr. Serials
has a (almost) complete set of the library and information science titles he
can (and has) created an WAIS index version of Library Literature. He calls it
Electronic Library Literature Index (ELLI). With ELLI you can search the entire
body of electronic library literature and retrieve articles from a number of
titles. For example, you can search for "NREN" can come up with articles from
Public Access Computer Systems News , ALAWON , and IR List Digest . There is no
reason why this idea can't be used for other ejournals and subjects. This index
is available at gopher://dewey.lib.ncsu.edu/11/library/disciplines/library/ell
or use your WAIS client to connect to dewey.lib.ncsu.edu on port 210. Request
database ell.