5.0700 CD-ROMS -- Networked in Libraries (2/125)

Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Wed, 19 Feb 1992 21:29:17 EST

Humanist Discussion Group, Vol. 5, No. 0700. Wednesday, 19 Feb 1992.

(1) Date: Mon, 27 JAN 92 16:57:14 GMT (19 lines)
Subject: CD-ROMs in libraries

(2) Date: Wed, 19 Feb 1992 08:15 CST (88 lines)
From: "Thomas C. Wilson" <LIB4@Jetson.UH.EDU>
Subject: CD-ROM Networking at Univ. of Houston Libraries

(1) --------------------------------------------------------------------
Date: Mon, 27 JAN 92 16:57:14 GMT
Subject: CD-ROMs in libraries

I've been reading the messages about CD-ROMs
and networks, towers etc. with more than a little interest,
and I wondered whether any of you knew of an up-to-date
source of information on what both university and public
libraries have available for running CDs? It would be most
useful for me to know what libraries have at the moment,
or are planning to get, and what their users expect in the
way of network facilities for CDs.

Please reply directly to me. Many thanks.

Ruth Glynn
CD-ROM Project Manager, Electronic Publishing
Oxford University Press
(2) --------------------------------------------------------------98----
Date: Wed, 19 Feb 1992 08:15 CST
From: "Thomas C. Wilson" <LIB4@Jetson.UH.EDU>
Subject: CD-ROM Networking at Univ. of Houston Libraries

I am responding to an item forwarded to me by a friend. I am not a member
of this list. My apologies for not responding sooner.

Judy Koren writes:

>Humanist Discussion Group, Vol. 5, No. 0621. Thursday, 23 Jan 1992.
>Date: Thu, 23 Jan 1992 11:04:25 +0200 (EET)
>Subject: RE: 5.0613 CDROMS -- Pricing, Library Use
>Lorne Hammond's comments on library supply of CD databases were very
>interesting. Lorne...saw a "jukebox" arrangement where a PC supplies
>10 CDs, though only one at a time, and is linked to the library LAN.
>Libraries are rushing to put CDs on the library/campus network these
>days the way they were rushing to computerize the catalog 10-15 years
>ago. But I'm surprised Houston opted for a jukebox arrangement.

Not true! As usual, in technical discussions, terminology is *extremely*
important, and frequently words are used that convey something different
than what may have been intended. I cannot speak to the exact description
of our network that Lorne was given, since it was not I who gave that
description. While it is true that we built our CD-ROM network to be
redundant in terms of equipment to avoid disasterous failure implications,
we did not use "jukebox" technology, and each drive/product is available
to every user all the time. We have two Meridian Data CD-ROM towers
(redundancy), one NetWare fileserver, and ten workstations in the library.
We will be expanding this arrangement significantly in the next six months.

>Most "professional" opinion holds that it's [jukebox] too slow for efficient
>service on a LAN, i.e. when several people are likely to want a disk from
>the same jukebox at the same time, and they'll have to wait because
>the jukebox PC must physically switch CDs for each one.

Agreed! There are, however, some applications in which a jukebox arrangement
on a network is appropriate (e.g., large numbers of discs that are rarely or
less often accessed, but still important).

>I could understand
>it better if the "tandem" arrangement Lorne mentioned meant that both
>PCs were supplying data at the same time, but s/he says one only steps
>in when the other crashes.

Current CD-ROM networking products do not support this type of automatic
fail-safe operation. Each of them requires a human to change physical or
logical mapping of the software and hardware. We decided to purchase
redundant equipment so that we could make these changes if necessary at
any point of failure. Fortunately, we have never had major failure.

>You can put that
>on the campus-wide LAN (as a matter of fact you can put as many of
>them as you like on the campus-wide LAN) and let 100 faculty members
>loose on them with relative impunity.

In theory, yes. But the technical details will in some cases limit the
number of servers one can implement on a given segment of a network and the
level of access one can provide.

>I'm sending this to Humanist because Lorne said "talk to your friendly
>librarians about it". Being one myself, I'm all for it; but don't,
>please, suggest a jukebox... suggest a tower.

CD-ROM technology is somewhat unique in that it hit libraries long before
any interest was expressed in the general computing industry. That has
changed to some degree now. Nonetheless, your local librarian may be an
excellent source of information on this topic. If you or your institution
are looking to network CD-ROM, please focus on the access issues (i.e., what
you want to gain by doing this, what your environment is like currently,
who you want to reach, how much you can afford, etc.), rather than the exact
technology implemented. The technology changes, sometimes rapidly, but the
access issues are more constant. That is not to say that the two sets of
issues are unrelated. They are! But what something is and what people call
it are frequently not related. The best evaluation of any project of this
sort is based on whether or not it, when fully implemented, improves your
access to information. Whether it is CD-ROM, jukebox, tower, network,
UNIX, VMS, VM, Mac OS, OS/2, or DOS is almost irrelevant.

If you have other questions about our implementation of CD-ROM, please e-mail
me directly. Thanks for your interest!

Tom Wilson (a librarian)
Head of Systems
University of Houston Libraries