[WEB4LIB] Crawford's comments on Earth's Largest Library

[Jerry Kuntz]

Although I missed the main flurry of comments on the ELL thread last week,
I've got to chime in with my laundry list of technical issues with Steve
Crawford's conception:

--Mention of OCLC as a model for ELL set off my mental alarms, since I've
long held the opinion that static catalogs (i.e. disconnected from item
status information) makes for a poor basis for ILL activity, especially in
the public library world, where libraries typically have over 25% of items
with a status other than "On Shelf". Moreover, in past years OCLC invoked
rather draconian fees in order to get libraries to maintain accurate OCLC
holdings--essentially asking libraries to maintain their catalogs on two
different systems.
--Steve Crawford doesn't mention the slow, prodding, but still substantial
progress that has been made in developing Z39.50 catalog connections and the
ISO 10160/10161 ILL standards. Products based on these standards can
delivery the functionality that Roy Tennant was mentioning last week, where
searches can gradually be scaled from local, to regional, to state,
national, etc. If Crawford had omitted the OCLC model and instead presented
a rallying cry for full speed ahead on Z39.50/ISO ILL implentations, I'd be
more willing to take ELL seriously. I'm sure Mary Jackson, among others,
will make this point at the Internet Librarian '99 presentation on ELL (and
if not, I'll pipe up from the peanut gallery!)
--From personal experience, I'll make a couple of general statements about
the economics of ILL: ILL's filled locally using locally devised delivery
methods are far more cost-effective than USPS, UPS, FedEx, etc. Therefore,
it makes sense to make the effort to build those local delivery systems and
to then conduct as much ILL as possible within those systems. This fits the
scalability of Z39.50/ISO ILL systems.
--As Karen Schneider mentioned last week, the whole ELL concept is talking
about delivery of physical artifacts, which I'm not sure is going to be an
issue in the long run. At least as interesting as what Amazon has been doing
is the idea of print-on-demand (in fact, either B&N or Borders--I forget
which--has been using print-and-bind-on-demand service units in some of
their locations.) I think this has a lot more exciting potential for
libraries than grandious physical artifact delivery systems.
--There are valid criticisms of library vendor catalogs vs. the value-added
features in Amazon's catalogs. As Walt and others have pointed out, though,
our library catalogs do offer better detail and more access points (if not
better usability--let's call that a toss-up). However, there are solutions
available that can add value to our catalogs. Products like: Ebsco's
Novelist, which searches a reader's advisory database and then can connect
to the catalog; enhanced MARC records with reader's advisory information,
such as offered by Syndetic Solutions or Fiction Digest on MARC. Amazon also
offers their nice Book Matcher reader's advisory database, which makes
suggestions from user input based on purchasing patters. Integrated library
systems could generate similar reading patterns...if we asked vendors for
this feature.
--Although I'd also characterize library automation vendors as slow and
plodding, I'm certainly not ready to throw them out (as mentioned in a post
by John Blegen last week.) They were slow to react to the Internet; and
still are trying to catch up; but at the same time are hindered by the fact
that their installed customers haven't had the infrastructure in place to
jump from text-based mainframe systems to networked systems.
Moreover, although watching the Open Source Systems for Libraries
initiatives with interest, I'd say they're a long way from offering the
functionality we have (and have come to depend on) from our established IOLS
vendor. At the same time, the established vendors are actually getting good
at paying more than lip service to adopting standards! [an aside:
periodically, I get to deal with "computer-saavy" library board members who
are CERTAIN that they can develop a low-cost circ system on their own. One
look at a MARC record usually shuts them up.]
Sorry, John, I'd be much more willing to toss OCLC on the garbage heap than
my IOLS vendor!
...enough on ELL...I guess Steve Crawford and Walt Crawford aren't related.

Jerry Kuntz
Ramapo Catskill Library System
jkuntz at rcls.org