Bearing with Old and New Standards
by Andrew Pace
I marvel at the screwdriver.
The tool, that is, not the drink. I marvel because whenever I use one, I think
about the effort it must have taken back in the nascent world of standards
to build a device that could be used the world over. No longer would screw
and driver be proprietarily paired.
Then I wonder, did Henry Phillips (of screwdriver fame) and Peter Robertson
(a little trivia: destined for much less fame than Henry, he invented the
socket-head screw) ever go out for drinks and argue the finer points of their
inventions? Did they joke about hammer-and-nail users?
"What do they care about standards?" Phillips would muse.
"They get their work done with brute force, and by flooding the market with
their cheap and crude solution," Robertson would reply.
Get a life, I know. Nevertheless, I think the cultural history of standards
development is somewhat fascinating. Library science is a profession obsessed
with standards, but its cultural interaction with them is often quite flaky.
Librarians preach and preach the benefits of standards, but sometimes we acquiesce;
sometimes we create them in vacuums and hold onto them beyond reason; sometimes
we look down our noses at them because no one invited us to the table to develop
The Innovator's Dilemma
Sometimes we willingly sacrifice standards for features. The browser wars
were the perfect example. Everyone knows that both big browsers (Internet Explorer
and Netscape, now the mammoth and the minuscule, respectively) developed features
beyond the HTML standard. Should they have waited for the W3C to formalize
HTML 2.0 before making tables work? Should they have been so forgiving of sloppy
coding? Borrowing Clayton Chistensen's book title,1 one might call
this an "innovator's dilemma."
Librarians making their way in the Internet age live constantly on the cutting
edge. Those who try to cut a swath through technology often sacrifice standards
If we're lucky, a standard will emerge to facilitate whatever it is we're
trying to accomplish. OpenURL is an excellent example. This brilliantly simple
standard for getting from citations to the full text of articles is the magic
bullet librarians have long sought.
Fortunately, stopgap measures were not obviated by the emergence of an acceptable
standard. E-journal management systems and purchased aggregated title lists
filled the service gap for a while, but OpenURL was a standard way to represent
what many were trying to accomplish. Version 1.0, just released, implements
many of the innovations that libraries and vendors created on their own after
using the 0.1 for almost 3 years. OpenURL succeeded because it did not represent
a challenge to existing standards or to years of devotion to an existing system.
Are We Losing Our Grip?
I took a lot of cataloging classes in library school. I wanted to be a rare
book cataloger. The book arts actually led me away from cataloging and toward
HTML 1.0. I should have been thinking about cataloging as a set of standards,
but I never did; they were simply rules. But HTML ... that was a standard to
me. Moreover, the dilemma was still there. As a training librarian, I viewed
the MARC record as completely monolithica beast that was beyond further
training. As I learned about Web protocols and relational databases, that 800-pound
gorilla was always in the corner of my eye, but I could not determine what
to make of it.
Throughout a decade in libraries, I have probably tried to explain the MARC
record to about 2 dozen programming, MIS, and computing professionals. Smart
people, all of them, but unanimously baffled by the stronghold the MARC record
had on the information profession. What's more, I had worked with integrated
library systems. I saw the dedicated efforts that went into cataloging modules
wasted by ineffective search engines and antiquated indexers, or ignored by
minimalist sort, filter, and display options in first-generation online catalogs.
Now before I get piles of hate mail from technical services folks, let me
re-emphasize my love of cataloging. But MARC is not cataloging, it's just a
standardized format. I don't hate MARC. But lack of proper use of the metadata
by indexes and applications, and blind loyalty to a 40-year-old standard can
prove infuriating at times, especially when the standard does not do enough
to meet the needs of patrons.
Librarians should be proud of the MARC record. We deserve to be snobs about
standards. We are the leaders of standardized creation, storage, and transfer
of descriptive metadata. Yet it seems unrealistic to expect technological advances
from IT staff and from automated systems vendors while demanding adherence
to and promoting longevity of aging standards. There is likely some middle
ground between preserving our past and bearing (and bearing with) the standards
of the future.
"WE DESERVE TO BE SNOBS ABOUT STANDARDS."
The Z39.50 standard presents even bigger challenges. While it remains the
best way to broadcast searches for bibliographic data, its use as a standard
for metasearching disparate information resources presents insurmountable problems.
Even information vendors who have run Z39.50 servers for years never planned
for the impact of metasearch engines. For others, CGI access to their data
was enough of a challenge; the notion of implementing a Z39.50 server for access
to citation data or proprietary data sets is too high a barrier of entry. There
are new challenges (as well as revenue opportunities) for software developers
and content providers. As with OpenURL, libraries, content providers, and software
vendors all share a common interest; the opportunity for cooperative development
of a new standard is excellent.
If we think of our 40-year trek through the standards circuit as ending now,
then our second time around this circle represents a real chance to do things
better. Librarians are desperate for better standards to support their core
functionality and day-to-day activitiesNCIP for patron data, COUNTER
for statistics, EDItEUR for acquisitions, OpenURL for citation linking, and
something for metasearch, just to name a few. Vendors are equally anxious to
provide services that adhere to desired standards, especially in a technological
climate that emphasizes interoperability and distributed systems. Not to worry,
a new era of cooperation and co-development is at hand.
No one can take more credit for this new era than NISO, the National Information
Standards Organization (http://www.niso.org). Beyond promoting standards development
by singularly interested parties, NISO has fostered cooperative efforts among
wider ranges of stakeholders, including content providers, vendors, middleware
developers, and librarians. NISO has a stake in nearly every standard mentioned
in this issue.
One of NISO's most interesting tactics has been to couple grass-roots efforts
at standards development with top-down analysis of business functions. The
metasearch initiative is an excellent example. Several interested parties met
in Denver in early 2003 to brainstorm and determine the issues surrounding
federated search, resource discovery, and authentication. That meeting was
followed up last fall by a NISO workshop with an even wider audience of librarians,
vendors, and publishers. Work continuesvia its "Metasearch Initiative," NISO
is capturing the momentum created by those meetings, the press that followed,
and community reaction.
One of my favorite sayings is: "When all you have is a hammer, everything
starts looking like a nail." Our profession has plenty of hammers and brute
force opportunities to build the things that we need. MARC is a hammer; stand-alone
development done in a vacuum is a brute force strategy. Our profession needs
1. The Innovator's Dilemma: When New Technologies
Cause Great Firms to Fail, Clayton M. Christensen,
Harvard University Press, 1997.
Andrew K. Pace is head
of the systems department at North Carolina State University Libraries. His e-mail
address is firstname.lastname@example.org. You can also reach him through his Web site