I know little about Duke's initiative, but I do remember the OSI library
projects in the 1980s, the expectations of immediate interoperability in
the early 1990s at the point when the first circulation system talked to
another vendor's catalogue with Z39.50, and even the much more recent
VIEWS project [1]. VIEWS, I believe, was also an attempt to define
interfaces for mixing and matching flows within the ILS. NCIP also seems
like an especially poignant and somewhat immediate exercise in the lessons
of defining machine interfaces for library functions.
I will admit that, like Dan Scott, I am biased towards Evergreen, but I
would urge the group behind "blowing the lid off the ILS" to take a good
look at OpenSRF [2], both the concept and the implementation. Outside of
the perception that it might be intrinsic to any particular system, the
reason to consider OpenSRF is simple.
OpenSRF could be to the ILS what HTTP was to the web.
A problem with defining specifications in the abstract is that metrics of
impedance and throughput are rarely worked in. There were a group of
engineers at Cray that defined a PC emulation for super computers, and
they were actually able to implement it. The only problem was that it took
two weeks of processing for the command prompt to appear. I am not, in any
way, suggesting that the Duke project is heading down this path but the
brilliance of HTTP, at least in part, was that its design was never far
removed from its deployment.
Also, I think that that there is a very good chance that the OSS world is
getting darn close to the commercial side in terms of the number of
options. NewGenLib [3], from India, doesn't seem to have received a lot of
attention in the West yet but it looks quite complete. Add this to
Evergreen and Koha, and factor in the options for smaller collections,
like OpenBiblio and such, and you have a slate that is, at the very least,
starting to match the vendor offerings.
I bring these systems up because I think the "worrying about needing to
get vendors to support" meme may be becoming more and more of a red
herring. A couple of years of following municipal politics has taught me
that it really takes a critical mass in order for a group to have any kind
of credible scapegoat. For better or for worse, it's not like there are
that many vendors to blame any more, and there may literally be less
excuses than ever before to create the world we want, at least in terms of
vendor lock-in.
I mucked up one of my arms for a few weeks and managed to get buried in my
newspaper life, so I may have missed some of the dialogue on this project.
I do see potential in aligning some demonstrably successful plumbing
concepts from open projects with definitions of what systems should
support, and that may be what is afoot here. If so, that could be a very
good thing.
art
---
Art Rhyno
Systems Librarian, University of Windsor Libraries
Co-owner, The Essex Free Press
1.
http://web.archive.org/web/20051230080851/http://www.views-consortia.org/
2. http://open-ils.org/blog/?p=36
3. http://sourceforge.net/projects/newgenlib
Received on Sat Feb 02 2008 - 10:20:30 EST