In this article, I take a look at some of the key issues in the demise of BIBSYS.
A consequence of the “BIBSYS behovsanalyse” [in Norwegian] is that the university libraries in Norway are now looking at Ex Libris’ Primo as a replacement for the aging BIBSYS Ask.
For those not familiar with BIBSYS, it is the Norwegian library mangement system for accredited universities and colleges; the data in use is shared by the various (100+) institutions. BIBSYS as an institution is government owned and placed as a part of the Norwegian University of Science and Technology (NTNU) — my institution — in Trondheim.
There have been a number of attempts to change — some have said obliterate — BIBSYS over the years, and it has been pointed out to me that this is due to the unpopularity of the connection between NTNU and BIBSYS. Whether or not this parochial wishful thinking or not can be debated by others — I say that it is just ireelevant.
BIBSYS’ downfall has been due to its openness — openness to its users and their opinions, and openness about what they are doing right and wrong. For example: I know what database platform BIBSYS uses; I know what programming languages are in use; I can go down to BIBSYS’ offices and have a meeting with the people there without feeling that anything special is happening; I can write a blog post and actually get a response from the head of programming at BIBSYS; I can affect changes in the system by lobbying for them.
All of this is of course really, really bad news. For BIBSYS. Hand in hand with this kind of flexibility comes the expectation that “my particular wishes and desires” will be implemented just because I asked for them — you’ve been listened to in the past, why aren’t they listening now? How very upsetting. Gosh! What rotters.
Another thing is that people get to see the mistakes made by the BIBSYS team — not everything, but you do get to see some of the errors they make. To err is human, but when you’re BIBSYS, something nasty hits the fan every time.
BIBSYS has itself to blame on other fronts too — it listens to the “think tanks” appointed to help them develop the system for the future. Listening to one such appointed think tank caused a furore last year, when a change was made in the interface according to the recommendations of the think tank, only to be universally decried by the national population of librarians at higher education institutions when it was so unfairly sprung on them. BIBSYS had really mucked up here and had to do a total U-turn: a public relations disaster. In all fairness, they had said that they would be making this particular change during the previous hearing round.
BIBSYS also does a lot of weird stuff — like providing multiple language support. I mean, why is that necessary? Doesn’t everyone speak English by now? Oh yeah, and specialized search options for minority subjects like music…fools.
The sad fact is — cutting through the irony (sorry) — that all of these things have indeed contributed to BIBSYS’ demise as an organization. Personally, I’m very, very impressed with BIBSYS as a system; it’s easy to use, the feedback you get from user support is professional and I think that it does more than any similar system. On the other hand, all of the openness has indeed left them “open”; attacks of the kind I heard at the lunch table the other day “their database is on the way out” are impossible with companies that guard their industrial secrets carefully. Foul-ups and so forth are quickly hushed up by PR people for commercial service providers.
To find examples of foul-ups, you need look no further than our own Metalib installation; disappearing clustering due to license expiry (!) and consistently absent multilingual support. Weak. Oh, and no-one uses it, which explains why we haven’t heard about it!
BIBSYS is used, it is widely used. Unfortunately BIBSYS has also fallen prey to the conception that BIBSYS Ask == BIBSYS; this is obviously nonsense — BIBSYS is a library management system, whereas BIBSYS Ask is an OPAC. The OPAC is an interface to the collection — Ask is an ugly one with bad search options (it doesn’t accept RPN or even search by regular expression and is thereby rubbish in my book [no pun intended]).
Now obviously, the BIBSYS Ask OPAC is a bit wet around the edges, and no-one really likes it very much (hey, it’s old, and has been spec’ed by committee). All this means is that the very public face of the system (as far as most library directors are ever likely to get) is ugly and naff. Oops.
I’m rambling now — so if you’re still with me, I will reward you: Primo isn’t a really great replacement for BIBSYS, it is a replacement for BIBSYS Ask. Sure, Primo searches for articles and provides ranking…actually, Metalib does the article searches ten compatible databases at a time (remember what I said about Metalib above?), and the ranking is just weird (try it at one of the open Primo installs).
I was at BIBSYS the other day, and they are scared. They were implementing Primo-like functionality in their Ask 2 beta. Only it was better — the ranking was gone and the resources being searched weren’t totally obfuscated. I thought it looked cool, but the guy I work with said it was stupid. He was right, I was wrong (I freely admit to this).
Why? After having done a search, where you know more or less what you expect to find. in Primo, go and do the same search in Google Scholar. The ask yourself (or “Ask” yourself if you feel like it’s time for another pun), why would I pay good money for this. The point is “looking for a replacement for BIBSYS Ask” is a futile exercise — it isn’t ever going to be relevant. Those days are over.
People find the literature they need in various ways; they paths they find typically suit their comprehension. No-one except a member of library staff expects to find something that they do not know the title or author of in BIBSYS Ask. Nor should they. A library catalogue is great for known entity searching (this is hardly new knowledge).
Databases of different kinds have different added value; the native interfaces provide helpful tips and constraints that help a person with knowledge of their subject to find literature on their topic. They use their subject-specific skills to analyze the results they get.
A student at 1st year level (OK, totally unrealistic in most Norwegian contexts — the first time most students venture beyond obligatory texts is towards the end of their third year) will not have the skills necessary to analyze the results they get from a “good database”, they need help. Google Scholar or Primo can provide this in the form of ranking. Nice.
Back up a moment: the majority of the literature that the library is paying for is probably so far beyond the level of comprehension of these students that there is no real point in presenting it to them. They need to get a few skills together so that they can recognize what is relevant to them (even if I put five relevant texts in front of you, you might not get it — I see Deep Thought here) — subject-specific information literacy — and then they can get away with using one of those “good databases” I was talking about.
So, Google Scholar and Primo; what are they good for? Known entity searches typically, because they provide links to electronic full texts and ordering via OpenURL (if you’ve implemented it that is). Winner.
Only problem is that Primo and Google Scholar don’t have a profile service that allows the registration of loans and so forth. Oops. Looks like you need BIBSYS again.
What to do? I don’t know, and I’m not sure that I care — as long as the library can provide efficient access to literature (“I’m only really interested in finding online articles” as one PhD student said) then we are doing our job. How people get hold of this is up to them.
All power to BIBSYS and Ex Libris, I’m sure they know what they’re doing. I remember however that I used BIBSYS a few times when I wrote my Masters thesis, otherwise I used CiteSeer and and commonsense. I was — in other words — much like any other student (just that most people will now use Google in one form or other instead of CiteSeer).
What have we learnt? Not much. What should we do? OAI-PMH and a simplified OPAC with a profile webservice and a nicely functional OpenURL resolver.
Did you like the article? Do you have any comments? Please feel free to contribute your opinions below. I can almost promise a verbose response in English or a Scandinavian language, if you must.