29 July 2004
A colleague passed me this message which echoes my sense that some protective clauses we fight to include in our contracts are nearly worthless in the real world.
About six months ago, [our] University Libraries was faced with a decision about continuing our access to what was formerly called Elsevier’s Academic Freedom Collection. We had subscribed to the package of Academic Ideal e-journal collection since 1998 through our consortium. Elsevier wasn’t willing to work through consortium arrangements and wasn’t willing to provide the group of journals as a package any longer. [Our] Libraries’ acquisitions budget had been cut and we were facing yet another reduction. We could not afford to subscribe individually to all the previously owned journals via ScienceDirect. We renewed 34 titles in print. We also could not afford to pay the annual access fee to maintain ScienceDirect linking to the backfiles. We chose the option of receiving the Ideal/Freedom journal backfile data that we had purchased by our several years of subscriptions. Elsevier sent us 8 DLT tapes about two months after our request for the data. After another frustrating two months of locating the outdated tape drives needed to open and access the tapes, we have discovered that there is duplicated data and that the data does not appear in any kind of rational order.
It will be interesting to see how they fare. I’m afraid that the data in vendor systems will get more and more complex and that a “dump” of this data will be less and less useful to anyone outside that vendor’s shop. To make these clauses meaningful will require that we develop some well known formats and then demand the offloaded data meet these specs. Of course, none of our contracts currently contain such requirements and in any case no such standards currently exist.
I have a similar concern with regard to source code escrow agreements. How much good does it do us to have source code without the suite of compilers, libraries, and tools that it takes to build a given application? Even if we could build it, would we have the skills to do so with confidence? In most cases, wouldn’t we migrate to an alternate vendor’s product before taking on maintenance of a defunct vendor’s product?