Tuesday 18 October 2005 @ 10:21 pm
Two-thirds of the way through Access! One more day to survive; unfortunately, it’s the day when I’m presenting. Here are some thoughts from today’s program:
- I wish I’d thought of “radioactive metadata”. As used in ZInterop2, described by Bill Moen this morning (who attributed the term to Sebastian Hammer), the idea is to craft diagnostic metadata records (in this case MARC but the idea would apply to any format) to test search systems: put unambiguous easy-to-search tokens in specific fields, mount them in your repository, index them, and then use standard search scripts to generate a report on which tokens were found and which ones weren’t. When I think of the time I’ve wasted trying to find OPAC records with specific patterns of terms to test whether a particular field is being indexed and how, I weep. The preliminary version is available.
- I mentioned yesterday the desirability of developing our websites as installable apps. This came home to me even more clearly today in the Martini presentation. It started with my custom-built process to mount Olive OCR items in our digitization project. I was then lucky enough to get two excellent developers in a row to slap it into shape, move the full-text indexing to Lucene, make it multilingual, make it handle newspapers as well as monographs, and generally professionalize it. This work is still in progress; but when it’s done, the next priority must be to make it METS-driven. Then Olive content will become just one XML content option among many.
- Roy Tennant told a sad tale with a happy ending about a quest to deploy a new service. He told his vendor, “We can’t make the interface do what we want, so we’re dropping it and building our own using your API. Your API doesn’t work either, by the way.” He was happy in his choice of vendor, and they have put considerable effort into developing the API to meet CDL’s needs. The future looks rosy. A lesson to us hasty hackers who delight in the improvised workaround: sometimes it’s better just to let the pros do it.
- Ross is wrong, there’s nothing specifically Canadian about disco balls. Just because there’s one in the dining room of the Faculty Club… When disco comes back, Edmonton will be ready.
Warning: Creating default object from empty value in /home/wandb/wallandbinkley.com/production/quaedam/wp-includes/comment-template.php on line 1015