Hmm ...
Begin forwarded message:
From: Daniel Chudnov <[EMAIL PROTECTED]> Date: February 25, 2005 5:56:01 PM EST To: [EMAIL PROTECTED] Subject: [unalog-general] dumb implementations of smart metadata
It might seem like unalog development has ground to a halt... in a way, it has, with o significant checkins in weeks.
But, we've been making various kinds of progress. I've tested a few different approaches to adding search, and today's winner (and probably a lock for the long-term choice) is PyLucene. It's lucene, it's fast, it's working, and there are some nice little UI tricks I think we can layer in to make it even more user friendly. Expect to see that checked in in at least rudimentary form sometime in the next 7-10 days.
Longer-term, the Yale Gang of Three (heretofore YGoT) has been meeting regularly to discuss The Metadata Problem. That being, "how do we expand our notion of what an entry is so we can save citations and route openurls and the like?" We're rather enchanted with LoC's MODS, and think we can upgrade the current model of unalog entries directly to MODS records and add features as we go. This will also allow us to wrap around things like bibutils to allow import/export in a few well-known formats, among other things.
The other thing we'd like to try to do with MODS is to find a way to track the "site-wide" "bibliographic" version of an entry (whereas users have a "holdings" record). The YGoT is thinking that we could store a MODS record wrapped inside METS, with changes over time noted as METS amdsecs. And we were also just thinking how much fun it would be to track tags, keywords, names, and subjects using MADS, too.
That's all a little ambitious. And the YGoT might be ambitious, but we're also project-laden and resource-cramped. Cranking out full and reliable implementations of each of these specs is no small task, especially as each is changing fairly rapidly, and none of us has a lot of experience with any of them. Also, there's the do-we-store-xml-natively-or-not conundrum, and we don't wish to fan the flames of the Hacker Culture Wars just now.
So the YGoT is probably going to spend the next few weeks producing some dumb implementations of these very smart metadata specs. It should be fairly simple to convert the current unalog "UserEntry" into a more MODS-like internal model. It should be fairly simple to save a site-wide "bibrecord" for each entry, and define a couple of workflows for whether users who add similar items a second and third time add to the bibrecord that's there, or replace, or just modify optionally. It should also be easy to keep a METSish model of changes over time, and a MADSish model of how tags relate to each other, and simple UIs for working with both of those. It would be particularly easy too if we just avoided the whole XML question for a while and just worked "in the spirit of M[A|E|D][D|T]S" without perfecting all those messy angle brackets just now.
Thus we might not go from 0-to-SchemaValid anytime soon, but I think we can crank out some stuff that will be More Useful and Not Irreversable and Something To Improve Upon if it goes well.
It's awkward to have these lengthy, insight-inducing YGoT meetings without having some or all of you in the room, and I know these occasional missives come off as long-winded and hard to follow. Suffice it to say, then, that we'll be cranking it up in the next few weeks, and anywhere and everywhere any of you want to pitch in, we'll try to find a way to make it easy. Even if I have to write succinctly.
-dc
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
unalog-general mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/unalog-general
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
