The criticism leveled at Nashville Public Library’s booksite lists is accurate. These lists are compiled independent of the library’s holdings, so some items in the lists cannot be located in the library’s catalog because they are unowned. This is a big problem that should be corrected with a better technical solution that removes these titles from the list to begin with.

However, the majority of the books in these lists are connected directly to the items available in the library catalog with the “Find In the Library” link. So, in most cases this service is helpful.

Still, it is not ideal, and does not approach the better bibliographic advisory services offered at Amazon.com.

Thanks!

Seriously, that’s exactly what’s needed. Convince Mellon to fund it, get Roy or someone to be the director, I’m on board in a heartbeat.

You see our users frequently “confuse” the web catalog and the website proper. That’s because almost everyone visits us online in order to search for materials. This should tell us that a a good catalog is much more important than a great web page. So many libraries roll out these shiny new websites, but as soon as you search for a book, you’re dumped into some god-awful interface.

Thanks for outlining the problem so well. I see a lot of challenges ahead in implementing a solution, notably technological expertise among staff. But developments like Biblio, SOPAC and Koha are encouraging as they achieve more traction.

“# There is no simple way for her to find the intersection of discovering new things that might be interesting to her and what the library has.”

I have a vision of combining an (enhanced) version of the LibX toolbar, with my Umlaut link resolver software (which in turn uses both your local library systems and services like WorldCat), to _add_ information about availability in your specific library to Amazon (and other pages). You could definitely add a link to get such information (oh, I’ve found a book I want, does my library have it? Or if not can I ILL it from my library?) to an Amazon page. You could probably add the info directly on the page too.

This is totally do-able, the building blocks are there, I think I could do it in a month or less of work (that’s a month or less of work on top of all my other usual software custodial duties!).

But even once done, the software you’d have to maintain/deploy locally might be too much for a public library, sadly.

But maybe some of our assumptions are backward.

For one thing, while directors seem to keep paying for proprietary software, they keep hiring programmers who embrace open source and open standards, even on the job. So it’s certainly possible that directors, in general, are agnostic and pragmatic when it comes to software licensing, and, all things being equal, they may even favor open source.

It’s also possible that software companies are slow to respond to requests not because they aren’t big enough — not because they have fewer than Lucia’s 110 programmers — but because they’re too big (and bureaucratic). I tend to get really good customer service from new start ups, even though many of them have just one to three programmers. The key to getting good customer service and useful enhancements is making these requests to the right start ups with the right programmers.

And I think that’s where libraries may be failing. We might well be following the wrong model. We’re doing agile development in a pseudo-corporate environment and holding out hope for Lucia’s 110 programmers. Which very well could be exactly the wrong way to get out of the situation we’re in.

We should be looking at Paul Graham’s YCombinator (here’s a nice audio intro to YCombinator, with plenty of links to more) or the Google Summer of Code model — small, talented, inexpensive, lightly supervised, and unencumbered (by job or life responsibilities) programming teams who are encouraged to take risks. The keys to YCombinator and Summer of Code are evaluation, volume, and incentives.

We have the programmers available to evaluate applicants: code4lib folks have the knowledge and drive, and they excel at making well informed, democratic decisions.

Volume is a question of money, though funding perhaps 20-30 small teams per year would cost a lot less than hiring 110 full-time programmers.

The issue may be incentives, though being given an opportunity to create interesting software for libraries seems to be enough incentive for many outstanding programmers. The prestige of being part of a library-funded, competitive YCombinator/Summer of Code-like program may well be enough to attract talented, hard-working programming teams from around the world who want an opportunity to prove themselves as coders, and help libraries in the process.

We are moving from iBistro to Bibliocommons in the next few weeks. Our customers will get:

- dramatically more relevant results (+spell checking)
- facets for narrowing results
- tagging, rating, reviewing, adding videos
- building lists, sharing lists
- fully integrated my account (which really facilitates the rating and reviewing of stuff you have out or have recently returned)

At this point BiblioCommons give the integrated experience of a vendor’s product (like iBistro).

Search results like endecca or aquabrowser

Well thought out web 2.0 interactivity

Shared user content from multiple libraries – the necessary critical mass to build activity that generates activity.

I am not aware of anything else in the market that brings all these pieces together.

Peter
Edmonton Public Library

The wonder to me is, that given the oft-mentioned cooperative leanings of librarians, they have, largely, not yet embraced open source development.

Excellent post, by the way.