Wednesday, July 25, 2007

Instant Podcasting with Odiogo

I finally got around to testing the Odiogo automated podcasting system (they call it "podiocasting") on this blog-- you may have noticed the "Listen Now" buttons for each post. Try them out and let me know your impressions!

I have been very impressed with the ease of setup for Odiogo: just set up an account, incorporate the provided javascript code onto your site, and it's done. Odiogo's text to speech system works quickly to produce sound files and store them in Odiogo-land, and the podcasting feature works flawlessly in my testing. Shortly after I save a post, the feed appears on my podcast receiver-of-choice (Juice). Nothing to do after initial setup, it's all automatic.

The "voice" is clear enough to understand, and better than most I've heard-- better than the MS Reader voice, for example. The Odiogo voice also skips reading out the punctuation marks, something I've always found distracting in other text-to-speech systems.

I have noticed that listening to the posts out loud is very different than reading them.
It's the difference between writing an essay and writing a speech-- it's possible that this distinction will go away as this kind of technology improves and becomes more widespread.

The company promotes this technology as a way to catch up on blog reading via podcasting while commuting, exercising, etc. Based on my experience, I think they're onto something-- though I'm thankful my commute (long as it is) is not long enough to listen to all my feeds being read!

Powered by ScribeFire.

Friday, July 20, 2007

AcqFest to Add Acquisitions and Serials Features to Evergreen

From the open-ils blog comes word of this week's Evergreen AcqFest, during which a group of coders and developers from the U.S. and Canada are meeting in Suwanee, Georgia to work on the acquisitions and serials system (code-named "Woodchip") for the Evergreen open source ILS.

The communal notepad is up as a Google doc and reveals some interesting ideas are floating about. For example, serials control systems have traditionally used calendar-based predictions for serial receiving, which would be adapted as they were used for the inevitable irregularities in subscriptions. Instead, the Woodchip people are looking into a collaborative system in which libraries would share predictive data for titles. Earlier subscribers, which would have worked out many of the quirks in a subscription, would share it with later subscribers, thus easing the workload. A small but smart idea.

Good luck to the Woodchip team. I can't wait to see the outcome!

Powered by ScribeFire.

Thursday, July 19, 2007

VuFind Open Source Discovery Software Goes Beta

Earlier today, Andrew Nagy on Code4Lib announced Villanova University's new open source discovery tool VuFind is available as beta software (officially release 0.5). Half the blogs and listservs I follow seem to have echoed this announcement by now, so I thought I'd join the crowd.

Checkout the Live Demo to see how well they've incorporated the long list of features librarians are looking for in "2.0" discovery tools, including patron tagging, user comments, links to reviews, and faceted browsing. But it also has (or soon will have) the back-end features it will need to function in the library computing ecosystem: it supports LDAP and database authentication, retrieves holdings status from the local ILS (currently with Endeavor Voyager), with support for other ILSs on the way (SirsiDynix, Evergreen, Koha, and "more to come").
Other features on the way include a script for a nightly data sync with the ILS, and holds and recall features.

It looks to be a very well-designed, feature-rich
Solr-based product, and congrats to the team that created it!

VuFind is just one of several such projects (although a sharp-looking one) coming from libraries today. The pace of these projects seems to be picking up, the feature lists are being refined, and the technical issues ironed out. It will be interesting to see how the slower-moving proprietary vendors compete!