In most cases, the suggestion that I travel across the country with my mom to attend a library technology conference would be greeted with incredulous shock. It would be about the same reaction I’d have to the expectation that I take my one and a half year old son across the country. However, that was the position I found myself last Monday, at 5:30AM, driving frantically from Chattanooga to Nashville so as not to miss our flight to Portland for Code4lib 2008.

Why would I subject myself to this? Well, a majority of my family lives in Portland and it seemed like a good opportunity for Che to meet his 97 year old great-grandmother. Since Selena couldn’t go, my mother volunteered to join me and take care of Che while I was in the conference. Surprisingly, despite ridiculously planned twelve hour traveling days, Che was a remarkably good traveling companion. Especially given that the day before we left, he was up all night vomiting from a stomach virus he picked up at daycare. But enough about my stupid traveling decisions for now.

Code4lib 2008 set an awfully high bar for Providence, RI to follow. Jeremy Frumkin managed to find an amazing hotel with both free breakfast and happy hour in the downtown of a remarkable city and somehow managed perfect weather in Portland at the end of February. Is it ok to name Jeremy president for life of Code4lib? Good luck, Brown University, Jeremy just made your job insanely hard.

I would say the overarching theme of this year’s event was “help us with our open source project!”. Or maybe, “fooled you! This actually all about RDF”.   Also, given how this is one of the more premier library technology conferences, we might want to give a scholarship to an MBA or two to teach us how to use a projector and Powerpoint.  I began to think there was a hex on the podium.

Rather than go over the presentations themselves (since that will be done more eloquently elsewhere, I’m sure), I’ll just make some observations. My perspective changed somewhat after Brewster Kahle’s keynote and Rob Styles’ presentation (during the following break) since my role in the conference went from ‘jerk in the audience making snarky comments in IRC’ to master of ceremonies when Jeremy had to go back to Corvallis to care for his sick daughter (or, rather, for his sick daughter and exhausted wife).

Rob was the first to focus on RDF and I think the constraint of our twenty minute presentation slots was both a blessing and a curse here. Rob did an excellent job of explaining why RDF is a good fit for MARC data (or, rather, the metadata that we are currently putting in MARC), but there wasn’t enough time for that and demonstrating why we would want to go through the effort of actually doing it (like, for linked open data and whatnot). It was a good overview of busting the MARC into its conceptual components and making connections between those components, however.

If any of my former colleagues at Georgia Tech still read my blog (hello!), I definitely urge you to take a look at Oregon State’s Interactive Course Assignment Pages (ICAP). This is exactly what I was trying to do when I wrote the subject guides application there. There is a lot of what appears to be NIH syndrome when it comes to subject/course guide applications; in reality I think these applications have to conform to a lot of local needs and expectations which is why “somebody else’s code” doesn’t always do the trick. ICAP is so similar to what Tech is running (although executed much, much more elegantly) that I don’t think there would be much of a leap for them to migrate. As an aside, the developer (Kim Griggs) mentioned that it took four months from proof-of-concept to production. I find this an amazing attribute of Ruby of Rails. The Umlaut took slightly less than six months. I should add these were both one person development teams.

Next up, David Walker talked about the new Worldcat API. What I found interesting about his talk is that he said his initial interest in the API was to be able to build ad-hoc union catalogs when he needed to. This was my original reason for trying out the Talis Platform when I was at Georgia Tech (although I could never figure out how to manage getting all the libraries I wanted to work with to get their holdings in a Platform store), so it definitely resonated. With the gravitational pull that OCLC already has in the U.S., this idea of ‘union catalog for specific problem set’ could really flourish. Of course, since not every library can afford to put their holdings in Worldcat (and therefore have access to the API) it may not be all that useful to the sorts of libraries that would gain the largest benefit.

While I was impressed by Winona Salesky’s and Michael Park’s presentation on XForms (there were definitely some slick features there!), my takeaway was that this technology seemed too difficult to deploy (when your options are browser extensions or Java servlets, I’ll pass on that). Am I missing something here? Java purists, you are free to mock me and my scripting language ways.

Interestingly, the Zotero presentation removed pretty much all references to RDF and the Semantic Web that appeared in their proposal.

The breakout session I attended on Tuesday was proposed by Jonathan Rochkind (who I co-presented with on the Umlaut) about finding a way to isolate only the open access content in OAIster. While maintaining a separate index of this data might be useful (sort of like what IndexData does, but they also include the non-free material; I am also harvesting some of this data with the intention of putting it into a platform store… but whoa does the clean-up of the data take time…), the group eventually decided that a web service that identified whether or not OAIster results were OA or not (via whitelists, blacklists, etc.) might be an easier first step. I’ve created a Google Group to carry this discussion forward here, if anybody is interested in participating.

I can’t remember which lightning talks happened on day one (is there a list published anywhere?), but it doesn’t matter. The day two lightning talks showed why having this style of presentation is so great. The reasons can be summed up by pictures of underpants and player piano midi files being played from musical scores generated from OCR. When the video files are up, definitely check those two out.

Did I mention that there were free drinks from 5:30-7:30 every day?

Karen Coyle opened day two with her keynote on RDA. Corey Harper closed the formal presentations the next day on the same topic. It was pretty neat how they managed to not step on each other’s toes. I think their message resounded pretty loudly: people like Code4libbers need to get involved in RDA and DCAM to ground it in reality. Ok, I’ll see what I can do.

LibLime’s MARC editor was quite nice. Granted, I’ve never used OCLC Connexion, but I would have to think something like this would be a strong competitor.

I’ll skip over Aaron Swartz’s presentation (based on the number of Flickr photos of his presentation, I figure there must be plenty about it already), but I plan on writing a bit about ThingDB and other document-centric databases really soon.

Skipping up to the DLF ILS API, both Emily and Terry’s presentation as well as the breakout session on this were incredibly useful.  I feel like the DLF has a pretty pragmatic approach to working on the problems of interoperability and Emily and Terry were really good emissaries to explain their goals to Code4lib.  During the breakout I introduced Jangle and later gave a lightning talk on it.  Since the cat’s out of the bag, I absolutely will have a post about that this week.  While I don’t exactly trust the vendors’ response to the DLF proposal (they meet on Thursday), I think that the potential of this is mutually beneficial to every party.  Jangle was quite well received, by the way.

Omeka seems exactly like what Greenstone promised to be, but wasn’t quite exactly (at least, in my mind).  I really like the focus and polish of this project.

Day three is where things started slipping for me.  Not only was my brain quite full at that point, but Che’s stomach virus apparently had taken a grip on me (and I apologize to anybody I paid that forward to).  I did manage to get out to see my family that night, though (who may not be as thankful to have seen me if I spread it to them, too).

This is a fun conference and I’m really proud to be a part of it.  Thankfully, Oregon State had everything so well organized that things still hummed along fine despite losing our ringleader.  I had a good time MCing, but we lost all of the dignity and professionalism when Jeremy had to leave.  “Slackerly” and “clownish” are more apt for me.  I know Roy Tennant and, sir, I am no Roy Tennant.