Code4Lib North Meeting at Ryerson

I attended my first Code4Lib North conference this past week, something I’ve never quite felt “geeky” enough to do in the past.  This year an implicit invitation to technical services librarians encouraged us non-coding types to participate.  The result produced a gathering rather similar to what I’ve experienced at Access conferences.

Kudos to the folks at Ryerson for hosting and putting on such a well-organized event.  It was great to meet and learn from participants from the local and regional community and the unconference/unthemed style worked very well.

There was a wonderful flow that emerged on the first day beginning with Alan Harnum‘s talk on the idea of Computational Thinking.  This served to set up the inclusive tone of the conference nicely noting that although some of us were not coders our activities as librarians, such as cataloguing and project management, draws on and applies computational thinking processes.

Alan drew on Jonathan Rochkind’s blog post as inspiration for his talk and outlined the steps of the computational thinking process: decomposition; pattern recognition; pattern generalization and abstraction and, algorithm design.  He emphasized the importance of analysis during the decomposition phase noting that because project failure can be so far along in the process considerable work can be done before it’s realized that this might not have been such a great idea.  Breaking the problem down allows potentially fatal flaws to be identified and corrected earlier in the life of the project.

Peter Senge’s book The Fifth Discipline came up during the Q & A and although coming out of the computing environment of the 1980’s, and therefore also pre-internet sounds like it might still be worth investigating.

Alison Hitchens then took the floor and talked about RDA.  She reminded us that the purpose of RDA is to create “data for discovery” and that the rewriting of the cataloguing code has resulted in a clearer identification of the bibliographic data elements.  She talked a bit about the RDA Vocabularies and how they have been used by the Bibliotheque Nationale de France in their linked data experiments.

MJ Suhonos, one of the conference organizers and a fellow linked data enthusiast, then talked about metadata formats.  He evaluated a number of formats against the criteria of serialization, encoding, and schema.  The JSON MARC-HASH format seemed to satisfy all three and is considered one of the better formats to use.

Cynthia Ng, another conference organizer and our host during the event, talked about the importance of web accessibility and accessible user-interfaces on different platforms.

Katie Legere from Queen’s presented a very interesting session on the “sonification of data” a perspective I had not considered despite my own musical background.  This is a relatively new area of research and Katie provided an example based on reference question statistics from a multi-branched library using snippets from Beethoven’s third symphony to sonify and express the data.  It struck me that combining this with data visualization might also be applicable in a future navigation scenario.  I’ll talk a bit more about this idea in a subsequent post.

The final “formal” presentation of the day was by Nick Ruest who talked about his recent work archiving web information from York University‘s news source Y-File and resources associated with the Dale Askey/Edwin Mellen controversy.  Nick introduced us to web archiving and the use of WARC files and their importance as historical sources for researchers in the future.  He also talked about the importance of advocacy and the librarian’s role as the “social conscience of the information world.”

Nick’s session inspired a reply and very informative talk on hashing and digital file identification/verification by Mat Trudel.  That Mat was able to quickly throw a presentation together and deliver it to the group is a testament to the real beauty of the unconference format.

The afternoon began with a detailed overview of SFX by Dana Thomas and then discussion groups and hackfest sessions began.  I sat in on the discussion about University Librarians and technology led by Mike Ridley.  Mike asked two main questions: what should the UL know about technology?; what should the UL know about those who work with technology.  Some highlights from the discussion: trust your staff especially when weighing information provided by vendors; also trust reports from staff at other institutions who have first hand experience with products you are considering; attracting and keeping “talent” when money is not available as an incentive; importance of fostering a culture of “yes”, i.e. to provide freedom to experiment and participate in projects that might not benefit the library directly; if you are operating in a unionized environment consider the union a partner rather than an adversary.  It was a good exchange of ideas.

That evening was spent in good company. Bill Denton took us over to the fabulous Arts and Letters Club on Elm Street for drinks.  Some computational drinking ensued with Ian Milligan, Alan Harnum, Giles Orr, Abe de Jesus, Anna St. Onge, Nick Ruest and Bill. We migrated to the Queen and Beaver for dinner managing to grab a table for nine in a small private room. Excellent!


Day two began with Andrew McAlorum and Ken Yang talking about a couple of digital collections maintained at the University of Toronto:  DEEDS (Documents of Early English Data Set) [still in development] and the RPO (Representative Poetry Online) Anthology.

We then saw a demonstration of an app called Book Finder that Ryerson has developed to help library users find physical books on the shelf. Here’s an example of the app at work:

The following session was about developing a “responsive catalogue” that is capable of rendering itself in a good way on any device that might be used to access it.  This is something that will be launched at U of T next month.  If you want to test your own site they recommended a device emulator called the Responsive Design Bookmarklet.  They also showed us how Chrome can also be used to do some of this testing through the built in user agent/device metrics feature.  And they noted that it is best to try things out on the actual devices.

Giles Orr from TPL talked about a number of “credit card sized” computers including: the Raspberry Pi; MK802, the Beaglebone Black, and the TP-Link MR-3020 wireless router. A fascinating colection of gizmos if ever I saw one …

John Fink showed us a way to do storage on the cheap with an array of consumer grade hard drives and a BackBlaze Pod. Effective but apparently a tad noisy.

Tim Ribaric and Jonathan Younker, both from Brock University, talked about the recent acquisition of EZProxy by OCLC and what this means for the authentication of off-campus access to library licensed digital resources.  I hadn’t heard about this development and was glad to see this discussion of options to paying the small annual fee that OCLC will now be collecting.

David Fiander then looked at three ISBN APIs comparing Library Thing, xISBN and Open Library. Interesting to see the different results obtainable through each of these services.  I believe it was Open Library that demonstrated the most flexibility.

After lunch there were a series of Lightning Talks beginning with a short discussion that I “led” on navigation and visualization in a linked data environment.  It was a half baked idea to be sure and I’ll elaborate on it and the discussion in a later post.  Other talks included: a very promising project that MJ Suhonos is working on called Ladder; David Fiander talked about using WebDav with Zotero; and, Dale Askey talked about using Passpack and YubiKeys to manage passwords and authentication.

I had suggested a hackfest dealing with the conversion of MARC records into a format(s) that could be employed in a linked data environment and was lucky enough to spend some time with MJ and see some of the inner workings of his Ladder project and the conversion processes he’s used there.  We also took a quick look at a couple of open source conversion tools: marc2rdf and Ross Singers’s marc2rdf-modeler.  Alison Hitchens, Mat Trudel, and Liam Whalen also took part in this discussion.  Many thanks to everyone for bringing me another conceptual step closer to filling the gap that exists between connecting our bibliographic data to the semantic web.

A fantastic couple of days!  Thanks again to Ryerson for the wonderful hospitality!!