Library Day in the Life - 5 - Day 1 - ORIGINAL TITLE!

So here we are again. Library Day in the Life number five! Monday is a work from home day. No audacious commute from Toronto to Hamilton today!


Morning soundtrack: BBC World Service podcast, TWiT - It's The New Sex Talk, CBC Spark Daniel Pink on Motivation 3.0, The Protomen - The Protomen

Catch up a bunch of email from last week, and finally got around to setting up Drush. Don't know why I never got around to it before, but it definitely worth the time of checking out if you manage a few Drupal sites. Watched a couple of screencasts on Drush by CivicActions to quickly immerse myself, then got around to updating modules for our dev site. Once everything was up to snuff, I started working on the a cuple of our final functional requirements for the new version of our digital collections site before we start theming it; allowing each record to have its own Dublin Core XML output and adding some Dublin Core meta information to each record's header html output. Mind you, I am a horrible programmer.

The header output code was pulled mostly from this Computed Field php snippet example. I managed to get DC.title, and DC.Date.X-MetadataLastModified working correctly, but the rest of the elements (descriptions, source, format, etc) were another beast entirely. I put off the Dublin Core XML until later in the day when I could rely one of our programmers for assistance, because mind you, I am a horrible programmer.


Afternoon soundtrack: Squarepusher - Hello Everything, Squarepusher - Just a Souvenir, Daft Punk - Discovery, Film Junk - Inception (spoilers portion of the podcast), BBC World Service podcast

Thought out the spec a lot more for the Dublin Core XML. Decided not to use CCK Computed Fields to make it happened. Don't know why I was thinking it would work, but one of those square peg in a round hold things. Contrary self - I could just make the peg round. Brainstormed a lot more with Matt (one our dynamic duo of programmers) on the Dublin Core XML idea. We agreed we just create a quick module to handle creating the XML. This will be our first custom work with the new version of the site. Due to many problems with the last iteration, while current production version, I wanted to move as far away from custom code as possible and we have been doing very well. But, this makes sense... maybe. There is always a million ways to solve something like this. Maybe tomorrow it will just be a View with a php snippet.

In the background of all wretched coding on my part, I was again working with my favourite module - Views Bulk Operations (VBO)!!! With the first iteration of the site, we made a couple of decisions that I have come to regret. They are not earth shattering or anything, just didn't setup some of the metadata fields how I would have liked them to be setup. For quite sometime I've been trying to thing about an easy way to merge some of them together. Epic mysql query dreams! JOIN, JOIN, INSERT, UPDATE, WHERE, BLERG! Anyway, some wonderful soul wrote a merge fields action for VBO! So, in the background all of today's work, I updated 14559 rows, a couple of times. It only took an average of 12153468ms each time!

Oh yeah, email was answered. Spheroidally.

2009 OLITA Digital Odyssey

I must say that the Digital Odyssey was the best one day event I have been to. Just a fantastic day with fantastic people talking about awesome projects. It cheered me up and gave me hope in these crap times. Best part of the day had to be Mike Ridley's keynote speech - The Age of Information is over. It is time for the Age of Imagination. It will be the library's job to nurture and foster creativity.

Workshops attended:

Walter Lewis - The Perfectibility of Data. I must say that Walter may be a bigger metadata fascist that I am. He showed some cool stuff that I didn't know about - media rss feeds! Then using Cooliris to visualize said feed. Also, finally realized how simple it is to provide proper data to interact with google earth & google maps. Just latitude and longitude coordinates!!!

Loren Fantin - Planning and Managing a Digitization Project. Lots of great stuff in Loren's talk. Don't see a blog entry on the Digital Odyssey site yet, so no link. Biggest lessons learned - scope creep!!! & digitization should be apart of collection management.

Art Rhyno - OCR Options for Scanned Content. Great session on the basics & overview of OCR, and OCR software options. Provided many examples from a variety of OCR software packages. Mostly ABBYY & Ocropus.

The text to my presentation, pdf of slides, keynote file, and powerpoint file.

blog image

library clouds in the sky with [diamonds]?


Sorry, had to get that of the way. Those who know, know. Those who do not, oh well. I will address it later... subtly???

Awhile back we got hit with the perfect-downtime-storm. A RAID controller battery randomly failed, and I was down for quite a few hours. Then a day or two later ... a brown-out occurred. Somehow, some way, this killed the brand new RAID controller on the DB server, and disemboweled the RAID controller on the web server. I was down for almost a week awaiting repairs vendors and IT. During this period of utter embarrassment and fury, I finally took somebody up on a long-standing offer to put all of my digital collections stuff on a BEEFCAKE server. I ordered my [twin node] BEEFCAKE and decided that high availability and redundancy was the way to go.

So, I began building a proof of concept: Tomax & Xamot [LAMP with hint of wonderful Tomcat, Java, Solr, and Djatoka for blooming ideas] are my sinister production machines with Heckle & Jeckle [HAProxy & KeepAlive] providing the load balancing. After many hours, the proof of concept succeeded. Kill apache and/or mysql on Tomax, Xamot will be right there still fighting for the Cobra Commander.

I've been sitting on BEEFCAKE for a week or so, almost ready to go to production. But for the last week, I have been diligent with my 99-part hearty diet of bacn, Batman, Green Lantern Corp, and Promethea. Combined with the nicotine patch, my head has been in the clouds - in a good way. I was pretty undecided about the Cloud for a long time and Stallman's talk at U. of T. threw me even farther to one side of the fence {GPL loophole], but Fink's idea-machine-brain rambling on about creating a Cloud at LibMac (another possible proof of concept) started turning gears. (Side note, Fink is more rabid about Open Source than I). The collections within the Digital Collections, (namely PW20C, Concentration Camp Correspondences, Bertrand Russell, Canadian Publishing, et al) are sitting on a fair chuck of metadata begging for something to be done with it. Add that to the Mass Digitization Project (DC, METS/ALTO, and fingers crossed TEI), and EVERGREEN!!! Oh what to do, what do to???