Code4Lib North: Day 2 Morning Notes

Morning presentations

Digital Humanities Project Showcase

• Andrew McAlorum & Ken Yang

Sorry, missed a bit. Was strangling my browser to work properly.

Uses

• Drupal
• Solr
• Blacklight
• Virtualization using Chef

PM Tools

• JIRA: Issue Tracking
• Confluence: Documentation
• GreenHopper Agile

Projects

• DEEDS: Documents of Early England Data Set
• RPO: Representative Poetry Online

Design Challenges

• visualizing data and its relationships
• making data accessible – primary target audience: scholarly research
• discoverable

Features

• responsive
• cleaner look
• browse list
• DEEDS
  • two column view: text transcript of original / original
  • visualization e.g. map charters – spider view of pointers, but working possibly on combining pointers into one pointer with number
• RPO
  • create own anthology in PDF, and share/print, which can also be sent to UofT Espresso machine to have it bound
  • can put own commentary
  • long term goal is the social aspect

Find Your Books Fast: Yet Another Bookfinder Presentation [my subtitle]

• Steven Marsden

I will add any new material, but otherwise, please see the full write up.

The basic information is pulled from Summon using the API, but the availability is screenscrapped from our OPAC.

The Library in your pocket: Responsive Catalogue at UofT

Bilal Khalid & Gordon Belray

• Faceted searching with Endeca(sp?).
• clean interface
• display only title/author with expandable details
• mobile view: hide facets
• still in the works: missing some core functionality e.g. MARC record display, stackmap, responsive images
• need to make sure works in older version of IE

Break Time

Credit Card Sized Computers?

Giles Orr

• many different small-sized computers e.g. Raspberry Pi
• Wireless Router: can install linux
• most run between 2-3 watts
• will be spending more than base price to buy cables, power, etc.
• most have flash memory, but otherwise, have to boot off USB

Big Giant Box of Disks: Mass Storage on the Cheap with Backblaze Pods

• John Fink

Slides

Open sourced design of hardware case: backblaze pod, which stores 45 hard drives. One part of the case is for a computer. No lights to indicate drive failure. Using RAID 5 or 6, very loud, not very fast.

OCLC Ate My EZproxy? …then charged me $500/year for the priviledge

• Tim Ribaric & Jonathan Younker

It is fairly minimal price for something that is very crucial to our work, but it’s the principle of the thing.

What else is out there? What are our alternatives?

Options

1. Pay the man
2. Have OCLC host it – but need to call in, added db definition will take 48 hours
3. Change to other commercial proxy – only heard of alternative is WAM of III OPAC, which isn’t very powerful
4. VPN – can be a hassle
5. Shibboleth – but this goes through central IT in their instance, also difficult to understand
6. Roll your own

Roll your own

1. Squid
2. Apache mod_proxy
3. LAMP (PHProxy, Glype, etc.)
4. Python +/- Twisted
5. nginx
6. Node.js
7. PirateReverse.info
8. ???

Need to run your own data centre though. Looking to see what’s the best fit for the library.

Comparing ISBN/Related Works APIs: LibraryThing, OCLC, and Open Library

• David Fiander

LibraryThing

• Can use LibraryThing API passing a link with ISBN and returns ISBN of all the different editions
• Non-commercial use only
• max 1/sec + 1000/day
• has good quality

XISBN by OCLC

• can get bibliographic information as well as ISBNs.
• Allowed for managing and enabling access to library resources/info, but cannot use bots, spiders, or other automated information-gathering devices.
• max 1000/day as well.
• has odd errors

OpenLibrary

• Can provide ISBN, which provides work ID, then resubmit with work ID, which provides all the ISBNs.
• means it’s a 2 step process
• Can download a dump.
• no license/rate information
• possibly lower

Not actually a technical question. Requires a lot of paperwork, etc. Other issues include data quality and integrity.