Code4libBC Day 1 talk notes.
Giving back to the community through transparency and a public handbook
See the slides and the whole script: https://cynthiang.ca/2024/05/16/presentation-giving-back-to-the-community-through-transparency-and-a-public-handbook/
Twitter archiving, post Twitter…
Speaker: Daniel Sifton, VIU
- Other talk credits: Dalys Darney, Dana McFarland, Sarah Ogden
- leveraging open source tools
- for many years archiving tweets on covid and wildfires
- a lot of manual work, not a lot of capacity, increasingly unreliable because searching images, so need to describe them
- have a few thousand items with varying
- looked at different ways of managing workflow, such as DocNow, which lets you archive tweets
- didn’t work great, need twitter API account
- the search to archive was often broken
- Fall 2022: meltdown
- we were in this space before, and archiving own work
- started doing 2009, ended 2022
- downloaded archive from twitter
- had 10 years of tweets
- had 7000+ files/folders
- was hoping could do something more lightweight
- looked into Harvard LIL thread keeper, JS app that will generate PDF and such
- plugged it into a DO droplet: oldtweets.viulibrary.ca
- doesn’t work with video, but will take a screenshot of video
- tweets in JSON
- used jq to pull links into a CSV file
- created a CSV file, one for each year
- used scraping tool to interact with save your threads app to plug in information
- used open PDF to amalgamate into one PDF per year
Saving circ data in MS Office during a system change
Trevor Smith, Douglas College
- moved from Sierra to Folio
- could transfer MARC and item records, but not book usage and price
- no historic circ data in Folio = no weeding
- Panorama (Tableau) has limited = limited choices, not great for weeding analysis over a decade
- Price data is stored in a note field
- exported all records from Sierra, save data in Access, dashboard in Excel
- Excel: pivot tables, charts, not great for storing data or large calculations
- Access: good at storing data, but not great to interface with
- Pivot table: quick way to summarize data
- has a lot of data in codes, use Access to “translate” to human readable info with SQL
- extend and normalize data
- can create charts with slicers in Excel with a connection to Access
- which then generates weeding list
- can also do it with raw CSV
Find a Space: A tool to help student identify and locate the perfect study spot, with a focus on accessibility and usability
Andy Peterson, Western Washington University Libraries
- building is a maze
- accessibility issues
- developed the tool partly to address the issues
- Phase 1: get information, images, maps and written directions on how to get to the spaces
- built in Drupal
- content made by student employee
- Phase 2: added library and other student services available through the building
- added restrooms
- Drupal with content types and taxonomies, direction chunks (decrease repetitive data), using feeds: loaded initial spreadsheet data and for ongoing hours import
- usability: interviewed student employees (no one knows what a “carrel” is), especially terminology
- accessibility issues a primary factor: wheelchair, assistive technology, visually impaired
- accessibility interview: got recommendations for different elevators, directions, great feedback on facilities
- things I would do differently: develop vocabulary and phrase guide earlier (example: turn right vs. take a right), more usability testing before started
- phase 3: maps in the catalogue, reuses a lot of the data, which was possible because we did it in chunks
- repurposed content, added new taxonomis and added them to maps, built new views to dynamically load maps
- in primo: borrowed code from CWU, did some JS customization
- next steps:
- room reservation (through sprinshare)
- more buildings
- more usability and other improvements
- more accessibility features (real time busyness, geolocation, lighting, focus on visually impaired and neurodivergent users)
- write up detailed “how-to” for Drupal piece
- spoke to the accessibility access centre, student group
Making it count: Implementing a Time of Flight People Counting at the UVic Libraries
John Durno, University of Victoria libraries
- security gates at front entrance
- people gate counter, inside one of the gates
- Everyday, one of the staff records the number and sends it to the business analyst
- unfortunately, can’t help with decisions, can’t break down by hour
- alternatives: new gates (expensive), EvolvePlus (bolt new sensors, talk to wireless input, server; but janky), Terabee Flow Counting (network attached sensors, power through ethernet, configurable)
- had to do privacy and security review
- need two sensors (subordinate, coordinator) over the top of gates with overlapping field of view
- workstation running Terabee management software > 2 sensors over gates > JSON (https) > web server > storage > sql table > analyst’s workstation running Power BI
- all local, no cloud
- security considerations:
- dedicated VLAN, restricted ACL
- disable wifi on the sensors
- restrict web server access via IP
- ensure sensors do not contact vendor servers
- stop all non-essential services on sensors
- strong admin password
- data shows the number of in and out since last reset (usually once a day)
- sending data once an hour, though can be set to shorter intervals
An SFU Library Dream team: Collaborating for a better public website
Janis McKenzie, User Experience Librarian
- manage online presence including
- Drupal
- in charge a lot of
- usability, online accessibility, training staff on publishing, etc.
- own general webpages
- while can look at stats, not the same as seeing it in-person
- go to the “dream team”, all library technicians, lots of info on where students get stuck, use the website for providing info
- ask once a week how reference and instructions is going, notice any trends
- together, try to solve problems
End of Day 1 talks
Seal See you tomorrow!
