Notes from the first day of lightning talk presentations at Code4lib BC.
Code4lib/BC: What’s it all about – Cynthia Ng
Go see my presentation post
Canadian Federated Data Repository – Alex Garnett
- FRDR – Federated Research Data Repository
- designing infrastructure for a national research data repository (partnership between Computer Canada CARL and Globus) using Compute Canada hardware
- support: place to deposit large data sets, any datasets, preservation for long term usability, archiving of data sets, discover and de-silo datasets,
- storage can be distributed, owned and operated by institutions. mutiple CC sites could run endpoints
- scalable
- harvester index of data repositories supports: OAI, CKAN, CSW with metadata crosswalked and indexed by the same harvester, generate JSON-LD for search
- goal is to supplement existing repository sites and improve discovery, breaking down silo-ing
- using the UBC digital collections search interface
- want to link out to data, so results points back to original site
- can deposit database using heavily modified DSpace
- authentication through GlobusID, Google, ORCID
- direct web upload interface has been replaced for now
- can sync from local computer or server using Globus Transfer
- major things that work: deposit, preservation, ORCID sign on, etc.
- DSpace license = MIT
Using Ms Access to make vendor reports or local workflows less painful – Trevor Smith
- comes from furstration with results out of ILS with data that can’t be shared
- move to Sierra provided direct SQL queries to database
- why Access? already part of MS Office
- can save, share, store, link to other Office products
- need to share differently depending on audience
- makes summarizing queries easy
- provides an easy way to create reports
- after pulling funds numbers
- to help spending and collections budget targets: can help see how much have spent vs. should have spent
- all about having Access do the math for you
Games Day in the Library – Kevin Brash
- ALA has yearly games day that they promote in November
- convinced WVML to participate
- why games in the library?
- have kids playing games in the library all the time; had user who wanted library to ban game playing, but did not want to start banning content
- but important part of culture
- year 1: mario event. Wii Mario Kart. Was going to connect to ALA server, didn’t get that to work, but kids still had fun
- was difficult to get IT on board because secure network was shared with rest of city
- board games were largely ignored
- lots of student volunteers
- year 2: got a little more budget
- had minecraft installed on computer labs
- two camps of kids: where they wanted to bash each other; creative group
- was going to join up with ALA, and didn’t happen, but that’s okay
- approached a meetup organizer from Extremely Shy and looking for friends (a group who plays board games in a pub)
- older audience (20s) played board games
- year 3: board games came back, and more board gamers
- still had some of the other games though many kids had moved on from WiiU and minecraft
- IT considerations
- a few barriers: particularly network access
What’s in the Bag? Indexing and querying Bags with Elasticsearch – Mark Jordan
- bags: something typically for digital preservation. put stuff in bag and put the bag in long term storage
- standard for organizing content in a directory
- bag has data and descriptive metadata, technical metadata, islandora object, bag creator info
- technical metadata primarily checksum
- have bags, yay! How to organize them?
- no prescribed way
- proof of concept on managing content to scale without having to look inside the bag; just want to index, not have to look inside each bag to know what’s in it
- questions: which bags were created on specific dat? contain specific file in their data direcotyr? with specific bagit version? specific keywords? etc.
- BagIt Indexer: index script puts into Elasticsearch, find: querying of index, watch: watches for new, moved, updated or deleted bags, and update index
- storage location of bags with watchers -> index (Elasticsearch); preservation staff can search index; automated processes that can also query index to use it
- demo: have directory with zips inside, want to index bags, index by specifying bags and what to index which in example is the MODS file
- find command can list all bags with name of bags and location, can tell you what’s in the bag.
- questions: how to integrate indexing and watching services into diverse workflows and storage platforms?
- how to represent changes across versions of Bags?
- how to use Kibana to visualize the index?
Break
Source: West, A. (2012). 001_Snapseed. https://www.flickr.com/photos/cheeqz/7694488774/ CC-BY-NC-ND 2.0
Restoring an interactive videotex art exhibition from the mid-1980s – John Durno
- grew out of project to restore art project in 2015/2016 made in telidon; successful with art exhibit that ran for 6 months
- during the course of the project, found out a number of artists whom work in the same format and thought were unrecoverable
- Microstar runs well in DOSbox
- got in contact with an artist who sent floppy disks
- problem is that some of the art works had some interactive pieces
- much of the hardware is now unavailable e.g. videotex
- had to build the interactive NAPLPS
- DOS Box can map to external ports, which were actual virtual ports, which then interacts with a custom script
- artwork set up using Raspberry Pi
IT things at FVRL: Sphero/Spark robots, Green Screens and VR Experience – Traci Monchamp / Lewis Vacek / Tamarack Hockin
- library of ‘things’
- how to do things in the background to bring digital and new technology to the public
- created a “Playground of things” as part of STEAM initiative
- a lot of new things require more expertise: Ukelele (70), green screen (1 per branch), VR (2), Sphero SPRK+ (112 + 84 coming + 24 for staff use + 21 branch use)
- didn’t realize how popular they would be
- technical challenges: budgets, hardware, software, management
- tablets: needed MDM (which had its own challenges; using Sophos), manage 50+ devices, training staff, lending, cataloguing and processing
- tablets are locked down
- cloud version was completely different from in-house version
- cataloguing was easy enough
- processing was difficult especially cases
ILS Migration Change Management: The Good, The Bad & The Ugly – Janice Banser
- ILS pieces are complicated
- moved from Millenium to Alma
- was on Millenium 20+ years with custom Discovery and Summon
- 2014: vendor presentations, 2015 steering committee, 2016 chose Alma/Primo, 2017 May – Go Live
- change management needed
- huge shift in staff workflows: emphasis move to e-resources
- hired a change management consultant using ADKAR model, training (3 day in-person + sandbox, video), CUPE retirement offer, internal communication
- big gap: faculty; assumed liaisons would get in touch with faculty; really missed the OPAC; working hard to get faculty comfortable with Primo
- Public Services task group, improved documentation, informal user testing
Contest of champions: Headless Battle – Daniel Sifton
- value of throwing into something that you didn’t know much about
- spent a summer trying to add and improve the vendor ebook access checker built by NCSU
- didn’t really know Ruby at the time
- original used JRuby and headless browser, require celerity, csv, highline/import, open-URI
- first effort to move away from Celerity to Ruby with Capybara/Poltergeist; but had long winded error handling, could invoke “headed” browser
- second effort: Ruby with Watir, Watir-Webdrive; but much slower, and error handling was so-so
- third effort: uses Python with BS4 import BeautifuSoup, Selenium Import Webdriver, CSV, time, RE, Requests; use straight scrape without browser headless or not, error handling decent; easy to maintain because in parts
- added GUI: put in vendor input, csv input, and specify csv output
- all available [on github](https://github.com/telezoic]
Getting Started with RFPs – Trevor Smith
- ILS servers within 1 year of EOL
- asked vendor for quote, but too much
- wanted to put out RFP
- got budget lined up, talked to procurement, IT, asked SFU for their RFP
- updated the SFU list, have procurement walk through RFP process, staff review their sections
- timeline: put out RFP in April, wanted to implement by September, and go live in December
- actual timeline: signed contract in October
- stayed with same vendor but did get better price, modern cloud service (hosted Sierra), more packages
- takeaways: get other people involved, talk to: procurement, IT folks, colleagues/staff
Lunch Time
Source: Trimming, P. (2013). Concentrated Eating. https://www.flickr.com/photos/peter-trimming/8566773596/ CC-BY 2.0