- David Bott, SCPL
- Sirsi-Dynix
- Horizon upgrade in 2006
- aging hardware
- end of life for Windows 2003
- sub-optimal use of resources
- Goals
- virtualize server environment
- upgrade OS to Windows 2008 and 2012
- faster disaster recovery
- optimize resources
- purchased Dell PowerEdge VRTX
- 4-blade chassis (5U)
- integrated 25-bay SAN
- redundant power supplies
- using VMWare Essentials Plus
- vMotion - move servers without interruption during maintenance
- thin-provisioning
- for guest OS used MS Windows Server 2012 R2 DataCenter
- bought through TechSoup Canada
- can have unlimited VMs per blade
- added backup solutions
- Dell DL-1000 - multiple snapshots
- Server-to-NAS (2008)
- Total project $42K+taxes
- included all project management, configuration and installation
- Dell was very good in scoping out the details and implementing the solution
- after things in place then did physical-to-virtual conversions for multiple servers
- still need to migrate
- domain controllers on Windows 2003
- IVR server (Talking Tech)
- all eggs in one basket - VRTX hardware failure takes down all servers; still have offline capabilities for CKO (PC Reliance); but also susceptible to other failures already (network switches, Internet, Hydro); Dell has 4-hour on-site response time
- discussion about documentation required and how big shops vs. small shops handle it
- discussion of blade servers
- compared VMWare vs. HyperV - similar pricing, but went with VMWare because more industry standard and mature; also VMWare had 5-year support package
- can't afford redundancy but spend on resiliancy - e.g. have 1 Internet connection, but have regular backups of important servers/data
- none of the cloud services have patron data - SCPL does the authentication using SIP server
- Whitni Watkins
- @_whitni
- community crowd-funded a Mac for Terry Reese so he could build MarcEdit in Mac
- preview build released last week
- great for making bulk edits of records in ILS - use regex
- Mac UI very similar to the PC version
- http://bit.ly/1HPKoSt for work in progress
- MARCTools is done; now working on MARCEditor
- requires Mac OS X or greater
- required Mono (.NET) MDK before installation of MARCEdit
- Alison Hitchens
- model of bibliographic data using linked data framework
- designed to replace MARC
- can be used to transform existing MARC
- can be used for new bibliographic descriptions
- core classes are: creative work, instance, authority and annotation
- www.loc.gov/bibframe/docs/model.html
- annotation = resource that enhances our knowledge about particular item
- modes of investigation
- read official docs, white papers, etc. at LOC (www.loc.gov/bibframe)
- BIBFRAME discussion list
- webinars (LC, ALCTS, etc.)
- Zepheira practitioner training ($$$)
- experiment with tools
- bibframe.org/tools - e.g. Comparison tool shows MARC vs. BIBFRAME output for record
- BIBFRAME Scribe from Zepheira (editor.bibframe.zepheira.com/static) - uses real-time calls to LOC linked data for the drop-down lists within the form
- BIBCAT (http://bibcat.org/)
- there are discrepencies in data output from tools - some data cleanup is required with the results
- study vocabulary
- bibfra.me vocabulary navigation tool
- see what others are doing e.g. BIBFLOW (http://www.lib.ucdavis.edu/bibflow/)
- https://renegadebibframe.slack.com
- https://github.com/RenegadeBIBFRAME/Vocab_review_docs
- Some barriers to participation
- do you know about data modelling?
- do you know about ontologies, vocabs, schema?
- do you know how systems use the data you create?
- have you done bulk data clean-up or analysis?
- do you know what others might do with library data?
- do you understand linked data?
- discussion about support of BIBFRAME in ILSes at present and in future
- discussion about how quickly BIBFRAME is being adopted
- David Fiander
- doing a research project of tracking researchers' personal collections vs. campus library collections
- List is programming language - uses parantheses for content separation
- Lisp started as "list processing" but has other data structures
- Lisp very good at memory management
- Practical Common Lisp (http://www.gigamonkeys.com/book) - available in Springer book collection
- there is no native Lisp library for Z39.50 - but there is a C library (yaz/zoom.h)
- foreign function interfaces in Lisp - i.e. call the Zoom C functions
- Emacs works well for programming Lisp (used SLIME https://github.com/slime/slime)
- dynamic environment in Lisp allows for interactive debugging of the C library
- restas is a framework for developing web applications in Lisp
- quine = a program that prints the source code of the program
- Lisp is compiled language
- Jeff Proehl from Canisius College, NY
- [email protected]
- used project to teach himself new things
- Canisius uses Summon as the discovery layer
- had to add the Where Is It? button in the Innovative WebPAC not in Summon record
- used combination of HTML5/CSS3, SVG (to draw map), JS (jQuery), JSON and PHP
- why they used SVG?
- each to maintain
- uses familiar language
- accessible
- allows for RWD
- open standard (included in HTML5)
- works in modern browsers, including mobile browsers
- used CSS3 :target to style item referenced through inpage anchor
- e.g. url#roomname will open map with roomname selected
- used JS to handle interactive selection of map content e.g. change picture based on room selected
- all room data is stored in a JSON file - the room ID matches the SVG room ID
- future - give users different result if item unavailable; add book information incorporated into the floor plan (esp. helpful for people using mobile)
- can use https://inkscape.org/en/ for creating vector graphics
- Warren Layton
- RVM = Répertoire de vedettes-matièr (from Universite Laval)
- LCSH = Library of Congress Subject Headings
- CSH = Canadian Subject Headings
- RVM is a French adaptation of LCSH and CSH
- two services
- access to RVM authority data (online and/or FTP access)
- translation service ($$$)
- Program or be Programmed, Douglas Rushkoff
- toolbox
- yaz-marcdump
- language of choice that can handle MARC
- RVM (and understanding of licensing)
- database (to store RVM)
- questions raised
- what if multiple matches in RVM for a given heading?
- generate file for records with unmatched headings?
- character encoding?
- Warren required RVM subscription to get the vocabularies
- discussion regarding the copyright issues surrounding metadata and bibliographic records
- Sara Allain and Thomas Guignard
- downloaded data set from http://cheese-fromage.agr.gc.ca/, but listed businesses didn't have address
- used the Google Places API to lookup the business name and province to get the lat/lng
- limited the data to 200 artisinal cheese makers
- used Open Refine for data cleanup and exported using templating option to create GeoJSON
- used LeafletJS to map the data with the GeoJSON file
- sallain.github.io/can-cheese-map
- Be it resolved that libraries need to invest in more cloud based services
- In favour
- expertise not in house and cost can be minimal
- makes information available everywhere on all devices
- using services that consumers already use (e.g. Google or Microsoft mail)
- Against
- definition of 'cloud' = give up control of system in name of simplicity
- have to go through multiple levels of vendor/IT to get stuff done
- privacy laws (e.g. Patriot Act)
- campus IT can stop library IT from having fun/experiment
- need clear agreement and pricing on 'exit services' i.e. exporting data when leaving the service