Page Comparison

2022-02-03 Product Council Meeting notes

• Most of the meeting was devoted to a discussion on MARC Authorities and the impact of the implementation of the new 'authorities' entity in mod-inventory-storage
• Khalilah gave an overview that was presented to MM as well and gave a good demo
• Khalilah explained how to search works; what data is displayed

• Presentation

• FAQs

• Recordings for PC are to be found here: https://drive.google.com/drive/folders/0B7G8S7WF6N20c0xjUXZSTGp3UTQ?resourcekey=0-KPI_grz6y_EdssNfJw172A
• Christie: no communication with entity management group, no involvement; there are still concerns
• Jacquie: has raised concerns as well; work that happened should be taken into account; development related to Authority management should have been communicated earlier
• Jessica: Khalilah was aware of the entity management work; the development was just about ebing able to be able to pull authorities in 
• Ann-Marie in chat: +1 to everything that Jessica said. This is not some plan to mess with Entity Management. And the abstraction layer is just to make them searchable in the UI. Not a big deal, and not closing off much more future work by the EM folks.
• Christie: still communication was missing; in chat: But if the abstraction layer was created for marc authorities without consultation, then what is the plan for making sure that the data model for the abstraction layer works for something other than marc authorities?
• Martina S.: the PC discussed the topic to clarify the process around adding functionality, involving the SIG and the PC at an early stage;
• Felix: adding to that, there is the question whether subgroups can decide on their own, without communicating back to the SIG, or not
• Ann-Marie: there is not a way to search in SRS, Invetory does not surface authority data;  that was the reason to add the develop 
• Jacquie in chat: But instead of developing a new abstraction layer, it would have been more impactful to develop a search on SRS data, which was my point in using the dev. time to create something we (SMEs) did not ask for.
  It is a matter of the prioritization of dev work that was not approved by this group is what concerns me.
• Khalilah will be invited for a related discussion to a future meeting

Raegan (Missouri State):

• one of the longest people who have been live
• highlights to think about:
  • do not leave your serials migration until the very end
  • be prepared that it is going to take a while for all processes to be in place
  • some functionality may not be available; if it is, it may not work straight away as expected (e.g. overlay records; linking orders to bib records)
  • Missouri State went live with Fameflower
  • had issues with single record import
  • it took over a year to get some of their processes going; needed external spreadsheet for a while in parallel
  • Feedback: be prepared that the system will not work perfectly for you at day one

Laura (Cornell University)

• continue to have challenged around scalability
• have a lot of data; things get slowed down
• single record import much slower than people expect it to be; that is slowing people down; has impacts on productivity
• unhappy with amount of scrolling and clicking
• someone developed Macros to solve that with Macro Express 3
  • a demo of the Macro creation with Macro Express is asked for → will be scheduled
  • can Macros be shared? → Laura will check that; Jacquie: The community using ME can likely export and share the macros, but they have to be version-specific.
  • Sara: certain key combinations match up with other things and certain characters cannot be typed any more
  • Laura: which version this happened with? → Sara: Macro Express Version 3.9a build 1
• sometimes people put information in two places instead of one (holdings, items) → more customization needed; hide elements that a library does not use
• Non-MARC data not surfaced in Blacklight → deduplication needed; using template to generate brief MARC records; unintentionly this gets edited from time to time and needs to be re-created
• Functionality around Caiasoft is working well
• Jacquie: are you doing pre-accessioning? which data flow is used?
  • Jacquie will check with an ACQ person at Cornell
• Felix: order records will be created as non-MARC records and then a template is used; would it be possible to overlay a source Folio with a MARC record → Laura: yes; the temlates are used when no MARC record is available
• Ann-Marie: libraries that get data from their vendors; they want to update their data with cataloguing data; key piece that is not there yet: there is no reliable match point; working on adding to data import to match to instances, holdings, items that are related to a POL
• POL knows which instances, holdings and items it is related to
• Charlotte: will the functionality include having multiple items; Ann-Marie: that is not planned for Morning Glory; feature is not yet prioritized
• Christie in chat: We are adding the vendor reference number in the other system identifier of the instance at the point of creation and using it as a match point: (GOBI)999xxxxxxx
• Jenn: how is the GOBI integration working for libraries? Not working for them
  • Joshua (Michigan State): having issues and will connect Jenn to the relevant people in his institution
  • Charlotte: GoBi integration. Would that be a topic for a tips and trick page?
  • Ann-Marie: need to get any bugs in Dennis Bridges' scope as he is the PO

Christie (University of Chicago)

• up for 2 weeks
• has been relatively painless
• people are in general really happy with things
• some pain points
• some things they did before the migration: practiced migration regularly
• migrated over a long weekend
• still ran into things on that weekend; 
• unintentionally data was left behind; have still access to previous system so can go and get data
• did load testing (2 or 3 times)
• unintentionally did batch loading at the same time as colleague
• good performance due to load testing; 
• simple record import is working fantastic for Chicago
• not able to delete tags when they have an instance where source is MARC;
• are creating brief Folio instance records at point of order
• hopefully going  to move to Kiwi shortly
• Laura: so VuFind is getting a MARC record generated on the fly from the instance record? → yes
• Magda: what are you using for export → Christie will check;
• definitely doing instance to MARC on the fly
• Magda: we added a few fields for the Lotus release, requested by Chicago; coverage with default mapping was improved for generating records on the fly
• Natascha: most worried about migration how staff would respond to it
• worked really well: teams Folio questions chat; was good because people work apart from each other, remotely
• answers are immediate and reach everyone
• chat died down in the last week so things seem to run quite smoothely
• Charlotte: Chicago has been a great customer; were very well prepraed; several dry runs, erevytime learned something
• Christie: has all been going very well
• Felix: when you were setting up and planning for locations and the 4 level hierarchy: was there many people involved?
• Christie: it was a small group; 3 level structure in previous systems
• got rid of locations they did not want to persist; a lot of clean-up work has been done
• Colleagues from MM, RA and LTS 
• Ann-Marie: great tip to use the opportunity to reset; locations, material types, fund strcuctures that are no longer needed can be looked at, consolidate or restructure
• Laura in chat: Cornell did a lot of clean-up on records for equipment (making them more consistent across locations and, we hope, easier to maintain) (these records do not display to the public)
• Christie in chat: That was on our list, but we did not get to it.
• lessons learned from migarting over and over again: cleaned up MARC records
• went live on first ILS in the 70's; so a lot of data that could not be migrated
• recatalogued a lot of material, if they cared about it
• Jacquie in chat: We are in the midst of an OCLC Reclamation Data Sync to help us make sure that our data is as clean as posslibe.  We have unexpectedly found LOTS of invalidly structured MARC data. We had 216,000 vendor records without LDR 05-07.  Fixing that via the data sync.
• Laura in chat: we had lots of language code errors also
• Felix: one of the few mandatory fields is the title; had some records without a title

MM related use cases for data sync

postponed to next meeting. Intermediate discussion will take place on Slack #metadata-management channel