Meeting Minutes for 05/26/2021

Announcements

• Aims Community College went live this month
• Cataloging for Non-catalogers training next Thursday
  
  • https://marmot.org/content/cataloging-non-catalogers
• OCLC Connexion 3.0 release in July/Aug
  • Support for old version ends in 2022
• Marmot is moving our documentation to Confluence. 
  • Revising documentation
  • Creating new documentation of the load profiles

Duplicates Team Update

• Asked III if we could add the 024a into the v index with 020a.  
  • They said this would require a reindexing project that would cost money
  • We could start collecting ideas for a reindexing project <<ACTION ITEM
• Brandon worked with Duplicates Team on new versions of the duplicate record finding tools
• Discuss the possibility of developing a method to track the number of duplicates over time so we have better data on the problem
• Discuss the possibility of a tool that would find bad dedupes
• Is the duplicate problem bad enough that we would like Marmot to hire a part time deduper again?
  • Lisa says we probably don’t need to hire someone because so many of our dups are intractable.
  • Lloyd says maybe a person focused on this task would make fewer mistakes than those who only do it now and then.
  • Rebecca and Carol say they like the idea.
  • Tallie asks why the position was eliminated.
    • Brandon hoped we could use more automation to keep duplicates under control.  He created the Tableau duplicate reports so the rest of us could be more effective at cleaning them up.
  • Nina says we don’t have good data about the problem, so we should try to gather information about how bad the problem is.  We can’t ask for money without data.
  • Lloyd says we should start by collecting data.  With data, we could experiment.
  • We could try a new policy and see what happens to the data, or we could do more training, or we could hire a temporary person and watch the data.  Without data, we don’t know what effect anything has.
  • Jamie says it is difficult to know how many bad dedupes there are since we only find them randomly.  There is no systematic way to find them.
  • Lloyd says we could record when we find them and keep track of them.  That is really hard to measure.
  • Lloyd has brought up with Brandon an idea for a tool that would find bad dedupes.  It could look for cases where there are item types on the same record that don’t belong together.  That would be complicated because it would have to know what every library uses to describe each thing.
  • Jamie says they did a project where they found a lot of records where the record in Sierra had a different title from the record with the same 001 in OCLC.  That’s another problem.
  • That was another project Nancy worked on, we think she did that for Adams.
  • If people want to work on a project like that, Lloyd can help you set it up.
  • Selene asks “Do we make a list of the different issues and see what we use to find them.  Look for commonalities and patterns between them”
    • Lloyd: No we don’t have anything like that.  We could start one.
  • Lisa says it would be great if we had a shared document for collecting issues.  She says she will create a google doc.
    • We have some discussion of whether everyone can access google docs because some institutions don’t use google.
    • We determine that anyone can access google docs at least anonymously, so we will try Lisa’s doc as a starting point.
  • Nina chats “It seems there are at least these variations of the problem: 
    • 1) multiple records on the database with the same ISBN/UPC/BUN; 
    • 2) bad deduping, e.g. Playaway on book record; 
    • 3) incorrect match between Sierra 001 field and the OCLC record #.”

35 minutes

FOLIO Update

• Virtual WOLFCon is next week
  • http://openlibraryfoundation.org/about/wolfcon/upcoming/
  • Nina asked what is WOLFCon?
    • It is the World Open Library Foundation conference.
    • This is the organization that is doing FOLIO, ReShare, VuFind, and some other projects.
  • Lisa asked if we are sure we are going for FOLIO?
    • No, nothing is sure.
    • The only Marmot who is planning to try FOLIO is Western and they have not committed to a date.
    • Western is waiting for the INN-Reach integration.
    • Any transition would be library by library.
    • We are working on setting up our own FOLIO test server.
  • Jo says there’s no FOLIO for consortia yet.
    • This is correct.  Most consortia functionality is being worked on in the ReShare project, which is setting up a system for peer-to-peer resource sharing between libraries on separate servers.
    • There is a consortium SIG.  They are talking about what a consortium would need out of FOLIO.  There isn’t much development for consortia.
• Iris version was released in May
  • People finding serious problems.
  • Missouri found that 006 field is not working correctly.  When you change the leader, it is automatically changing the 006.  The programmers didn’t realize that the 006 was supposed to be a different material type.
• Programing is happening for Juniper.  This will be a very quick release because Iris was delayed.  Juniper is due for release in August.
• UChicago is very bothered by the lack of communication between MARC record storage and Inventory.  The two apps don’t really communicate after records are created unless you force them to.  You can update Storage, and you can update Inventory, but those changes are not necessarily communicated to the other one.  Sounds like Chicago is freaking out.  I think they are planning to go live on Juniper.
• They are replacing the old search function with something called ElasticSearch.  It seems the old search function was not bringing back accurate or complete results.  
  • ES is working well to search the Inventory app, but it is not clear how well it can search MARC record storage.  That has to be tested.  
  • ES does not allow for searching more than one app at a time yet.
• Index Data is working with a closed group of FOLIO users to develop a Library Data Platform (LDP).  The LDP is where most reporting and statistics functions would exist.  Index Data’s LDP is open source.  EBSCO’s is not.  We will have to experiment with setting up an LDP server with the FOLIO test server we are working on.
• There is a complaint in the migration SIG that FOLIO changes significantly with each release.  This means the 3rd party migration apps they are building require significant change as well.  FOLIO is not supporting this.  This complaint is from an EBSCO employee.  EBSCO has a team of 7 people to keep up with all the changes.
• PALNI, GALILEO, and Stanford are all setting up test servers.  They hope to set up a test ReShare network between them.  If we get our test server up, we can join the ReShare test too.  
• Consortia SIG is discussing whether we need a central office app, or if central office people just need to be superusers on each tenant.  There are many functions people would like in a central office app.
• INN-Reach integration is meeting in two weeks to hopefully demonstrate actual software they hope to have ready to test in Iris.
• They do have serials checkin, but you can’t create patterns yet.  You can only manually create predictions.
• Missouri State is very unhappy about how the system keeps multiple copies of MARC records in MARC record storage.  Apparently, when you call on a MARC record with SQL you can’t be sure which one you are getting.
  • This may be fixed in an Iris hotfix.
• UChicago is worried they would not be able to get their data out again if they want to quit and move to another ILS.
• Global edit function is not on the road map.  It’s very highly rated by implementors, but it is also a huge job, so they don’t know when it can happen.
• Missouri and Chicago are worried about data corruption issues due to the poor interaction of Inventory and Record Storage, and other weird things they have found.  They are going to make a proposal that development is slowed down until this can be fixed.
• Lloyd says that this all seems bad, but he thinks it is still worth pursuing because the commercial market continues to become more and more monopolistic.  ProQuest, who owns III, just got bought by some other company.  It seems unlikely that Sierra will really continue to be supported adequately.  
• There are a lot of major universities joining the project.  CU Boulder is becoming an implementer.  They are hiring 2 developers to work on the project.

 Discussion Topics

• Currently, the field we use for a call number prefix, or pre-stamp, is subfield d.  This is the correct field if you are printing labels with the Sierra old printing system.  However, we just realized that the new print templates system gets this information from subfield f.  So if anyone who used prefixes were to switch to print templates they would have to change this field.
  • Update the cataloging standards document?
  • Group agrees this should be updated on Standards Document << ACTION ITEM
• Lloyd brings up the question of whether people prefer the current system where Lloyd types up minutes, or would people rather see a recording of the meetings?
  • People like getting the minutes, and they don’t express interest in getting recordings.
  • Lloyd will send people copies of the recording if they individually ask.

New Action Items

Ongoing Action Items

Next Duplicates Sub-committee meeting: June 9

Next UCC meeting: June 23