Meeting Minutes for 12/18/2019

Digital Archive Committee Minutes
December 18, 2018
1:00 PM

Demo and discussion about archive outage (Lead by Pascal)

  • Last week there was a mystery as to what happen to the archive.
  • On Monday,  Marmot received a report that a compound object or collection was not displaying.
  • On Tuesday, Tammy looked into the issue and discovered that collections were not displaying in Islandora.
  • Marmot also discovered that compound objects (an object that has two constituent parts) would not display parts in Islandora or Pika.
  • On Tuesday, Marmot restored backups from various parts from the previous week thinking some sort of setting in Islandora had been changed to cause this disastrous effect. 
  • Restoring the backups had not effect.
  •  It turned out to be the Islandora index for Fedora data, that stores all the metadata in what is called a triplestore, was missing two kinds of data.  
  • There is more than one index in Islandora.
    • There is a Solar index for searching.
    • There is a Fedora index which relates metadata to different things.
  • The Fedora index was missing two types of metadata.
    • It was missing sequence number for the member of a collection. Islandora could not display objects without a sequence number.
    • For compound objects there is another metadata type called “isConstituentOf” that relates the two pieces of a compound object together.
  • Marmot talked with the Islandora foundation, and their point of contact pointed us in the right direction.
  • On Thursday, Mamot had to rebuild this index, which restored the data that was missing in Islandora and for compound objects.
  • Question: Can this  happen again?
  • Answer: It may happen on an infrequent basis, but this is the first time it has happened for our instance. Recovery from it is easy now that they know what caused the problem.

Discussion about adding born digital content (Lead by Alysa)

  • Bud Werner partnered with all of the organizations in Routt County to build a community portal called Yampavalley.info.
  • This portal was created before any nonprofit organization could build a website, or do anything easily online like you can today.
  • Yampavalley brought in 75 nonprofit organizations to build their website within the framework, and a number of historians put their content about local history on this portal as well.
  • When Bud Werner took the Yampavalley.info site down, they made sure that all the organizations had migrated to a new website.
  • There was still a lot of born digital content that does not exist anywhere except on Yampavalley.info, which is no longer on the web.
  • Alysa would like to discuss how they would take born digital objects and pull it into Pika for the Digital Archive.
  • Alysa wondered how they could take content from Yampavalley.info and incorporate into the Digital Archive everyone is building with Pika, and eventually get it on Digital Public Library of America. Nothing is scanned, there are no paper documents, and there is nothing to upload. It is just text with an occasional picture.
  • Alysa would also like the Routt County cemeteries or gravesites migrated as well. Each cemetery link may have a photograph, GPS coordinates, a description to the history of the area, and it can be tied to the genealogy database with all the people buried in those locations.
  • Alysa would like an open discussion and get some feedback from Pascal and Ashley about any insights on how they might take on something like this and get into the archive.
  • Pascal will have to do some research to find the best way to do this type of project.
  • Brandon asked is the content could stay where it is currently located, and if they just needed to link to it, or could it all just be manually pulled into the archives as person or place entities.
  • Alysa replied that the site is no longer on the web, and is no longer accessible.  This does lend itself to flushing out an entity.  However, it is authored by someone who would need to be credited for that work. 
  • Alysa can put these into PDFs and upload them to create a fake book, or fake compound object within Islandora. However, since we are not sending our entities to the Digital Public Library of America they would lose that ability for that object to be shared at DPLA. Alysa would like these to be in the Digital Public Library of America as well.
  • Brandon asked Aylsa if she wanted to make the people entities more like objects.
  • Alysa and John talked about turning this into a book and each page would be sheriff or cemetery, and tie those to the appropriate entities so they are findable. However, the only way to do this would be with scanned documents or PDF.
  • Brandon replied that Islandora does have a solution pack called Web Archives, but he is not sure if it is installed on our instance of Islandora.
  • Brandon is not sure what can be included in the Web Archives solution pack. 
  • To Brandon, Web Archives sound like the original source of content that would be a book or web archive. Being able to identify what the source of the content was, and provide acknowledgement to the original authors (along the lines of the initial contributor and secondary contributor) makes sense.
  • Pascal would like to create a new type of object, because it is a web page with html in it. He thinks they could preserve it as the master data.  With html they should to be able to display it in a digital archive. He would like to do something different than what currently exists.  He thinks it would basically be an iframe.
  • Alysa likes the idea of treating  it as web archive to credit the Yampavalley.info, which was an enormous resource in their community for a period of time.  This could allow them to create entities for that effort, as well as preserving Joyce and Roger’s incredible history efforts on behalf of their community within that resource.
  • Alysa asked what would be the next steps? Would Marmot have to reachout to our hosting solution for Islandora to see about adding another object type? 
  • Marmot would have to go through the process to add the Web Archive Solution Pack.  They would need to tie the forms to the solution pack. It will take a little work to get it all set up.  
  • R&D will do some research to see how long it will take to step this up in a planned way in case they need to rollback of Islandora.  They will install it on the test server first. There would also be a level of Pika development as well, since this would be a new type of thing to display.  
  • Alysa mentioned that this would be the next project for Bud Werner.  She wondered about the timeline for Marmot getting this to work, or wondered if Marmot needed to do more research first.
  • Marmot would definitely need to figure out how much time this would take, before making a commitment.  
  • Brandon suggested that if adding the solution pack is going to be quick that they move forward with that part of the project, so they can get the content displaying in Islandora. The Pika display could be a project that is taken on as it fits into the R&D schedule.
  • Alysa agreed with Brandon’s suggestion.  The Yampavalley.info project is part of the original Three Wire Winter plan, and Bud Werner needs to get this information added as well.
  • Brandon will reach out to the group to let them know how this process will fit within the Marmot schedule, and how long it will take to get everything in place.
  • JB and the R&D team will add this to their future projects, so they can plan for it.

Next meeting is Tuesday, January 15th at 1:00 p.m.

Meeting Date: 
Tuesday, 2018, December 18
Documentation Type: 
Meeting Minutes
Committees: 
Digital Archive Committee