...
- As documented in this mailing list thread, UF's problem with re-indexing it's data in 1.5.1 hasn't been resolved. Possible root cause is bad data in MySQL evidenced by repeated errors in logs, but the logs just point to what should be allowed escape characters (linefeeds) that should not cause Java, Jena or Solr to throw exceptions. Should this be the topic of dedicated development meeting? Who wants to be involved and what time slot would work – does the 2 pm ET Thursday slot work for people?
- Chin Hua also needs to have this bug resolved to be able to debug the visualization issues Florida and Penn have been experiencing
- Problems with non-Roman characters – make sure your server.xml contains URIEncoding="UTF-8" in the active <Connector> element. more
- Transport error in connecting to Solr (resolved – was an outdated configuration)
- Updating Solr to be able to use new modules/add-ons
- just dropped in the new Solr WAR file and changed the build.config to point to the new version
- everything worked in VIVO
- when pushed the update to the Colorado Springs campus, a data update failed so had to wipe the webapps directory for VIVO and Solr, not just the Solr directory – was just giving an internal system error during the add process
- fixed the issue of searching with wildcards and will be other new features to take advantage of – there's a changelog for Solr online
- Connecting VIVO to a REST-style web service to migrate existing profiles to VIVO
- Generating a list of keywords from VIVO
- Last data update date for the entire site at CU Boulder (are not allowing individual editing, and update the entire site every other day)
- might in the future want to have a tooltip that would show where more granular data elements in different graphs or even on a per individual basis
- is this a feature others would like – a more robust version of this might change a number of files
- Jim – others might or might not need that particular feature, but would love to see something more generalized to make arbitrary information available to Freemarker templates so can put that information on any page, or in the footer
- we already have an investment in our data-getter class that allows you to configure it with a sparql query and the name of the Freemarker variable you would like to put the returned value(s) into
- just need to figure out how to have a query run to appear on any page, and how to set up one or more data getters to run every time any page is generated, as opposed to running queries within specific page templates that the controller invokes just for that template
- is this as generalizable as a key-value pair
- are using different named graphs as of 1.5.1, so can replace one data element and can diff the DAT files and only re-run ingest processes where there have been changes. Has enabled reductions of the ingest load from 3-4 hours to about an hour
- can add statements to the graph for provenance on batch updates of the entire graph
- might in the future want to have a tooltip that would show where more granular data elements in different graphs or even on a per individual basis
- Other issues?
...