This article is more than 5 years old.

Before commenting on today’s topics, I thought I would say few things about the pre-conference section on “Solr: what’s new?” that I attended yesterday.

Although there were other interesting pre-conference sections offered concurrently, I chose to attend the one on Solr because of the important role that its SolrMarc utility plays in vufind record indexing. If you wonder how this works, basically, SolrMarc reads in records from an imported voyager marc load file, extracts information from various fields and adds that information to the vufind Solr index. I wanted to learn more about Solr and see if there is any new update that can be used to move our vufind implementation to the next level. The preconference was rather technical. Erik Hatcher from Lucid Imagination was the presenter and he talked about how Solr has drastically continued improving over time. Some of the new features in development include SolrCloud which relies on Zookeeper, a centralized service for maintaining configuration information, to provide a shared, centralized configuration and core management to programmers. He also talked about pivot/grid/matrix/tree faceting which is a hierarchical way of providing facets that would branch out to other facets (“sub-facets”) to further narrow down a search. Another cool feature that Solr has improved is the date faceting and that is going to be seen in our upcoming vufind upgrade.

The actual conference started today and Erik has already blogged about all the important subjects. I was interested in what Damian had to say about vufind.

The idea about centralization of code introduced by Erik Hatcher was also embraced by Damian Katz when he talked about the redesign goals for vufind. He is aiming for a centralization of marc specific code to facilitate replacement. Just to be a little bit controversial, Damian stated that “MARC must die”. He wanted to say that library data is not limited to marc data but also consists of other data types that are becoming more and more popular. He expressed pride in the ability for the upcoming release of vufind (vufind 1.1) to provide among others, full Open Archive Initiative Protocol for Metadata Harvesting (OAI-PMH) server support capability, which will enable harvesting metadata into a directory for further data manipulation. Here, Damian provided a solution to the question “where is my data?” . His solution, “grow the toolkit” that will solve the problem of obtaining records from remote sources, process harvested files, and index arbitrary xml records. According to Damian, understanding record drivers gives the programmer a lot of control over vufind.