I was in SF, California last week to dig deeper into Solr and to meet with its core developers (now working for Lucid Imagination). It was a great week with many new connections and new insight.
First of all, as you may know, Lucid Imagination is at the moment the primary commercial marketers, promotors and evangelists for Solr. They bring to the scene what an open source offering often lacks – a professional and polished image, nice wrapping and not least – commercial support. That is great for the future adoption of Solr with customers that need just that kind of safety. Being Lucid’s Norwegian parner, Cominvent AS now offers all of this in Norway as well.
In the upcoming version 1.4 of Solr, there are a lot of good stuff. Let me ierate some of them:
- Better built-in support for replication to multi node
- Improved performance of facets (navigators)
- Greatly improved performance of numeric sort and range queries
- A number of bug fixes
- Import of Word, Excel, PPT, PDF +++ formats through Tikka
This promises good for the enterprise use case, which both needs scalability and multiple document format support. The new index replication feature is implemented in Java and is socket based, promising to work on all platforms. The old implementation was script+rsync based and would only work on *NIX systems.
Also, being able to work with MS Office document formats is important. There might be commercial document conversion filters which support more formats and at a better quality, but having a free open alternative by default and an commercial upgrade path is all about choice and flexibility, isn’t it?
Cominvent AS, an independent search consulting company, is partnering with Lucid Imagination, offering professional support and training to the Norwegian market. We also offer consulting- and training services for FAST ESP.