As you may have already heard, IBM and Yahoo! today released a new product named IBM OmniFind Yahoo! Edition<http://omnifind.ibm.yahoo.net/productinfo.php>. It is a free-of-charge search engine for web sites and file systems, which builds on Lucene and other components such as UIMA <http://incubator.apache.org/uima/> and IBM's LanguageWare<http://www-306.ibm.com/software/globalization/topics/languageware/index.jsp>text analytics. Myself, I am Andreas Neumann, working for IBM, and one of the developers of IBM OmniFind Yahoo! Edition.
Having worked on various other search technologies before, I knew that a search engine, while appearing so simple to end-users, is a highly complex piece of software. The more was I thrilled by Lucene. Its flexible architecture and elegant interfaces make it a pleasure to work with. At the same time it delivers space-efficient, very low latency indexing and high-performance querying. Not to mention that it is pure Java, relieving us developers from the otherwise tedious task of porting our code to a variety of platforms and compilers. Lucene is one of the few pieces of software that prove that performance-critical software can indeed be written in Java. But the best thing about Lucene is its community. When we started working with Lucene less than a year ago, we were entering unknown territory. We had some concerns, were looking for a few features, missing some bug fixes... By now, all of the patches that we've been wishing for have either been committed, or they are under discussion now. This community really rocks! Andreas.
