Hi all, I'm in day 2 of a 4-day basic Lucene training seminar and conference. The instructor is Grant Ingersoll, one of the foremost Lucene/Solr experts around. I was describing to him some of the ideas that we've had regarding offloading fulltext indexing, maybe faceting, etc., to lucene/pylucene. He wondered why not just Solr. I tried to describe our concerns about sending that many ids or bibcodes over http. He asked why that was necessary. I tried to explain how we would need to rerank results according to metadata in Invenio. He thought that was odd and wondered why that additional ranking data couldn't also be indexed. That kind of stumped me. I know I've heard the arguments and reasons why it needs to be that way, but I'm not doing a good job of explaining the situation. Can someone try to summarize again. I'd like to take advantage of picking this guy's brain as much as possible.
Feel free to throw any other obscure lucene/solr questions at me as well, and I'll try to get answers. -- ****************************************************** Jay Luker Astrophysics Data System (ADS) jlu...@cfa.harvard.edu Center for Astrophysics 617-495-4588 60 Garden Street MS 67 617-495-7356 fax Cambridge, MA 02138 ******************************************************
