I'm skeptical about this as the wave of the future. The voice summary would be good while driving or otherwise visually engaged, but I can read a lot faster. A lot also depends on the quality of voice synthesis for unusual terms, I'd hate to hear a lot of words mispronounced. Visually, it's pretty but not exactly brilliant, as I don't need to see Google's logo yet again. The choices of images really matter, if I was looking up biochemistry, I'd hate to see photos of the universities where the labs are, rather than the chemical compounds. The main question I have is based on the demo: is there any way to navigate and refine the search? For the San Francisco one, I might want maps or tourist attractions rather than political entities; for the Chipmunks one, I might well want the animal rather than the musical group. The strict sequence of the demos makes me wonder. === All good points, Avi - I think it's a good start in alternative ways of presenting search results, more than simply a great search engine - yet, at least. The 'publication' model, where there are a number of related links, is interesting to me and has been requested by a number of folks I know.. Time will tell, i suppose! /Miles

Rappoport, with an O please! (and thanks for the attribution) HOy! FIXED:) My (lack of) typing come(s) back to haunt me badly! Sorry Avi!:)

I would love an ad system smart enough to know that I hate anything that blinks flickers or scrolls, that I'm a consumer who can be appealed to in terms of quality and enviro-friendlness, and that I skip ads 99:99% of the time and mute them if I can't skip (thankgoodness for tivo). Because that's not attractive to advertisers, I bet they never will cater to me. In return, I use AdBlockPlus and Click2Flash, and keep my sanity.

That's really interesting. I wonder if they are scoring ties, so they're showing some kind of default order within the scores. Can you get the detailed scores? And I so want to try this on the Open Relevance Project, http://lucene.apache.org/openrelevance/

Great write-up, I seem to have missed it before. The Meetup was heavily technical, I agree, but I think those are the people who saw the invitation, rather a completely representative group of Lucene/Solr people. And some things have to be discussed in technical terms, like TrieRanges and Query Parser Frameworks -- though I'd be happy to explain any/all of them, if someone wants to pay me to do it! Thanks Avi! It was good to see you there as well! Agree about Autonomy: "Mike Lynch's secret algorithms are not why people buy IDOL anymore. Things like giant reference accounts, professional services, and commercial grade spiders have a lot more to with why big companies still pay six figures for search technology." In fact, they haven't really been interested in search for about a decade, which is kinda sad. I hope the spider reference is to Ultraseek, because all those years of care and tweaking deserve to live on. Avi Rappoport (please note spelling?)