Re: RoadMap for next Beagle releases


Daniel Naber wrote:
In a web context, Lucene is usually used to fetch the first 10 results (if the first page displays 10 results). If the users clicks "next page", the same search is re-executed, the first 10 hits are ignored and the next 10 are display. This is because the search itself is often very fast but actually getting the results is slow because it requires one disk access per document. Maybe this would be a useful approach for Beagle, too. Fetching 100 results sounds too much unless they are really immediately displayed.

The main reason why we fetch 100 results is that we display results by type and date rather than by relevance. 10 is too few to be really useful for most people, and more than 100 is too many. We could possibly make it 50 or 75, but the speed impact is pretty small at that point.


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]