Camelia, the Perl 6 bug

IRC log for #ilbot, 2013-06-17

| Channels | #ilbot index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
20:18 * moritz is making some progress on the 'normalized' branch
20:19 pdurbin cool
20:19 moritz nearly everything is ported, except for the search
20:19 moritz and it's way faster
20:20 moritz I now have a table which stores combinations of days and channels
20:20 moritz and I can cache number of lines per day there
20:20 moritz *the number
20:20 moritz which makes the #perl6 index page much, much faster :-)
20:20 pdurbin :)
20:27 moritz but search is going to be tough, because I'm now using InnoDB as a backend, which doesn't support fulltext indexes
20:27 moritz so it'll probably Lucy right away
20:27 moritz *be
20:29 pdurbin as I've mentioned, I've been working with Lucene. would be happy to help if I can
20:33 moritz I'll surely ask when I need help
20:34 moritz ah, I have a question right away
20:34 moritz one can only search in one channel at a time
20:34 moritz would it make sense to build a separate search index per channel then?
20:42 moritz (I think so, because I don't think search engines do so well in partioning the output data based on exact matches, at least compared to RDBMS)
20:59 pdurbin moritz: I know you're not excited about Solr or ElasticSearch but you might want to try one of them as you prototype the solution. they have nice RESTful interfaces for adding and querying documents
20:59 pdurbin once you setting on the way you want to do it, you could switch to Lucy

| Channels | #ilbot index | Today | | Search | Google Search | Plain-Text | summary