give an example of the failed searches

diff --git a/doc/todo/lucene_search_engine.mdwn b/doc/todo/lucene_search_engine.mdwn
index 694c81c4cba316ef8f2cf6f20021b39e82a1a71c..bac9f91307001bbd2200096b1171ee8e151c176d 100644 (file)
--- a/doc/todo/lucene_search_engine.mdwn
+++ b/doc/todo/lucene_search_engine.mdwn
@@ -1 +1 @@
-There are [some issue](http://www.branchable.com/bugs/Exception:_Cannot_open_tables_at_consistent_revisions_at___47__usr__47__lib__47__perl5__47__Search__47__Xapian__47__WritableDatabase.pm_line_41./#comment-c159ea3f9be35fcd9ed0eeedb162e816) with the current search engine. Sometimes the database gets corrupted and it's not very good at weighting say, the title against the content. In [[different_search_engine]], there was the idea of using Lucene - is there any reason why we should have both, or at least let lucene live in contrib?
+There are [some issue](http://www.branchable.com/bugs/Exception:_Cannot_open_tables_at_consistent_revisions_at___47__usr__47__lib__47__perl5__47__Search__47__Xapian__47__WritableDatabase.pm_line_41./#comment-c159ea3f9be35fcd9ed0eeedb162e816) with the current search engine. Sometimes the database gets corrupted and it's not very good at weighting say, the title against the content. For example, [searching for pagespec](http://ikiwiki.info/ikiwiki.cgi?P=pagespec) in this wiki doesn't lead to the [[ikiwiki/pagespec]] page in the first page... but in the third page. In [[different_search_engine]], there was the idea of using Lucene - is there any reason why we should have both, or at least let lucene live in contrib?