anonymous wrote : You could do something like use meta attributes that hibernate tools support to annotate which things you want to be searched and users can then either use reveng.xml to indicate which fields needs indexing or use a programmatic strategy to compute it. This is seems be better. I have a smattering knowledge of Hibernate Search. anonymous wrote : So if the generation is pretty simplistic (i.e. nothing fancy) then it could be integrated into reverse engineering/code generation but not sure if its really useful/will be used. I think it could be usefull. anonymous wrote : What other good index search tools are there and why should we generalize them ? Hibernate only support many db's because the db's has a common language (SQL, tables, rows, columns,..). Index search tools doesn't afaik...and again - why not just focus on Lucene ? This is could be a topic for a long discussion... In a general case - index search tool consist of index (analog of DB) for some documents/items, and query in a common case is just a string (Lucene generalize this, but in common case this is just a string) - no SQL just a string in a common case (of cause it is possible to use some preconditioning). What does mean a good index tool? There are many parameters here as portability, index speed, search speed, maximum number of documents in index, relevancy of results, simplicity to configure, etc. I have to test a lot of index tools cause this is intresting question for me. Lucene is a most famous, it has a god infrastructure and many projects use it, but it is not the best for all parameters which I describe above. I will not admit some other index search tool - cause this is not a main topic here, but I admit that there are others, and possibility to select is a greate posibility for the user, for developer. "focus on Lucene?" why not :) View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4203377#... Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&a...