Re: search.postgresql.org
От | John Hansen |
---|---|
Тема | Re: search.postgresql.org |
Дата | |
Msg-id | 1143269128.8226.3.camel@livecd.geeknet.com.au обсуждение исходный текст |
Ответ на | Re: search.postgresql.org ("Marc G. Fournier" <scrappy@postgresql.org>) |
Список | pgsql-www |
On Sat, 2006-03-25 at 02:24 -0400, Marc G. Fournier wrote: > On Sat, 25 Mar 2006, Tom Lane wrote: > > > "Joshua D. Drake" <jd@commandprompt.com> writes: > >> I was looking at our search configuration and I noted that out of a > >> weight of 100 we are only weighing archives at 50 and varlena at 25. > >> That seems a little low as archives is the primary source of pratical > >> and real world issues and varlena is a great resource. > > > > Um, what other resources are rated higher? If you'll pardon my > > ignorance, what other resources is the search engine considering at all? > > And maybe a short explanation of what these 'weights' are all about? Does > that just mean that when i search for something, archives results will > always preceed varlena's? And, if so, and we are only doing archives vs > varlena, as long as archives is weighted higher then varlena, does it > matter what the weights are? Could be 1 + 2, no? That's right, the siteweights are just for making sure results from one site is displayed before another. A search on search.postgresql.org/archives.search ONLY considers the archives anyways. search.postgresql.org/www.search, however, returns results from all the indexed sites, in order of relevance, with priority according to the siteweights. > > ---- > Marc G. Fournier Hub.Org Networking Services (http://www.hub.org) > Email: scrappy@hub.org Yahoo!: yscrappy ICQ: 7615664 > > ---------------------------(end of broadcast)--------------------------- > TIP 9: In versions below 8.0, the planner will ignore your desire to > choose an index scan if your joining column's datatypes do not > match
В списке pgsql-www по дате отправления: