News for hackers. Stuff that matters.
Re:Hstore development for 9.4 release by bramanga on Thursday November 07, 2013 @12:40PM (#6690845)
IMHO this seems like building a better apple by making it taste like an orange. This could definitely be useful for storing json from javascript frameworks like backbone/angular/ember, but I don't see how this is going to improve the lives of PostgreSQL users in general. Hopefully this leads to some weird cool emergent use case that make PostgreSQL way better than other databases, but I'm not sure at this point that it will. Does anybody have more context on the long-game for this approach? I'm not sure what the goal is for adding this feature.
Re:Hstore development for 9.4 release by garenp on Thursday November 07, 2013 @02:38PM (#6691876)
Look at some of the linked PDFs. It's the indexing technique and on-disk format that is an enabler for several notable use cases; personally, I'm looking forward to a time when I can use pg for full-text search without having to resort to another external system like Lucene or Sphinx. What it won't have is the query "smarts" like Lucene or Sphinx do today, but you need speed first.
Re:Hstore development for 9.4 release by baudehlo on Thursday November 07, 2013 @06:50PM (#6693400)
Pg's built in full text search works amazingly well and has all the smarts. It's also faster than lucene.
Re:Hstore development for 9.4 release by rpedela on Friday November 08, 2013 @12:30AM (#6694621)
Huh? Please come up with about ten references (one isn't enough) to back up those claims.
Re:Hstore development for 9.4 release by jeltz on Sunday November 10, 2013 @12:34PM (#6707438)
I do not think so, but it might be after the GIN optimization work is committed. After it they beat Sphinx in several benchmarks.

http://www.sai.msu.su/~megera/postgres/talks/Next%20generati...

Re:Hstore development for 9.4 release by baudehlo on Sunday November 10, 2013 @06:04PM (#6708884)
I didn't say it was faster than sphinx (yet) but it is definitely faster than lucene in all the benchmarks I've thrown at it.
Re:Hstore development for 9.4 release by mml on Thursday November 07, 2013 @04:38PM (#6692591)
I'm very excited to put my 3m documents into hstore, as they are currently in a horrible, unqueryable, unwritable (mostly) blobs of compressed, b64encoded serialized ruby hash structures (long story, predating nosql).

This makes me very, very happy.

Edit:

I should point out why: -my document structures pretty much match 1:1 with hstore now (we have nested arrays of hashes) -i can query & report on the data with normal tools -i can update individual fields on my data (currently impossible) -i can do relational joins & such-like in the db (currently impossible) -i can have sensible transactions