News for hackers. Stuff that matters.
Re:Why HStore2/jsonb is the most important patch of PostgreS... by jmnicolas on Sunday February 23, 2014 @06:38AM (#7285627)
PostgreSQL is the best open source relational database and I hope it remains that way.

While I don't mind seeing some non relational additions like JSON, it shouldn't be detrimental to Postgres relational core.

It's not a pissing contest : if MongoDB is more popular than Postgres for non relational (NoSql) data so what ?

It's not like Postgres is a commercial product and need wider adoption to keep the parent company afloat.

Re:Why HStore2/jsonb is the most important patch of PostgreS... by masklinn on Sunday February 23, 2014 @07:04AM (#7285674)
> While I don't mind seeing some non relational additions like JSON, it shouldn't be detrimental to Postgres relational core.

The patchset in question is an improvement of hstore[0] (typed hstore, nested hstore) and unification of json and hstore (by using the binary hstore format for both). There is no more detriment to the relational core than when hstore or XML were originally added. Also, next-GIN[1] by the same people[2], not sure it's part of the patch mentioned.

[0] [PDF warning] http://www.sai.msu.su/~megera/postgres/talks/hstore-dublin-2...

[1] [PDF warning] http://www.sai.msu.su/~megera/postgres/talks/Next%20generati...

[2] Who previously contributed GiST and GIN, FTS, ltree or intarray

Re:Why HStore2/jsonb is the most important patch of PostgreS... by jeltz on Sunday February 23, 2014 @09:26AM (#7285964)
The GIN improvements were submitted as a set of separate patches and several of them have been merged into 9.4.
Re:Why HStore2/jsonb is the most important patch of PostgreS... by masklinn on Sunday February 23, 2014 @11:35AM (#7286407)
Good to know, thanks a lot for the info.
Re:Why HStore2/jsonb is the most important patch of PostgreS... by jeltz on Sunday February 23, 2014 @12:45PM (#7286674)
I know that the GIN compression patch has been committed and some of the patch for speeding up searches by skipping posting list entires, but I have not seen anything about the third part, speeding up FTS by adding ordering info to the indexes.

https://commitfest.postgresql.org/action/patch_view?id=1357 https://commitfest.postgresql.org/action/patch_view?id=1382

Re:Why HStore2/jsonb is the most important patch of PostgreS... by twic on Sunday February 23, 2014 @10:53AM (#7286211)
They've still got a .su address! Old school!
Re:Why HStore2/jsonb is the most important patch of PostgreS... by einhverfr on Sunday February 23, 2014 @09:47AM (#7286018)
> PostgreSQL is the best open source relational database and I hope it remains that way.

Yes, and it will. In fact the best post-relational features of PostgreSQL work best when used in a relational way.

> While I don't mind seeing some non relational additions like JSON, it shouldn't be detrimental to Postgres relational core.

As I usually tell people "Good object-relational design is usually good object-oriented design and good relational design, and that makes it really, really hard."

> It's not a pissing contest : if MongoDB is more popular than Postgres for non relational (NoSql) data so what ?

I would agree. The fact is that for the last 10 years, PostgreSQL has been the choice of open source rdbms's in business apps for the last 12 years and I don't see that changing.

> It's not like Postgres is a commercial product and need wider adoption to keep the parent company afloat.

Mindshare is important though.

Re:Why HStore2/jsonb is the most important patch of PostgreS... by jeltz on Sunday February 23, 2014 @09:31AM (#7285976)
Adding to what masklinn said, JSON is just another type with related functions and operators. It uses the already implemented extensible type system, and is no different from any other type including varchar and integer.