Re: [HACKERS] cidr - Mailing list pgsql-hackers

From Paul A Vixie
Subject Re: [HACKERS] cidr
Date
Msg-id 199807210510.WAA04382@bb.rc.vix.com
Whole thread Raw
In response to Re: [HACKERS] cidr  (Bruce Momjian <maillist@candle.pha.pa.us>)
Responses Re: [HACKERS] cidr
List pgsql-hackers
> This could clearly be a KILLER APP/TYPE for us.  This is a pretty
> sophisticated use of our type system.  Indexing should present no
> problems.  We supply the comparison routines and plug them in, and the
> optimizer automatically uses the indexes.

i'd like that to be true.  but the section of the manual which describes
this isn't as clear as the examples (the COMPLEX type in particular) in
the contrib/ directory at the time i started the work.  figuring out what
OID my comparison operators happened to get and plugging these values into
a PG_OPERATOR insert was just more than i could figure out how to automate.

other than the OID thing i really love the postgres type system, btw, and i
can't see why anybody would ever use MySQL (or Oracle) unless it has the
same feature.

pgsql-hackers by date:

Previous
From: Paul A Vixie
Date:
Subject: Re: [HACKERS] cidr
Next
From: Bruce Momjian
Date:
Subject: Re: [HACKERS] cidr