From: Jeremy Evans <code@...>
Date: 2008-10-15T00:42:49+09:00
Subject: Re: ANN: Sequel 2.6.0 Released

Jos Backus wrote:
> On Sun, Oct 12, 2008 at 10:27:35AM +0900, Jeremy Evans wrote:
>> * Schema parsing was refactored, resulting in a huge speedup when
>>   using MySQL.  MySQL now uses the DESCRIBE statement instead of the
>>   INFORMATION_SCHEMA.  PostgreSQL now uses the pg_* system catalogs
>>   instead of the INFORMATION schema.
> 
> I'm curious: given the huge speed improvement using their native 
> metadata
> tables, does this mean that INFORMATION_SCHEMA support (which is the 
> standard
> way to access this type of information across database engines) is 
> poorly
> implemented in MySQL/PostgreSQL?  Or is INFORMATION_SCHEMA itself poorly
> designed?

In MySQL's case, it probably isn't implemented well (surprise, 
surprise).  I think the INFORMATION_SCHEMA is implemented fine on 
PostgreSQL, since there isn't a noticeable speed difference.  The reason 
we switched to the pg_* system catalogs on PostgreSQL was to get the 
primary key information, not for speed related reasons.  It's possible 
that the primary key information is available somewhere in the 
INFORMATION_SCHEMA, it was just easier to switch to the pg_* system 
catalogs than do the necessary research.

Jeremy
-- 
Posted via http://www.ruby-forum.com/.