Received: from candle.pha.pa.us (maillist@s5-03.ppp.op.net [209.152.195.67]) by hub.org (8.9.3/8.9.3) with ESMTP id KAA76084 for ; Tue, 1 Jun 1999 10:31:25 -0400 (EDT) (envelope-from maillist@candle.pha.pa.us) Received: (from maillist@localhost) by candle.pha.pa.us (8.9.0/8.9.0) id KAA21629; Tue, 1 Jun 1999 10:30:55 -0400 (EDT) From: Bruce Momjian Message-Id: <199906011430.KAA21629@candle.pha.pa.us> Subject: Re: [SQL] Column name's length In-Reply-To: <199906010733.AAA00923@netcom12.netcom.com> from Zalman Stern at "Jun 1, 1999 00:33:30 am" To: Zalman Stern Date: Tue, 1 Jun 1999 10:30:55 -0400 (EDT) CC: tpham@mail.priority.net, pgsql-sql@postgreSQL.org X-Mailer: ELM [version 2.4ME+ PL56 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit > > And I assume it is working fine? > > Yes. It fixes the column length too I think. The following is a psql > interaction: > ----- > ssmldb=> create table > more_than_64_bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb > (more_than_32_bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb boolean ) ; > CREATE > ssmldb=> select * from > more_than_64_bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb > ; > more_than_32_bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb > --------------------------------------------- > (0 rows) > ssmldb=> insert into > more_than_64_bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb > ( more_than_32_bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb ) values ( true ) ; > INSERT 25353 1 > ssmldb=> select * from > more_than_64_bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb > ; > ^Mmore_than_32_bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb > --------------------------------------------- > t > (1 row) > ----- > (I did some tests to make sure all the characters are significant. They > are.) > > Here are the two diffs that up the "name size" from 32 characters to 256 > characters. (Once I get bit, I try to fix things real good so I don't get > bit again :-)) > ----- > diff postgresql-6.4.2/src/include/postgres_ext.h postgres-build/src/include/postgres_ext.h > 34c34 > < #define NAMEDATALEN 32 > --- > > #define NAMEDATALEN 256 > 37c37 > < #define OIDNAMELEN 36 > --- > > #define OIDNAMELEN 260 > ----- > diff postgresql-6.4.2/src/include/storage/buf_internals.h postgres-build/src/include/storage/buf_internals.h > 87c87 > < #define PADDED_SBUFDESC_SIZE 128 > --- > > #define PADDED_SBUFDESC_SIZE 1024 > ----- PADDED_SBUFDESC_SIZE is gone in 6.5beta, meaning we probably now have things tuned for NAMEDATALEN directly. > > After this, I did a clean build. I am pretty sure this produces a Postgres > that *is not* binary compatible with existing databases. You have to do the > dump/reload thing. (We discovered this issue very early on so I was able to > trash all my existing databases and start from scratch.) > > I do not know what performance or space usage impact this has. If it is not > too terrible, this would be a great change to see in a future version of > Postgres. Yes, I was thinking this too. -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026