Re: when does CREATE VIEW not create a view? - Mailing list pgsql-hackers

From Ross J. Reedstrom
Subject Re: when does CREATE VIEW not create a view?
Date
Msg-id 20000822151601.B27521@rice.edu
Whole thread Raw
In response to when does CREATE VIEW not create a view?  (Brook Milligan <brook@biology.nmsu.edu>)
List pgsql-hackers
Brook, 
This smells like a identifier length limit problem to me. Let's see:

precipitation_english_verify is 29 characters, default NAMEDATALEN is
32. Creating a view creates a table, and attaches a SELCT DO INSTEAD
rule to it, named _RET<tablename>, so that tacks 4 characters on, giving
us 29+4 = 33, bingo, rule doesn't get made. All your other attemps were
longer, except for xxx.  You'll find that replacing english with xxxxxxx
won't work, either (and it's not the vchip).

Sounds like a missing error check, or truncation, in the CREATE VIEW
rule generation code.

Ross

On Tue, Aug 22, 2000 at 12:40:21PM -0600, Brook Milligan wrote:
> I am trying to create a view and have run across a, to me, bizarre
> occurance.  One CREATE VIEW statement creates the view fine; changing
> the name of the view and repeating the same statement does not.  This
> has nothing to do with conflicting names as appropriate DROP commands
> are issued first.
> 

-- 
Ross J. Reedstrom, Ph.D., <reedstrm@rice.edu> 
NSBRI Research Scientist/Programmer
Computer and Information Technology Institute
Rice University, 6100 S. Main St.,  Houston, TX 77005


pgsql-hackers by date:

Previous
From: Brook Milligan
Date:
Subject: Re: when does CREATE VIEW not create a view?
Next
From: "Ross J. Reedstrom"
Date:
Subject: Re: when does CREATE VIEW not create a view?