Re: [HACKERS] Re: ALTER TABLE DROP COLUMN

At 01:22 PM 2/29/00 +0100, Jan Wieck wrote:
>I wrote:
>
>>     fully  SQL3.  I just had an idea that might easily turn it to
>>     do the right thing.
>
>    ON  <event>  RESTRICT  triggers  are  now  executed after the
>    statement allways, ignoring any explicitly set deferred mode.
>    This  is  pretty close to Date's SQL3 interpretation, or IMHO
>    better. Date says that they are checked BEFORE each ROW,  but
>    that  would  ignore  the SET character of a statement.

Perhaps that's actually the point of RESTRICT?  Sacrifice the 
set character of a statement in this special case in order to
return an error quickly?

Since RESTRICT wasn't in SQL92, and since it's very close to
NO ACTION, it reeks of being an efficiency hack.  

I dread digging into that part of the standard again...this is
a case where the various proposals and justifications that were
before the committee at the time would be useful since the 
actual words that made it to the standard are opaque.



- Don Baccus, Portland OR <dhogaza@pacifier.com>
  Nature photos, on-line guides, Pacific Northwest
  Rare Bird Alert Service and other goodies at
  http://donb.photo.net.