Neil Conway wrote:
>On Mon, Aug 04, 2003 at 10:47:36AM +0200, Andreas Pflug wrote:
>
>
>>AFAICS the current implementation still doesn't have a way to access the
>>affected rowset, so it'a pretty much like a SELECT without a WHERE.
>>
>>
>
>Yeah, unfortunately I didn't get a chance to implement this functionality
>during the 7.4 cycle (and it apparently wasn't important enough for
>anyone else to do it while I've been away). I'll probably get around
>to doing it some time before the 7.5 release.
>
>The current STATEMENT trigger code is still useful, IMHO -- I think
>your suggestion that we should remove any mention of this feature
>from the docs is certainly going too far. Should I submit a patch
>for the documentation to include a note about what functionality
>hasn't been implemented yet?
>
That's best, I believe. For people that came from other DBs, not
familiar with rules, they should be directed to the advantages of rules.
Places that mention triggers and need review:
- CREATE TRIGGER
- Chapter 36.6 Rules versus Triggers
- Chapter 37 "Server Programming - Triggers"
- Chapter 40 "PL/PGSQL"
- Chapter 41 "PL/Tcl"
(not ch. 42 - PL/Perl doesn't support triggers)
- Chapter 43 "PL/Python"
Phew! And this list is possibly not complete...
Regards,
Andreas