On Tue, Oct 9, 2012 at 2:04 PM, Simon Riggs <simon@2ndquadrant.com> wrote:
> On 9 October 2012 21:35, Peter Eisentraut <peter_e@gmx.net> wrote:
>> On 10/9/12 5:09 AM, Simon Riggs wrote:
>>> Anyone want to check for any other missing IF EXISTS capability in other DDL?
>>
>> TRUNCATE is not really DDL. If we allow TRUNCATE IF EXISTS, what is
>> stopping someone from requesting DELETE IF EXISTS or INSERT IF EXISTS next?
>
> I'm not involved in the planning or justification for this patch, and
> have no opinion.
>
> I discussed applying it because it was an uncontentious patch. It
> clearly is not....
I also read Simon's approach as not a push for inclusion, but
defaulting to commit for smaller patches that basically look
mechanically legitimate with no objections to streamline
communication. Since pgsql-hackers has a good record objecting to
patches that require objection in a timely manner, I think that's
reasonable. The cost of revert would not be that high, either.
Clearly those conditions were not met, but I don't think it's
justified to jump on Simon for this approach on a patch like this.
--
fdr