Alvaro Herrera <alvherre@commandprompt.com> writes:
> Excerpts from Robert Haas's message of mar jun 07 08:16:01 -0400 2011:
>> Probably. I guess the question is whether we want this to fail in (a)
>> the parser, (b) the planner, or (c) the executor.
> Really? I thought it was the job of the parse analysis phase to figure
> out if table and column names are valid or not, and such. If that's the
> case, wouldn't it make sense to disallow usage of a table that doesn't
> "exist" in a certain sense?
If you hope ever to support the proposed UNLOGGED-to-LOGGED or vice
versa table state changes, you don't want to be testing this in the
parser. It has to be done at plan or execute time.
regards, tom lane
