Bruce Momjian <pgman@candle.pha.pa.us> writes:
> 1 - All SETs are rolled back in aborted transaction
> 2 - SETs are ignored after transaction abort
> 3 - All SETs are honored in aborted transaction
> ? - Have SETs vary in behavior depending on variable
My vote is 1 - roll back all SETs.
I'd be willing to consider making the behavior variable-specific
if anyone can identify particular variables that need to behave
differently. But overall I think it's better that the behavior
be consistent --- so you'll need a good argument to convince me
that anything should behave differently ;-).
There is a variant case that should also have been illustrated:
what if there is no error, but the user does ROLLBACK instead of
COMMIT? The particular case that is causing difficulty for me is
begin;create schema foo;set search_path = foo;rollback;
There is *no* alternative here but to roll back the search_path
setting. Therefore, the only alternatives that actually count
are 1 and ? --- if you don't like 1 then you are voting for
variable-specific behavior, because search_path is going to behave
this way whether you like it or not.
regards, tom lane