Re: serializable isolation level behavior question

On Oct 17, 6:07 pm, HansF <Fuzzy.Greybe..._at_gmail.com> wrote:
> On Tue, 17 Oct 2006 07:53:17 +0000, Laurenz Albe wrote:
>
> > But that is not standard compliant, is it?
>
>When did the truncate command become part of the SQL standard?

The issue at hand is not whether truncate is part of standard SQL. The issue is that if a standard SQL client is doing a serializable transaction*, and some other client does a truncate or anything else, standard or not, should the tx client expect oracle to either deliver on the specified isolation level guarantees or notify the tx client of a failure? Is it acceptable that Oracle allow a silent failure of the tx? As described, if a serializable tx gets different results for repeats of the same query, that is already a silent failure.

Joe Weinstein at BEA Systems

• (which does include a guarantee of repeatable reads)