Re: BUG #11732: Non-serializable outcomes under serializable isolation

Kevin Grittner <kgrittn@ymail.com> wrote:

> Either way, what Tin
> is seeing should not be possible, because what I just described
> should have caused a dangerous structure for which should have
> caused either Tpivot to be canceled.

Gah!   Something shiny must have come into my field of vision while
I was editing that sentence.   My apologies.   This stuff is hard
enough to understand without such mangled editing, so I'll try
again:

Either way, what Tin is seeing should not be possible, because what
I just described should have caused a dangerous structure which
should have caused either Tpivot or Tout to be canceled.

--
Kevin Grittner
EDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company