On 2013-12-01 12:45:14 +0100, Andres Freund wrote:
> b) the old xmax cannot be an update, we wouldn't see the row version in that
> case. And in any way, ISUPDATE_from_mxstatus() isn't sufficient to
> determine whether the old row was an update, one needs to look at
> LOCK_ONLY as well, no?
That part's bogus, I missed part of the branch above the quoted code.
Greetings,
Andres Freund
--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services