On Wed, Aug 25, 2004 at 05:10:30PM -0400, Bruce Momjian wrote:
> Christopher Kings-Lynne wrote:
> > > OK, TODO updated:
> > >
> > > * Implement dirty reads or shared row locks and use them in RI triggers
> >
> > Can someone explain to me what a dirty read is and how it relates to RI
> > triggers?
>
> Dirty read allows you to see uncommited rows. I think RI triggers need
> it so they can know if someone has a current transaction that is going
> to conflict with the RI trigger action, or something like that. Right
> now I think we hang waiting for the transaction to complete.
Yes, we hang, but we behave differently depending on the commit status
of the transaction we are waiting for. We don't know that in advance,
which means we would have to "do something" at our own transaction end
to check that status.
Personally I think this is a bad mechanism to hide our lack of shared
row locks. We should be solving that problem instead, which I would
like to attack next.
FYI, I leave tomorrow morning to give a talk on Postgres at the
'Encuentro de Linux Norte' here in Chile, and will be back on monday.
--
Alvaro Herrera (<alvherre[a]dcc.uchile.cl>)
"La fuerza no está en los medios físicos
sino que reside en una voluntad indomable" (Gandhi)
