On Jan7, 2014, at 09:45 , Heikki Linnakangas <hlinnakangas@vmware.com> wrote:
> Overall, I'm leaning towards biting the bullet and always detoasting everything in master. Probably best to just
leavethe stable branches alone.
+1
The fact that de-TOAST-ing can happen lazily is, at least to me, an
implementation detail that shouldn't be observable. If we want to
allow people to use lazy de-TOAST-ing as an optimization tool, we
should provide an explicit way to do so, e.g. by flagging variables
in pl/pgsql as REFERENCE or something like that.
best regards,
Florian Pflug