> > Here is an overview of the SITC method:
> > http://momjian.us/cgi-bin/pgsitc
>
> A pretty fundamental problem is that the method assumes it's
> OK to change the CTID of a live tuple (by swapping its item
> pointer with some expired version). It is not --- this will break:
I am having difficulty visualizing that. The plan is not to change
CTID's
(only the CTID's offset into the page is to be changed).
The CTID of the new version is one that is up to now invisible to all
backends,
so noone can actually have remembered that CTID.
Also you would first insert the slot content and then change the CTID
offset
(this offset change might need to be made atomic).
Andreas