On 06.05.2021 17:28, David G. Johnston wrote:
On Thu, May 6, 2021 at 6:44 AM Tom Lane <tgl@xxxxxxxxxxxxx> wrote:
This case is the reason we invented the "stable" attribute to begin
with. People have since misinterpreted it as authorizing caching of
function results, but that's not what it was intended for.
This is a good paragraph...if something like it gets added to the create function documentation mis-interpretations are likely to decrease.
I found additional details in the documentation. In particular about the index scanning for stable functions:
https://www.postgresql.org/docs/13/xfunc-volatility.html
The link to this section there is in the create function page. Maybe that's enough.
-- Pavel Luzanov Postgres Professional: https://postgrespro.com The Russian Postgres Company
