On 11/14/24 08:20, Richard Guo wrote:
On Wed, Nov 13, 2024 at 9:59 PM Andrei Lepikhov <lepi...@gmail.com> wrote:
Even So, it would be better to insert changes, induced by new
feature by one commit.
I'm not sure I understand what you mean by this sentence.
Yeah, it's my bad English. nothing special, I just wanted to say that
this code can be added later.
By the way, Do you know if anyone gave a speech on the current state of
internal plan parameterisation and its possible development directions?
It would be helpful to have such an explanation.
Not that I'm aware of, but I found the "Parameterized Paths" section
in optimizer/README to be very helpful.
Although parameterisation itself quite clearly described, the concept
and code, related to re-parameterisation (in presence of
inheritance/partitioning) seems complicated.
--
regards, Andrei Lepikhov
