On 15.05.24 08:13, Michael Paquier wrote:
> On Tue, May 14, 2024 at 01:33:46PM +0800, jian he wrote:
>> thanks for the idea, I roughly played around with it, seems doable.
>> but the timing seems not good, reverting is a good idea.
>
> Please note that this is still an open item, and that time is running
> short until beta1. A revert seems to be the consensus reached, so,
> Peter, are you planning to do so?
I'm on it.
Here is the list of patches I have identified to revert:
git show --oneline --no-patch 144c2ce0cc7 c3db1f30cba 482e108cd38
34768ee3616 5577a71fb0c a88c800deb6 030e10ff1a3 86232a49a43 46a0cd4cefb
6db4598fcb8
144c2ce0cc7 Fix ON CONFLICT DO NOTHING/UPDATE for temporal indexes
c3db1f30cba doc: clarify PERIOD and WITHOUT OVERLAPS in CREATE TABLE
482e108cd38 Add test for REPLICA IDENTITY with a temporal key
34768ee3616 Add temporal FOREIGN KEY contraints
5577a71fb0c Use half-open interval notation in without_overlaps tests
a88c800deb6 Use daterange and YMD in without_overlaps tests instead of
tsrange.
030e10ff1a3 Rename pg_constraint.conwithoutoverlaps to conperiod
86232a49a43 Fix comment on gist_stratnum_btree
46a0cd4cefb Add temporal PRIMARY KEY and UNIQUE constraints
6db4598fcb8 Add stratnum GiST support function
Attached are the individual revert patches. I'm supplying these here
mainly so that future efforts can use those instead of the original
patches, since that would have to redo all the conflict resolution and
also miss various typo fixes etc. that were applied in the meantime. I
will commit this as one squashed patch.
