Hi Evgeny
> we are causing the target solution to become staled
I think the community is not positive enough about xid64,There is also a lot of resistance, such as the wal log becomes larger
>I also hope you and the community will support attempts to retain the
> whole xid64 solution in the consistent state because only the consistent
> state allows us to continually test and fix the xid64 solution.
My point of view is the same as yours, I just kindly remind you that the community's point of view is to split into small patches, like logical replication ddl, which was even more popular, resulting in no news now
Thanks
> Hi Evgeny
> xid64 path split several threads ,The current one should be
> this:(https://www.postgresql.org/message-id/flat/CACG=ezaWg7_nt-8ey4aKv2w9LcuLthHknwCawmBgEeTnJrJTcw@mail.gmail.com)
> ,We can do some tests on path so that can merge earlier
> Thanks
Hello Wenhui!
Thank you for pointing out the specific topic developing the Multixact
ID Offset MXIDOFF64 solution. I agree to the fullest extent with
community's decision to split the xid64 into smaller parts and
subsequently merge them into the master. In the meantime, abandoning the
full version of the original patch and developing only its fragments, we
are causing the target solution to become staled. That is why I have
tried to put the whole solution together. The xid64 patch (v56-v58)
includes the MXIDOFF patch as a distinct files. The first four
patch-files of the xid64 are exactly the 9th version of the MXIDOFF64.
If new versions of the xid64 patch update the parts relevant to
MXIDOFF64, we will add them into MXIDOFF64 thread as a next version. I
also hope you and the community will support attempts to retain the
whole xid64 solution in the consistent state because only the consistent
state allows us to continually test and fix the xid64 solution.
Best regards,
Evgeny Voropaev,
Tantor Labs LLC.
