| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Jim Nasby <jnasby(at)upgrade(dot)com> |
| Cc: | "Burd, Greg" <greg(at)burd(dot)me>, Nikita Malakhov <hukutoc(at)gmail(dot)com>, Hannu Krosing <hannuk(at)google(dot)com>, Postgres hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Support for 8-byte TOAST values (aka the TOAST infinite loop problem) |
| Date: | 2025-08-06 23:16:28 |
| Message-ID: | aJPiTG0leC0izFte@paquier.xyz |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Mon, Aug 04, 2025 at 02:37:19PM -0500, Jim Nasby wrote:
> On Fri, Aug 1, 2025 at 4:03 AM Michael Paquier <michael(at)paquier(dot)xyz> wrote:
>> - Addition of separate patch to rename varatt_external to
>> varatt_external_oid and VARTAG_ONDISK to VARTAG_ONDISK_OID, in 0003.
>
> Since you're already renaming things... ISTM "ondisk" has the potential for
> confusion, assuming that at some point we'll have the ability to store
> large datums directly in the filesystem (instead of breaking into chunks to
> live in a relation). VARTAG_DURABLE might be a better option.
Hmm. I don't know about this one. Durable is an ACID property that
does not apply to all relation kinds. For example, take an unlogged
table: its data is not durable but its TOAST pointers would be marked
with a VARTAG_DURABLE. With that in mind ONDISK still sounds kind of
OK for me to use as a name.
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Michael Paquier | 2025-08-06 23:18:26 | Re: BF mamba failure |
| Previous Message | David Steele | 2025-08-06 22:30:50 | Re: Return pg_control from pg_backup_stop(). |