On 2/5/23 3:01 PM, Tom Lane wrote:
> "Jonathan S. Katz" <jkatz@postgresql.org> writes:
>> On Feb 4, 2023, at 10:24 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>>> “Prevent clobbering of cached parsetrees…Bad things could happen if…”
>
>> While I chuckled over the phrasing, I’m left to wonder what the “bad things” are, in case I
>> need to check an older version to see if I’m susceptible to this bug.
>
> Fair. I was trying to avoid committing to specific consequences.
> The assertion failure seen in the original report (#17702) wouldn't
> occur for typical users, but they might see crashes or "unexpected node
> type" failures. Maybe we can say that instead.
I did a quick readthrough of #17702. Your proposal sounds reasonable.
Based on that explanation and reading #17702, I'm still not sure if this
will make the cut in the release announcement itself, but +1 for
modifying it in the release notes.
Thanks,
Jonathan