On 04.03.23 17:35, Jeff Davis wrote:
> On Thu, 2023-03-02 at 09:13 -0500, Dave Cramer wrote:
>> I'd like to open up this discussion again so that we can
>> move forward. I prefer the GUC as it is relatively simple and as
>> Peter mentioned it works, but I'm not married to the idea.
>
> It's not very friendly to extensions, where the types are not
> guaranteed to have stable OIDs. Did you consider any proposals that
> work with type names?
Sending type names is kind of useless if what comes back with the result
(RowDescription) are OIDs anyway.
The client would presumably have some code like
if (typeoid == 555)
parseThatType();
So it already needs to know about the OIDs of all the types it is
interested in.