Alvaro Herrera <alvherre@alvh.no-ip.org> writes:
> I tried this out. It looks good to me, and I like it. Not translating
> the labels seems correct to me.
> +1 for backpatching to 16, given that it's a psql-only change that
> pertains to a backend change that was done in the 16 timeframe.
Agreed. In the interests of moving things along, I'll take point
on getting this committed.
> Regarding the controversy of showing SET for previous versions, I think
> it's clearer if it's shown, because ultimately what the user really
> wants to know is if the role can be SET to; they don't want to have to
> learn from memory in which version they can SET because the column is
> empty and in which version they have to look for the label.
Seems reasonable. I'll go with that interpretation unless there's
pretty quick pushback.
regards, tom lane