Robert Haas <robertmhaas@gmail.com> wrote:
> Jeff Janes <jeff.janes@gmail.com> wrote:
>> I found the following message somewhat confusing:
>> ERROR: read-only property must be set before any query
>
> I think what we need here is two messages, this one and a similar
> one that starts with "read-write property...".
Done. I started out by being cute with plugging "only" or "write"
into a single message, but then figured that might be hard on
translators; so I went with two separate messages.
Also, I noticed we seemed to be using "transaction read-only mode"
and "transaction read-write mode" elsewhere, so I made this
consistent with the others while I was at it. Hopefully that was a
good idea.
>> When a subtransaction has set the mode more stringent than the
>> top-level transaction did, that setting is reversed when the
>> subtransaction ends (whether by success or by rollback), which
>> was discussed as the desired behavior. But the included
>> regression tests do not exercise that case by testing the case
>> where a SAVEPOINT is either rolled back or released. Should
>> those tests be included?
>
> +1.
Done.
-Kevin