Tom Lane wrote:
> Florian Pflug <fgp.phlo.org@gmail.com> writes:
>> Tom Lane wrote:
>>> I think txid_current_snapshot should read ActiveSnapshot. If the user
>>> wants to get a beginning-of-xact rather than beginning-of-statement
>>> snapshot from it, he should be required to call it in a serializable
>>> transaction.
>
>> Hm... does txid require that the snapshot it uses a valid in the sense that
>> its xmin follows OldestXmin? If not, we could keep the snapshot around for
>> txid, but still update our published xmin - which seems to be the main
>> reason we care about getting rid of old snapshots at all.
>
> Why should we complicate the main code like that for txid? I have not heard
> any argument why the function should be examining SerializableSnapshot
> instead of the current transaction snapshot.
I wouldn't know. I just wanted to say that even if it needs to examine
SerializableSnapshot, that won't clash with the xmin optimizations planned for
8.4, as long as the snapshot won't be used for actual queries.
greetings, Florian Pflug