> How about limiting the number of fetches? We already track that, so a
> limit would be computationally cheap. Numbers ecords read is another
> possibility, but it would have to be records fetched rather than records
> returned. Number of update operations, maybe?

User, even very experienced, can't prognose such things as number of
fetches (or page reads). Statement execution time however is usually
known to him. At least he can say for sure about upper time limit after
which statement execution must be stopped. Per statement timeouts can
be very useful for dynamic statements generated by application at run-time
(especially if application has end-user constructor) and at time of development.

IMO of course

Regards,
Vlad