2010/1/12 Boszormenyi Zoltan <zb@cybertec.at>:
> Tom Lane írta:
>> Alvaro Herrera <alvherre@commandprompt.com> writes:
>>
>>> But it would be broken in very obvious ways, no? It's not like it would
>>> be silently broken and thus escape testing ...
>>>
>>
>> Well, if we wanted to adopt that approach, we should add the count to
>> *all* SELECT tags not just a small subset. As the patch stands it
>> seems entirely possible that a breakage would escape immediate notice.
>>
>
> Can you give me an example that would return
> plain "SELECT" after my new patch? I added
> one more change to the patch, is it enough to return
> "SELECT N" in every case now?
I just tested this, so I can say definitely: no. I hacked psql with
the attached patch, and if you just do a plain old SELECT * FROM
table, you get back only SELECT, not SELECT N.
...Robert