On 10/05/2015 11:15 AM, Tom Lane wrote:
Peter Geoghegan <p...@heroku.com> writes:
I'm annoyed and disappointed that the patch committed does not even
begin to address the underlying problem -- it just adds an escape
hatch, and fixes another theoretical issue that no one was affected
by. Honestly, next time I won't bother.
The problem as I see it is that what you submitted is a kluge that will
have weird and unpredictable side effects. Moreover, it seems to be
targeting an extremely narrow problem case, ie large numbers of queries
that (a) have long query texts and (b) are distinct to the fingerprinting
code and (c) fail. It seems to me that you could get into equal trouble
with situations where (c) is not satisfied, and what then?
I'm certainly amenable to doing further work on this problem. But I do
not think that what we had was well-enough-thought-out to risk pushing
it just hours before a release deadline. Let's arrive at a more
carefully considered fix in a leisurely fashion.
FWIW, (a) and (b) but not (c) is probably the right description for my
client who has been seeing problems here.
cheers
andrew
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
