On Thu, Nov 17, 2022 at 7:09 AM Peter Eisentraut <peter.eisentr...@enterprisedb.com> wrote: > > Note that the above code was just changed in dce92e59b1. Thanks!
> I don't know > how that affects this patch set. With dce92e59b1 it would be much easier to find a bug in the compression patch. Some more notes about the patch. (sorry for posting review notes in so many different messages) 1. zs_is_valid_impl_id(unsigned int id) { return id >= 0 && id < lengthof(zs_algorithms); } id is unsigned, no need to check it's non-negative. 2. This literal {no_compression_name} should be replaced by explicit form {no_compression_name, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL} 3. Comments like "Return true if should, false if should not." are useless. Thanks! Best regards, Andrey Borodin.