Thread: several attstattarget-related improvements
Here are a few patches related to attstattarget: - 0001: Change type of pg_statistic_ext.stxstattarget, to match attstattarget. Maybe this should go into PG16, for consistency? - 0002: Add macro for maximum statistics target, instead of hardcoding it everywhere. - 0003: Take pg_attribute out of VacAttrStats. This simplifies some code, especially for extended statistics, which had to have weird workarounds.
Attachment
On 6/28/23 16:52, Peter Eisentraut wrote: > Here are a few patches related to attstattarget: > > - 0001: Change type of pg_statistic_ext.stxstattarget, to match > attstattarget. Maybe this should go into PG16, for consistency? > > - 0002: Add macro for maximum statistics target, instead of hardcoding > it everywhere. > > - 0003: Take pg_attribute out of VacAttrStats. This simplifies some > code, especially for extended statistics, which had to have weird > workarounds. +1 to all three patches Not sure about 0001 vs PG16, it'd require catversion bump. regards -- Tomas Vondra EnterpriseDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company
Tomas Vondra <tomas.vondra@enterprisedb.com> writes: > On 6/28/23 16:52, Peter Eisentraut wrote: >> - 0001: Change type of pg_statistic_ext.stxstattarget, to match >> attstattarget. Maybe this should go into PG16, for consistency? > Not sure about 0001 vs PG16, it'd require catversion bump. Yeah, past beta1 I think we should be conservative about bumping catversion. Suggest you hold this for now, and if we find some more-compelling reason for a catversion bump in v16, we can sneak it in at that time. Otherwise, I won't cry if it waits for v17. regards, tom lane
On 28.06.23 23:30, Tomas Vondra wrote: > On 6/28/23 16:52, Peter Eisentraut wrote: >> Here are a few patches related to attstattarget: >> >> - 0001: Change type of pg_statistic_ext.stxstattarget, to match >> attstattarget. Maybe this should go into PG16, for consistency? >> >> - 0002: Add macro for maximum statistics target, instead of hardcoding >> it everywhere. >> >> - 0003: Take pg_attribute out of VacAttrStats. This simplifies some >> code, especially for extended statistics, which had to have weird >> workarounds. > > +1 to all three patches > > Not sure about 0001 vs PG16, it'd require catversion bump. committed (to PG17 for now)