Opened 7 years ago

Closed 7 years ago

#3734 closed defect (duplicate)

PostgreSQL 10 regress_selectivity fails

Reported by: robe Owned by: pramsey
Priority: blocker Milestone: PostGIS 2.4.0
Component: postgis Version: master
Keywords: postgresql 10 Cc:

Description

Now that I got PostGIS at least compiling against PostgreSQL 10, back to troubleshooting regressions. We have 2 currently, this one appears to be a new one, but not sure when it started since our postgresql 10 tests could not be run because it couldn't compile.

This might not be an issue — strk how do these selectivity function thing work?

The actuals always agree, but I looked and saw those are just constants you have like:

select 'selectivity_02', 'actual', round(1068.0/2127.0,3);

so no big surprise there.

--- regress_selectivity_expected	2015-07-30 14:50:49.211290800 -0400
+++ /projects/postgis/tmp/2.4_pg10w64/test_64_out	2017-04-08 02:47:01.285321700 -0400
@@ -2,16 +2,16 @@
 ERROR:  stats for "no_stats.g" do not exist
 ERROR:  stats for "no_stats.g" do not exist
 ERROR:  stats for "no_stats_join.g" do not exist
-selectivity_00|2127
-selectivity_01|1068
+selectivity_00|715
+selectivity_01|245
 selectivity_02|actual|0.502
-selectivity_03|estimated|0.502
-selectivity_04|161
+selectivity_03|estimated|0.638
+selectivity_04|55
 selectivity_05|actual|0.076
-selectivity_06|estimated|0.076
+selectivity_06|estimated|0.119
 selectivity_07|81
 selectivity_08|actual|0.038
-selectivity_09|estimated|0.038
+selectivity_09|estimated|0.231
 selectivity_10|actual|0
 selectivity_09|estimated|0
 selectivity_10|actual|1

Change History (1)

comment:1 by robe, 7 years ago

Resolution: duplicate
Status: newclosed

nevermind not new already have this in there.

This is dupe of #3694

Note: See TracTickets for help on using tickets.