Opened 14 years ago
Closed 13 years ago
#967 closed defect (fixed)
[raster] raster regress failures - raster functions not being installed right during regress
Reported by: | robe | Owned by: | pracine |
---|---|---|---|
Priority: | medium | Milestone: | PostGIS 2.0.0 |
Component: | raster | Version: | master |
Keywords: | Cc: |
Description
This is a followup to Paul's comment.
http://postgis.refractions.net/pipermail/postgis-devel/2011-May/013066.html
I took get the same failures — run tests: 53 failed: 40
I apologize for not having brought it up earlier. I just assumed it was an isolated mingw issue.
Anyrate on further inspection — the errors are largely because raster fails installation partly.
I have no raster operators installed under regress install (though all installs fine under regular conditions).
Just to point out the bogusness of these errors
select st_isempty(rast) from empty_raster_test;
fails with function st_isempty(raster) does not exist
Works fine via my script install.
When I compare the postgis_reg to my standard install, I see that for postgis_reg, none of the raster casts are installed though the raster type does exist.
So all the tests that rely on raster 2 geometry casts fail.
The only thing I see suspicious in the regress log is this
psql:c:/projects/PostGIS/trunk/regress/../raster/test/regress/rtpostgis.sql:1471: ERROR: cannot change name of input parameter "pixeltype" HINT: Use DROP FUNCTION first.
But I don't get that error during my regular install of raster
Change History (5)
comment:1 by , 14 years ago
comment:2 by , 14 years ago
Pierre,
I was wrong I think I am getting that error on install. Guess I missed it. It's actually a bug in your code that only PostgreSQL 9.0+ would pick up because it supports named arguments. I'll put in a ticket for that.
Though still haven't figured out why the regress fails for Paul and me so badly.
My 8.4 raster regress doesn't work at all for some reason just crashes.
comment:3 by , 13 years ago
I have this occurring on 9.0 because of the last ST_MapAlgebra function (line 1827). Once that function is commented out, running "make check" completes with no failures.
comment:5 by , 13 years ago
Resolution: | → fixed |
---|---|
Status: | new → closed |
Just recompiled everything (PostGIS included)(but I still have PG 8.4.1) and everything passed fine except for the three tests Bborie is working on.
Run tests: 53 Failed: 3