Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-09-03 Thread Michael Glaesemann
On Sep 3, 2006, at 12:34 , Bruce Momjian wrote: OK, I worked with Michael and I think this is the best we are going to do to fix this. It has one TSROUND call for Powerpc, and that is documented. Applied. As I was working up regression tests, I found a case that this patch doesn't

Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-09-03 Thread Tom Lane
Bruce Momjian [EMAIL PROTECTED] writes: Is this non-datetime integer only or both? I cannot reproduce the failure here. On HPPA with float datetimes with today's code, Michael's case works but it took me less than two minutes to find one that doesn't: regression=# select interval '14 mon' *

Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-09-03 Thread Michael Glaesemann
On Sep 4, 2006, at 4:45 , Bruce Momjian wrote: Another question. Is this result correct? test= select '999 months 999 days'::interval / 100; ?column? - 9 mons 38 days 40:33:36 (1 row) Should that be: 9 mons

Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-08-30 Thread Michael Glaesemann
On Aug 30, 2006, at 12:50 , Bruce Momjian wrote: Yea, I see that -122:23:60.00. After applying your patch, I believe that on my machine it's the contribution from the day component that is producing the 23:60.00. For example, select interval '-12 days' * 0.3; ?column?

Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-08-29 Thread Michael Glaesemann
I think I've got it. I plan to update the regression tests this evening, but I wanted to post what I believe is a solution. select '41 mon'::interval / 10; ?column? --- 4 mons 3 days (1 row) select '41 mon 360:00'::interval / 10 as pos , '-41 mon -360:00'::interval / 10 as

Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-08-29 Thread Michael Glaesemann
On Aug 30, 2006, at 1:13 , Bruce Momjian wrote: Uh, I came up with a cleaner one, I think. I didn't test --enable-integer-datetimes yet. Cool. It's indeed much cleaner. Thanks, Bruce. I'm about to head to bed, but I'll look at it more closely tomorrow. I also noticed that my regression

Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-08-29 Thread Michael Glaesemann
On Aug 30, 2006, at 7:12 , Bruce Momjian wrote: Here are the results using my newest patch: test= select interval '41 mon 12 days 360:00' / 10 as quotient_a , interval '41 mon -12 days -360:00' / 10 as quotient_b , interval '-41 mon 12 days 360:00' / 10 as

Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-08-29 Thread Michael Glaesemann
On Aug 30, 2006, at 12:50 , Bruce Momjian wrote: Here is a test program. What does it show for you? The output for me is: 4.100 2.989 3.000 Here's what I get. Just to make sure I'm doing this right, I'm including how I

Re: [PATCHES] [HACKERS] Interval aggregate regression failure (expected seems

2006-08-28 Thread Michael Glaesemann
On Aug 26, 2006, at 11:40 , Bruce Momjian wrote: I used your ideas to make a patch to fix your example: test= select '41 months'::interval / 10; ?column? --- 4 mons 3 days (1 row) and test= select '41 months'::interval *