Re: Nearest Common Ancestor Report (XDb1's $1000 Challenge)

From: Neo <neo55592_at_hotmail.com>
Date: 3 Jun 2004 11:26:19 -0700
Message-ID: <4b45d3ad.0406031026.1b60c54b@posting.google.com>

> I couldn't care less if your next version will create a full report on one
> million things and 10000 classes in less than a nanosecond on a 80286
> processor.

While not quite less than a nano or 286, compare with performance on a Pocket PC (PPC):

Small Report Generation Summary

Solution Time(ms) Platform Notes

-------------- -------- ------------   ------------------------------
RM#2 SqlSrvr2K  11.0    1.3 Ghz PC    -Data not normalized to symbols
                                       Avg of 10 conseq runs.
                                       (?Last 9 possibly cached?)
XDb1 4.5.9       6.561  233 Mhz PPC   -Data normalized to symbols
                                       Avg of 10 "virgin" runs.

Large Report Summary - 200 goats (5 gen x 40/gen, each having two parents, except root). Generates 28,940 rows in RM's result table.

Solution Time(sec) Platform Notes

------------- --------- ------------   ------------------------------
RM#1 SqlSrvr7   15.2    500 Mhz Srvr  -Data not normalized to symbols
                                       Avg of 10 conseq runs.
                                       (?Last 9 possibly cached?)
XDb1 4.5.9      16.971  233 Mhz PPC   -Data normalized to symbols
                                       Avg of 3 "virgin" runs.

Received on Thu Jun 03 2004 - 13:26:19 CDT