said: 
>Yes, this was done on purpose.  The code to write the records to an 
>output file would be identical in both cases, hence including it in the 
>benchmark would be irrelevant. 
 
The code would be identical, but the system I/O (i.e. head movement, 
caching etc.) would be very different.  It would only help the buffered 
method. 
 
Here's some results with output file I/O included: 
                         ---- Seconds (uSec/Rec) to Read ----  Speed File 
Size   Record Size  Individual Records  Buffered Records  Ratio ==========  
===========  ==================  ================  ===== 
   1000000           10          20.500000          11.110000 
   1000000          100           2.130000           1.190000 
   1000000         1000           1.150000           0.250000 
   1000000        10000           0.090000           0.110000 
   2000000           10          42.200000          22.580000 
   2000000          100           4.290000           2.440000 
   2000000         1000           0.800000           0.420000 
   2000000        10000           0.570000           0.200000 
   4000000           10          84.360000          46.090000 
   4000000          100           8.730000           4.950000 
   4000000         1000           2.250000           0.830000 
   4000000        10000           1.040000           0.450000 
   8000000           10         168.250000          92.620000 
   8000000          100          17.490000           9.870000 
   8000000         1000           5.180000           1.740000 
   8000000        10000           2.610000           1.150000 
  16000000           10         339.650000         187.320000 
  16000000          100          35.640000          19.700000 
  16000000         1000           9.420000           3.320000 
  16000000        10000           6.000000           1.800000 
Except for the 1000000/10000 case which I don't yet understand, the 
buffered method retains its overall advantage longer. 
Steven 
--  
---------------------------------------------------------------------- 
"Steven Levine"   MR2/ICE 2.37 #10183 Warp4/FP15/14.093c_W4 
www.scoug.com irc.webbnet.info irc.fyrelizard.org #scoug (Wed 7pm PST) 
---------------------------------------------------------------------- 
===================================================== 
To unsubscribe from this list, send an email message 
to "steward@scoug.com". In the body of the message, 
put the command "unsubscribe scoug-help". 
For problems, contact the list owner at 
"rollin@scoug.com". 
===================================================== 
<< Previous Message << 
 >> Next Message >>
Return to [ 29 | 
July | 
2003 ]
The Southern California OS/2 User Group
P.O. Box 26904
Santa Ana, CA  92799-6904, USA
Copyright 2001 the Southern California OS/2 User Group.  ALL RIGHTS 
RESERVED. 
SCOUG, Warp Expo West, and Warpfest are trademarks of the Southern California OS/2 User Group.
OS/2, Workplace Shell, and IBM are registered trademarks of International 
Business Machines Corporation.
All other trademarks remain the property of their respective owners.