Re: This is weird: program appears to run faster from a "real" library than from QTEMP

Hi James,

What was the correction? You said you ran the uncorrected version vs the corrected version, so I'd suspect something to do with the correction before I'd think of which library it was running out of.

And of course, the usual stuff with performance issues: what else was running at the time? was the data the same? are you comparing clock times or CPU times?

--
*Peter Dow* /
Dow Software Services, Inc.
909 793-9050
petercdow@xxxxxxxxx <mailto:petercdow@xxxxxxxxx>
pdow@xxxxxxxxxxxxxx <mailto:pdow@xxxxxxxxxxxxxx> /

On 5/27/2016 10:31 AM, James H. H. Lampert wrote:

This is weird.

I've got a program running on a customer box. Earlier, I'd recompiled it into QTEMP, in order to get a compilation listing, so I could determine why it was locking up. I put in a correction, and recompiled it into its usual library.

I then ran what I thought was the corrected version. But because that particular box has QTEMP at the top of the default user library list, I was in fact running the uncorrected version I'd compiled into QTEMP.

After it locked up, I got rid of the QTEMP version, and re-ran the corrected version. It seems to be running at least twice as fast from a "real" library as from QTEMP.

Any idea why that could be?

--
JHHL