Ray asks:
> Gavin, did you do the DSCOPY first, and then the FTP?  And if yes, do you
> think that the file being in memory might have made some of the diff?

I purged and recreated the source file in the same way before each test to
eliminate this issue.  The performance difference is so dramatic that you
don't even need to be this careful.

But other than that it was a very quick test just to verify that what I
already believed was in fact the case (at least in the environments I've
looked at).

I'd be happy to hear of cases where :DSCOPY wins, of course.

If :DSCOPY is doing its own acknowledgements and FTP isn't, then it almost
doesn't matter how either program is implemented.  This simple algorithmic
difference means that FTP will always win.

:DSCOPY - The KERMIT of HP3000 file transfer protocols.

G.

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *