In a message dated 11-19-94, Dean Roddey said to Raetus Egli:

DR>Thanks Raetus Egli for your msg about Colorados/2\Warp, on Tuesday,
DR>11-15-1994!

DR>Do you make use of any of HPFS386's security features when running
DR>locally? Maybe that would explain your differences of opinion. Some
DR>folks would certainly consider local file security enough of a reason
DR>to give up some performance, and maybe without any security in place
DR>HPFS386 is faster locally.

Yes, I don't use local security.  I haven't tested HPFS386 performance
using local security.  At every new operating install I make sure that
I remove all local security.  Sometimes HPFS386 thinks that local
security is enabled after a Lan Requester install and before HPFS386
is enabled.

Whatever the issue on local security, I have never heard of HPFS16
performance being comparable, though it does appear that with
WARP the gap may have closed a little more.

Note that Albert Shan and Craig Swanson have experienced very similar
performance numbers as I have.

It should also be noted that even HPFS386 performance doesn't appear
to be in the same ballpark as NTFS or the 386 file access methods
in Win95.  Albert Shan ran some very eye opening benchmarks, most
of the improvements can most likely be attributed to the use of
a dynamic cache.  

However, my point was that IBM ought to include their fastest file system
with their OS as the competition has obviously left them in the dust.
I understand the marketing considerations for Lan Server but if OS/2
compares very unfavorably to Win in disk benchmarks Lan Server will
not be an issue much longer either.

...Raetus
___
 * MR/2 2.1 #179 * Delivered by Electronic Sled-Dogs.....Woof!

--- Maximus/2 2.02