Thanks to Phill Hugo for this
link.
“Further to your agreement that the inclusion of a discussion of
Price/Performance between NT and Linux is a good idea in the next
report, it seems to me that another good idea is to repeat the
benchmarks with Windows NT clients.”
“Designing experiments to show what users can expect from
choices they may have to make is essential if you are to be a truly
useful source of information. It seems from the details of the
“Open Test” that you realize this and have made provision for the
testing to cover a single CPU system. I applaud this inclusion and
would like to see other variables involved. NT clients seem to be
the biggest request from the Linux interested and your quotes of
certain ‘Linux Experts’ only wanting to it benefit Linux seem to me
simply PR stunts. If this statement is true, that NT clients are
only desired to favour Linux/SAMBA, then the converse is also just
as true – that Windows 9X clients were chosen as they favoured
NT.”