On Tue, 2003-02-25 at 19:26, et wrote:
> On Tuesday 25 February 2003 05:01 pm, Joe Braddock wrote:

> > Anyone who dual boots with windows on the same hardware knows that windows
> > performance is noticeably better than Linux, and that this is most likely
> > due to the highly x86 optimized Intel compiler.

> Please... this just is not a fact. this is basically WRONG in my personal 
> experience. M$ just plain sucks compared to Mandrake-Linux on at least _MY_ 
> hardware, I don't claim to speak for "anyone who dual boots windows, and as 
> long as I am included in your statement, neither do you.

You are not the only one debunking the bunk here, ed; I've got a dual
boot system and the performance of 98 has NEVER been better than LM82.

In fact, I'll relate to you a little story that you may find
interesting.  I loaded LM82 on a bud's machine, about the same time that
we loaded 98 on his machine with all the latest wiz bang VIA mobo
drivers and 98 patches; other than that it was a vanilla install.  It
was a dual boot on the same HD, an IBM Deskstar.  To go a step further,
we were using 98Lite, so when I got through, Internet Exploder was taken
out.  This speeded up 98 even more.

A few weeks later he started telling me that 98 was loading faster than
LM82.  "Is that so?" I said.  He assured me that it was indeed the
case.  I asked him how he came by his conclusion.  "It just looks
faster," he proclaimed.

"OK," I said, nonplussed.  I went back to what I was doing.  The next
day he showed up again, and bragged once again that LM82 load time was
considerably more than 98.  "OK," I said.  This time I was pissed.  I
went out to the shop.

"Where are you going?" he asked.

"To get my stopwatch."

So we went back to his machine (which I had constructed, btw), and we
timed the boot time on the stopwatch.  LM82 was a full 25 seconds faster
loading at boot than 98 was, starting from power on to GUI load.  And
after that bench, I discovered that I had left named, iplog, and a few
other services running that he didn't need.  So I did away with those in
chkconfig and widened the lead even further.

The moral?  Never ever believe the user's perceptions of speed unless
they have been quantified with benchmarks. Same goes for your own
perceptions of speed.

> >
> > I thought anything after Windows NT runs on Pentium and above (or K6 and
> > above if you use AMD).  If that's the case, then it's not anymore optimized
> > than Mandrake, so any increased performance on a Windows box is not do to
> > optimization.

> not true, you need much more starting with win ME

Ah, 98 as well.

LX


-- 
���������������������������������������������������
Kernel  2.4.21pre4-6mdk       Mandrake 9.1 RC1
Enlightenment 0.16.5-12mdk    Evolution 1.2.2-1mdk
Registered Linux User #268899 http://counter.li.org/
���������������������������������������������������


Want to buy your Pack or Services from MandrakeSoft? 
Go to http://www.mandrakestore.com

Reply via email to