Very Computer

Quote:>I'm not slamming MS but they need to rework their strategy a little if
they're
>going to compete with FreeBSD/Linux and UNIX.

        For the desktop, yes.  For the server market, no.  Microsoft is
        pushing NT hard at the mid level server market, mostly dominated
        by Unix right now, with good reason.
--

BSD:  A psychoactive drug, popular in the 80s, probably developed at UC
Berkeley or thereabouts.  Similar in many ways to the prescription-only
medication called "System V", but infinitely more useful. (Or, at least,
more fun.)  The full chemical name is "Berkeley Standard Distribution".

Quote:> Like why should we go out and spend more money to get a Pentium II
> 400-500MHZ with 128 Megs RAM to run Windows NT 4 when you can make
> do with an existing Pentium MMX and run FreeBSD/X for practically nothing?

Greetings,
                                Ripley
--
http://www.in-berlin.de/User/nostromo/
==
"You don't say what kind of CD drive or hard disks you have, but since it is
causing you trouble I'll assume it is IDE."  -- comp.unix.bsd.freebsd.misc

If you are running 32 bit PCI instead of 16 bit ISA, this is more
like 264%.

That's 13.2 seconds scaled from the 5 you reported, or 8.2
additional.  20 - 5 is 15; 30 -5 is 25.

So that's between 33 and 55% of your time right there.

8-).

The obvious conclusions we can draw from this is that clock
multipliers are evil, and the major bottleneck in modern
systems is I/O, not CPU speed.

                                        Terry Lambert
---
Any opinions in this posting are my own and not those of my present
or previous employers.

This is essentially true, but clock multipliers are not the whole
story.  For one thing, it's time for the ISA bus to go.  Period.  An 8
MHz bus (typically sharing bandwidth with 2-3 cards) running on a
system with a 400 MHz CPU is just ridiculous.  Try a PCI soundcard
sometime and run something like MS Flight Simulator, where there is
constant sound output (aircraft sounds, ATC talk, etc)... the frame
rates and speed of everything increase significantly, due to the sound
card being on the PCI bus with massively increased bandwidth available
and 4 times the speed.  So the issue *IS* I/O, but it's more complex
than just CPU multipliers.  Faster hard drives will help, phasing out
the ISA bus will help, faster bus speeds OVERALL will help... there
are many factors.  Also, the less mechanical/moving parts in a PC, the
better (some kind of non-mechanical mass storage at consumer prices,
like flash memory in hard drive sizes, is long overdue.  Imagine your
OS booting in 2-3 seconds from a flash memory device that retains
storage with power turned off :-).  CD-ROM and DVD has gotta go too...
we don't need motors.  I'd love to see a PC without a single motor in
it, even for cooling purposes.  What we need is EVERYTHING solid
state.  *THEN* PC's will fly like the Almighty itself.

--

  Chris Gray

72MB DiskOnChip modules are widely available; 144MB modules
are starting to show up.  Several vendors have IDE interfaces for these
chips; you could build a machine today which boots from a solid state
disk.

: > some kind of non-mechanical mass storage at consumer prices,
: > like flash memory in hard drive sizes, is long overdue.  

    I find it amusing that since the time I first speculated about having
enough battery backed RAM to replace my hard disc memory prices have indeed
fallen so far that the RAM (today) would be cheaper than the hard disc (then).

Unfortunately 20Mb hard discs are no longer considered mass storage :(

    I rather suspect that no matter how cheap 'fast' memory gets there will
always be 'slow' and 'bulk' memory that is cheaper.

I thought you were "Leaving" this newsgroup.  Believe me, it was so much
quieter and more pleasant with you gone.  Please "leave" again.

--

V xabj jub lbh ner.  V fnj jung lbh qvq.

I agree, but that is tangential.  I was assuming a 25MHx bus to
get the high percentage number I got.  For an 8MHz bus, the
machines are simply incomparable.  If that was the premise,
the conclusions are less than unsupported: they are ludicrous,
since the 486 was outperforming the Pentium, were that the case.

                                        Terry Lambert
---
Any opinions in this posting are my own and not those of my present
or previous employers.

 -----
|No spam protection here, I have a "trash" button on my mailreader (and
|I don't post to the "trash" newsgroups).  If you want to Email me, just
|reply normally.  My website is at http://www.serv.net/~fewtch/index.html -
|something for everyone there, take a look if you have the time. Ooga booga.

I *HAVE* "left."  This message has been crossposted for some reason to
comp.unix.bsd.freebsd.misc (check your headers).  I certainly don't
use FreeBSD anymore and don't intentionally post to c.u.b.f.m... but
I'm too lazy to check headers on a regular basis, so if somebody
crossposted it, don't blame me, blame the person who added it to the
list.  Probably some moron "out to get me" for my comments on the
2.2.7 piece of crap^h^h^h^h^h^h^h^h^h^h^h^h^h^h release.  I'm still
getting residuals from that almost a month later, believe it or not.
Some people have nothing better to do with their time than defend an
OS that, in their own viewpoint, doesn't even need defending.  BTW (as
long as I'm here for the moment), any backlash yet from the bugs that
went out on the 2.2.7 CD?  Anyone with any info, I'd be curious to
know about it.  Personal flames will be posted permanently on my "hall
of shame" web page (along with the flamer's Email address), indexed by
at least 100 search engines all over the world.  IOW, comments
welcome, flames not.

Further replies from me on this thread will have the crosspost to
c.u.b.f.m removed.  I'd encourage everyone to do the same.  What does
FAT32 have to do with FreeBSD anyway, it's basically off-topic there.

My opinion?  I still think Maxtor and Western Digital will be Flash
Memory companies 10-15 years (max) from now.  NOTHING motorized will
ever be 1/1000 as reliable or long-lasting as a RAM chip (or any other
form of nonmechanical storage).  A RAM chip basically lasts forever
(although I thought I remember reading somewhere that silicon-based
chips begin to fail after 30 years or so - could be mistaken).

 -----
|No spam protection here, I have a "trash" button on my mailreader (and
|I don't post to the "trash" newsgroups).  If you want to Email me, just
|reply normally.  My website is at http://www.serv.net/~fewtch/index.html -
|something for everyone there, take a look if you have the time. Ooga booga.

 -----
|No spam protection here, I have a "trash" button on my mailreader (and
|I don't post to the "trash" newsgroups).  If you want to Email me, just
|reply normally.  My website is at http://www.serv.net/~fewtch/index.html -
|something for everyone there, take a look if you have the time. Ooga booga.