Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 14 Mar 2001 21:34:51 -0500
From:      "Andrew C. Hornback" <hornback@wireco.net>
To:        "Mike Meyer" <mwm@mired.org>
Cc:        <chat@freebsd.org>
Subject:   RE: Now a little OT but RE: FreeBSD and Linux (More Questions!)
Message-ID:  <02d401c0acf8$833e77f0$0f00000a@eagle>
In-Reply-To: <15024.1411.79596.364926@guru.mired.org>

next in thread | previous in thread | raw e-mail | index | archive | help
> -----Original Message-----
> From: Mike Meyer [mailto:mwm@mired.org]
> Sent: Wednesday, March 14, 2001 6:58 PM
> To: Andrew C. Hornback
> Cc: chat@freebsd.org; tedm@toybox.placo.com
> Subject: Re: Now a little OT but RE: FreeBSD and Linux
> (More Questions!)
>
> [Redirected from -questions.]
>
> Andrew C. Hornback <hornback@wireco.net> types:
> > > >also remember working with snow white and the seven dwarfs, and
> > > >already knew that interoperability was something you got
> > > in a computer
> > > >line, at least until the manufacturer decided to play
> > > shell games to
> > > >kill the third party hardware market.
> > 	Hmm... shall we say, IBM here?  Just how many
> manufacturers adopted
> > MCA?  IBM, Tandy did a couple of machines with it ('course, Tandy
> > always followed IBM's lead with strange hardware
> designs... Tandy 1000
> > SX followed the PC jr., need I say more?).  I believe Siemens,
> > Northridge, and possibly Fujitsu produced machines with
> MCA support.
>
> Actually, I was thinking more along the lines of IBM moving the
> controller smarts from the system box into the drive in the 360/370
> line. A federal judge told them their near-monopoly
> position made that
> an anticompetitive practice, and hence illegal. The other case that
> comes to mind was their adding instructions for an OS release to
> prevent Amdahl from using the OS release directly. Amdahl's response
> was to provide a version of the OS with the "new" instructions
> unrolled inline. I'd call that Open Source software at it's
> best, but
> I don't know that it met the current definition for Open Source.

	Egads... someone's showing their age... or am I showing my lack
thereof?

	I don't remember this... but it sounds like it was from the late
70s/early 80s.  Heck, I barely remember Amdahls at all... I've only
seen pictures of 'em, as far as I know.  *grins*

> > > I don't remember if there was a lot of crossover between
> > > the mainframe and
> > > PC people back then - I think the markets were pretty
> > > foreign to each
> > > other.
> > 	There never really was a crossover.
>
> There may not have been a crossover of manufacturers, but there was
> pretty clearly a customer migration. I watched it happen - and my
> career followed, going from big iron to minis to desktop
> workstations. So far I've managed to avoid anything that
> looks like MS
> OS support, and have no plans at all to change *that*.

	In the installations that I've seen, the move was from 'frames to
client/server.  Sign of the times, the colleges that I went to just
decomissioned their ES/9000 and Vax installations in the past couple
of years.  To make matters worse, both institutions are stuck with
thier old big iron until the next remodel of the buildings come along.
Seems that in the last remodel at each facility, someone figured that
the big computer would be there for years upon years, and they did the
remodel in a manner that wouldn't allow either machine to leave the
premises.

> The original comment was about standardization and the PC market.
> Most of the people playing with PC hardware I knew in the early days
> had done the migration. They took incomparability between
> manfacturers
> as a given, and they didn't push for any kind of standardization.

	*gets a feeling that "back in the early days" refers to back when the
closest he got to a PC was watching IBM sponsor Face the Nation on CBS
on Sunday mornings...*

	I'm out of my tree here, probably, since I've only been working with
PCs for about 16 years... and back then, it was a Timex/Sinclair 1000
with the 2k RAM pack and interpreted BASIC... *Grins*

> > Gotta look at the big mainframe producers and see where they are
> > now.  Where's Cray?  Last I heard they were part of SGI, who was
> > also going in the tank over their line of NT based workstations
> > (what the hell were THEY thinking?).  IBM's Mainframe
> > business... how many ES/9000s do you see being sold each year?
> > They've moved to the AS/400 and RS/6000 lines.  And
> Digital?  Now a
> > division of Compaq, simply because Compaq couldn't build
> a high end
> > machine to save their life and didn't know what good customer
> > support was.  Then there was Unisys and their Clearpath
> line, which
> > may still exist, not really sure.  But I know they've cut
> out their
> > consumer PC division, and used a lot of resources to build the new
> > high end servers that have CMP technology.  What about
> Data General,
> > or Wang, or ...
>
> Yup - if your customers migrate and you're not selling where they're
> going, you're going to be in trouble.

	It's kind of sad, seeing the big iron makers get bought up.  Their
technical excellence gets strewn about, bastardized and implemented in
a corporate environment before integrated into a new chassis with a
familiar label that doesn't do them justice.

	As an example, just look at what Gateway did to my favorite server
manufacturer, ALR.  'nuff said.

> > 	The mainframe is getting to be like a classic car
> from the 50s.  Nice
> > to look at, the new generation Oohs and Aahs over it, but no one
> > really wants to touch it for fear of breaking it and no
> one wants to
> > support it any more due to the costs of replacement
> parts, etc. (not
> > that mainframe parts were ever cheap...)
>
> Actually, mainframe replacements aren't very expensive any more. You
> don't buy mainframe, you buy a stock Wintel box with a 3081 - or
> whatever - on a card, and run your legacy (industry jargon for
> "working") systems on that.

	I've never seen anything like that, although, I'd be interested in
finding vendors of these sorts of pieces.  In the back of my mind, I
have a sort of Frankenstein project of taking an old Intel based
server and adding a Sparc emulator card, an Alpha emulator card, etc,
etc. and hopefully get to the point where I have each slot filled with
a different type of processor, just to see if it can be done.
'course, I'm also building a 6X6... *evil grin*

> > 	Which reminds me... has anyone seen the new Intel
> vision of what a
> > consumer PC is going to be?  It's basically a stack of
> boxes, like an
> > Aztec temple, each one holding a component or two.
> Foundational box
> > holding the motherboard, processor and memory.  Next step
> up holding
> > the DVD-RAM drive, followed up the next steps containing
> the HDD, the
> > other removable media drive (looked like a Zip drive), and the top
> > being the control and I/O panel with all of the ports on
> top.  Gone
> > are your PS/2 ports for mouse and keyboard, replaced by
> USB.  Gone are
> > your serial and parallel ports, replaced by USB.
>
> Ok - what's the drive interconnect? Are they actually running IDE to
> external boxes?

	They're using USB to interconnect the drive hardware, and since you
can just add on USB devices whenever and have up to 127 possible
devices per USB chain... they think that'll make for all the
expandability you'll ever need...

> > 	This is what they want the PC to be once the IA-64
> hits mainstream.
> > Which basically means that when the Itanium gets out
> there and into
> > the hands of more than the technophiles, nearly everything that we
> > have now is going to be obsolete.
>
> Actually, it all already is, in the sense that you can buy
> newer/faster/better hardware.

	Newer/Faster/Better isn't always the case.  Intel's 8xx motherboard
fiascos and the introduction of RIMMs... need I say more?

> For instance, I moved my
> primary printer
> (an LJ5M) to USB to attach an old color printer, and the I/O numbers
> on the postscript benchmark went through the roof. I assume
> that if I
> had a real USB printer, it'd be even better. So if I replace the
> antique color printer that's on the parallel port, I'm going to buy
> USB, and disable the parallel port in the BIOS.

	*nods in agreement*  Some new things are good... USB, higher speed
SCSI, higher speed drives, memory, RAM, etc.  But, there are always
"innovations" that crater the runway like a Paveway II.

> > 	Thank you Intel!
>
> Have you forgotten that Apple has already done that - and apparently
> survived?

	Hmm... comparing the world's largest CPU manufacturer to a company
that's dropping like a stone and has only been propped up by a large
infusion of cash from Microsoft and Pixar in addition to the
production of some nice workstations made of cheap plastic... I don't
think we're talking apples to apples here.

> Personally, I like going to USB for that stuff,
> as it frees
> up IRQs, which are a precious commodity. I'll be replacing my ISA
> hardware with USB by attrition, though.

	I guess I'm more "old school" when it comes to things like that.
'course, that's probably based on the fact that now I can "play" with
the hardware that I always wanted as a kid but never could afford...

--- Andy
(CC any replies to me, as I'm not on -chat list, thank you.)



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-chat" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?02d401c0acf8$833e77f0$0f00000a>