NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/sys/amiga/hardware-part1/8634

Text File | 1996-08-05 | 10.6 KB | 212 lines

Newsgroups: comp.sys.amiga.misc,comp.sys.amiga.hardware
Path: netnews.upenn.edu!dsinc!scala!news
From: [e-mail] (Dave Haynie)
Subject: Re: Dump the crappy hardware!!! (was: Haynie joins AT team)
Sender: [e-mail] (Usenet administrator)
Message-ID: <[e-mail]>
Date: Thu, 28 Mar 1996 21:28:15 GMT
Reply-To: [e-mail] (Dave Haynie)
References: <4j3ltf$[e-mail]> <[e-mail]> <[e-mail]>
Nntp-Posting-Host: gator
Organization: Scala Computer Television, US Research Center

In <[e-mail]>, [e-mail] ("Darren M. Greenwald") writes:
>In article <[e-mail]> [e-mail] (NeuroMancer) writes:

>>Michael van Elst, regarding your message ' Re: Dump the crappy hardware!!! (was: Haynie joins AT team)'  -
>>
>>>[e-mail] (Kevin Pluck) writes:

>>>>AT are making the wrong move with trying to design hardware for the amiga
>>>>when they should be making the amiga designed for other hardware.   Why
>>>>reinvent the wheel?

>>>Because you cannot make money with selling an operating system.

>>What's microslop doing then ?

>>Before you say they are making money off the apps running under
>>windoze, what makes you think that a new amigados could not do the
>>same ?

First thing, you have to realize how Microsoft got into the position
they currently occupy. At one time, Microsoft made pretty much all
their money selling the OS. Of course, at one time Microsoft was a
tiny company. I interviewed with them while I in college -- the had
about 500 employees, but doubled in size every year.

So sure, you can make money on an OS, under the right
circumstances. Being small is one of those circumstances, having
little to no competition is another. It'll be a real neat trick for
anyone to pull off that latter one. MS made money because the hardware
companies needed them to sell their hardware. Not big piles, at first,
but even pocket change amounts to something if you get it a 100
million times over.

Next came upgrades. While PC vendors paid peanuts for MS-DOS (and, for
a time, paid per unit produced, whether or not it has MS-DOS on it),
users paid more. Pretty early on, Microsoft learned two good rules of
the computer business: if it's bundled, you'll make a killing with
little extra work, and you can make more money on upgrades than
original sales. Ever since, Microsoft has worked more and more of
their products into that "bundle". And they more for an upgrade to
Word than they do for the whole Windows 95, and that's just one
application. Which makes sense -- they need to upgrade the OS or you
don't upgrade any application. But at the right price, they can afford
to let you think about upgrading the apps for a little while.

>There you go, keeping in mind the following:

>1.) Microsoft is the one making the money off the applications.
>The Commodore and Apple way of doing business has been quite the opposite.
>Commodore and Apple made/make money selling hardware, the OS being
>required for that hardware, and do not make any significant money
>selling application software.

In fact, Apple's OS was a way to sell their proprietary hardware; at
one time, they gave it away for free. Now, in the early days, it was a
comparatively cool OS and, because it did stuff no other personal
computer OS did, they could charge a vast fortune per unit. Now
they're getting eaten by their failure to realize that, since they
chose to be a hardware company (for a time, they could have done well
moving toward becoming a software company), they would have to compete
with the rest of the hardware world, as software caught up.

C= was always a hardware company, from the ground up. When everyone
made their own systems, C= did it better and cheaper, and the OS was
pretty much just a way to sell the computer. Fortunately the Amiga
started out with a very good OS, but that's not something C= typically
was concerned with from the top.

>2.) Microsoft does not sell computers (they do sell a few
>pieces of hardware of the kind that can be used in all of the
>rest of the PC's sold).  Again, opposite of what Commodore and
>Apple do.  While Commodore and Apple had/have to compete
>with PC prices, which translates in to decreased profits per
>system, Microsoft only benefits from the competition among
>PC clone vendors.

And it's was worse for C= and Apple than the typical PC company. While
they had more control over their hardware designs, they never tapped
into the whole economy of scale thing that happened in the late 80s
and early 90s. Under the old model, every PC vendor did most of their
own work: systems design, graphics, etc. Gradually, independent chip
companies filled the niches, much like Motorola and Intel has always
provided the CPUs. So today's system houses, for the most part, don't
develop their own chips or system designs. Some OEM whole systems or
motherboards, others may tweak an existing system design here and
there. This results in a major cost savings that Apple or C= could
never tap into.

And then there's software. A PC system house may do some BIOS work,
maybe even a device driver or two, maybe a few simple value added
attractor programs to get you to pick a Packard Bell over a Compaq
when you're wandering the aisles of Computer City. They don't, for the
most part, do real software development. C= and Apple, on the other
hand, did tons of software development, some of it even ground
breaking. Especially for Apple, the ground breaking part was
necessary, since under their model, the only reason you bought a Mac
was to get at this great software. Unfortunately, they couldn't move
on new technologies as fast as Microsoft, so even on the OS they're
lagging in many areas.

>3.) An operating system like Windows, or OS/2 must run on
>a much wider range of hardware configurations than Amiga OS or
>System 7.  While Commodore and Apple had/have control over
>the hardware and can test OS releases along with the hardware,
>it is a more difficult problem to build an OS that works with
>all of the various PC configurations out there.

Of course, neither Apple nor Commodore ever attempted to do a port to
the PC platform. They certainly could have, and for Apple at least, a
MacOS-86 in 1987 or so might have made this a different world. Unless
you're doing a port to the PC, though, this isn't a big issue. The
ability to support standard parts, as Apple is starting to do and
Amiga Technologies must if there's to be an Amiga future, is not a
promise to support everything that exists on the planet today. For the
PC, you have to. OS/2 and Windows came along and found a zillion
pre-existing cards on the PC. Windows leveraged itself in by using DOS
drivers, and with volume managed to get a good degree of industry
support. OS/2 has been much less successful, placing more of the onus
of support on IBM. Not that IBM can't handle it, but they're not going
to do as complete a job, no matter what.

>The fact is that many PC hardware companies will not
>write drivers for new Operating Systems (not that they
>always do a great job writing drivers for the OS they do
>support!).  This is an expense that Microsoft has been
>able to share with the 3rd party developers, but for
>a new OS like OS/2, it took a long time to build up a
>large library of mature drivers.

>This is the big problem for a company like Commodore or
>Apple.  If they do adopt industry standards, drivers may
>still cost a lot to develop in-house.

Sure they do. But there's no immediate need to support everything that
exists. Apple ships their PCI based machines with an ATi graphics
card, and a driver for it. They don't expect 100 different PCI cards
to run in the Mac, they don't really need to support 100 different
cards in order to be successful -- their initial reason for using
standard parts is volume pricing, not end-user variety. Some 3rd party
companies, such as Number Nine, look at the PCI-Mac market as place to
sell high-end card, and their Mac driver is a cheap price of
entry. Others totally ignore the Mac, but then again, unless they
offer some advantage, there's little for them or Apple to gain from
support. Again, you're creating a new market with existing parts, not
dropping into a fully formed existing market.

>Say for example Commodore or Apple where to write drivers for a dozen
>or so 3rd party video cards.  These are included in the OS.

Where is the advantage of doing this? Every machine, obviously, needs
a graphics card. If Apple's sales model requires bundling a card with
every system, they only need to supply drivers for the cards they
ship. If they ship unbundled, they can offer a small number of
supported cards, they don't need to tackle every card there is. Some
third parties will support their own cards, just like Radius and now
Number Nine on the Mac, Phase V and MacroSystems on the Amiga. The
only difference is that the hardware is fairly standard, the 3rd party
company needs only concentrate on the software (which they always have
done, anyway).

This is exactly what Be, Inc. is doing today. They introduced
developer machines with support for one specific card, the Number Nine
GXE64. Clearly, as long as they're selling unbundled systems, they
should offer support for a few choices -- people will appreciate a
choice of performance versus economy (and, of course, they need to
support a card that's still in production, the GXE64 is no longer
made).

Some of this will come together, on PCI at least, with Open
Firmware. Using this system, a 3rd party manufacturer writes one
device driver for non-Intel systems, period. Such drivers are OS and
CPU independent, an increasing concern as more OSs live on more
different kinds of RISC-based computers, all supporting the PCI
bus. Apple's sloth in moving to a standard PowerPC system has retarded
this to some extent, but it's coming.

>Apple is not so much unlike Commodore really.

And unfortunately for Apple, and perhaps the industry as a whole,
looking more like them everyday.

>Apple sells you a system, and that may be it.  Not another dollar
>ever from that customer.

That is the hardware-company model. Maybe you buy an upgrade OS from
an Apple or C=, maybe a board (but that's going away, thanks to the
end of proprietary expansion buses), that's it. IBM, Compaq, and the
rest of them never see you again, at least until it's time for the
next system buy.

It's little wonder Microsoft intentionally kept out of the system
hardware business...

Dave Haynie          | ex-Commodore Engineering |   for DiskSalv 3 &
Sr. Systems Engineer |  Hardwired Media Company | "The Deathbed Vigil"
Scala Inc., US R&D   |    Ki No Kawa Aikido     |     [e-mail]

         "Feeling ... Pretty ... Psyched" -R.E.M.