I asked for a quote for another new PC today and while going over the
spec the salesman told me that teh PC was PC97 compliant.

I asked him what he was talking about and he said that MS have started a
PC certification system under which they will not support any software
on any system which is not certified by them.

Oh dear, it seems that Billy is trying it on.

I wonder if a part of the PC97 definition will be the absence of non-MS
OSes.

Jason
--
Please note that my return email address is jason.cliff...@domgen.com

As a service I provide analysis of viruses and poor grammar to senders
of unsolicited commercial e-mail at a rate of $500.00 per hour. Delivery
of said correspondence constitutes a request for the aforementioned
services at said price. Supply billing address.

: I wonder if a part of the PC97 definition will be the absence of non-MS
: OSes.

why not find out before posing this kind of ignorant crap?

Does that really matter ?
Even today we rarely need to use any M$ products.
Soon, we will never need those.
So, all that M$ support or sertification will be only for old donkeys
who will still use buggy Widdoze and co - and never for us.

  Jason> I asked for a quote for another new PC today and while going
  Jason> over the spec the salesman told me that teh PC was PC97
  Jason> compliant.

  Jason> I asked him what he was talking about and he said that MS
  Jason> have started a PC certification system under which they will
  Jason> not support any software on any system which is not certified
  Jason> by them.

  Jason> Oh dear, it seems that Billy is trying it on.

  Jason> I wonder if a part of the PC97 definition will be the absence
  Jason> of non-MS OSes.

Remember that there was a similar system for Windows 95 -- seems
benign to me.

Yes and the rest of the world who do use MS products because those who
sold them their hardware made a big point of the fact that their systems
are designed to comply to MS 'standards' will avoid OSes like Linux,
OS/2, FreeBSD etc because they will believe the MS marketing which will,
no doubt, tell them that the most powerful and efficient software for
their hardware is MS.

Given the effectiveness of the MS marketing department it matters a
great deal.

Jason
--
Please note that my return email address is jason.cliff...@domgen.com

As a service I provide analysis of viruses and poor grammar to senders
of unsolicited commercial e-mail at a rate of $500.00 per hour. Delivery
of said correspondence constitutes a request for the aforementioned
services at said price. Supply billing address.

Obviously the use of irony was a bit more than you could deal with.

The reason for my posting was that this kind of move by MS could do real
damage to the attempts being made to introduce Linux to the wider
market. If you cannot work out why this is so you have completely missed
the significance of the fact that MS has one of the most effective
marketing teams there is.

Jason
--
Please note that my return email address is jason.cliff...@domgen.com

As a service I provide analysis of viruses and poor grammar to senders
of unsolicited commercial e-mail at a rate of $500.00 per hour. Delivery
of said correspondence constitutes a request for the aforementioned
services at said price. Supply billing address.

The PC97 spec is a multi-vendor hardware-based initiative designed to reduce
the complexity and number of options that need to be supported in today's
PC's. Most of the code-bloat in MS OS products (Win95 checks in at over 15
million lines of code) is due to the "need" to not only support thousands of
devices that have been manufactured by dozens of vendors over the last five
years or so, but also to do so in a way that does not require manual
intervention, that is, via "plug-n-play". See

        http://www.zdnet.com/pcweek/news/0715/15winpc.html

for an (old) review of the spec in Computer Shopper magazine, or search the
MS WWWeb site for more details (the spec is several hundred pages in size,
or I'd have stuck it in here for you!).

Let me say that I am most emphatically *not* a MS supporter, yet as a
software engineer, I can't help but feel for them at times. I have done
operating systems design work, and understand some of what they face in
trying to support everything. They tried to simplify the code base, reduce
complexity, and increase stability and security in Windows NT ("only" about
5 million lines of code), and look at all the criticism they get when a
driver for such-and-such is not available.

For all of the "advancements" that have been made in microcomputer
technology in the last 25 years, especially since 1981 when IBM introduced
its original PC, the changes in the underlying architecture have been
minimal. Sure, we have processors that provide advanced memory management
features, but that still (for backward compatibility) segment RAM. The CPU
runs at insane clock speeds (the "revolutionary" IBM PC in 1981 ran at a
mind-numbing 4.77 MHz; as late as 1987, there was not a production
microprocessor on the planet clocking faster than 20MHz), but there are
still only 4 COMM ports and 2 LPT ports available. Networking? ISDN and
cable modems? 1024x768 at 16 million colors? These were all unimaginable in
the late '70's, yet all of them (and much more) have been shoehorned into or
tacked onto the original PC specifications. Know why you have so much
trouble getting all of your IRQ's straight? There are still only 15 IRQ's
available, and that number was achieved only when the AT architecture
piggy-backed two interrupt controllers onto each other (read your
documentation carefully; ever wonder why they say that IRQ's 9 and 2 are
equivalent? That's where the "bridge" occurs). The BIOS, keyboard
controller, mouse, PCI controller, and various other system devices have
already taken most of them before you even open the box and try to connect a
modem, sound card, network card, and who knows what else. Intel, MS, and
others are so paranoid about upsetting the boat that has kept them so
profitably afloat for the last 15 years that they are scared stiff to change
anything about the underlying design, even to jettison archaic baggage that
no longer serves any useful purpose (the 15-channel, piggybacked IRQ
controller, the one-byte serial UART, and the segmented memory model -- in
the world of 32-bit virtual-memory operating systems -- come immediately to
mind).

The point of this diatribe is that, as a programmer and occasional tech
support specialist, I am in favor of any effort that will reduce the
complexity of these machines and make them easier to use. Again, I'm not a
MS fan, but the PC97 spec is not theirs alone. I'm also not particularly a
MAC fan, so please don't write back to me to tell me how you never have any
trouble with your trusty SE or IIcx. I'm just a guy who is getting tired
trying to keep sticking new technology onto an aging platform. I understand
the marketing forces behind the concepts of an "installed base" and
"backward compatibility", but I feel that the time has come to streamline
things at least a bit. If we have to "cut our losses" here and there, so be
it. I work with clients all the time who are reluctant to abandon stuff that
has been working for them for years, but who want the advanced features of
new technology; it's my job to find and implement a migration path that
allows them to get what they want.

MS is a hog, and uses monopolistic practices to control the market. However,
even that does not automatically qualify any initiative they propose as
"evil". The PC97 spec will make it easier for MS to develop OS products and
support its customers. It will also mean that consumers will have much
better odds of getting a system that "works", especially if they upgrade
after purchase. And yes, it will mean that some products from some vendors
(maybe even your favorite graphics card or RAM manufacturer) are not
"certified" and will probably disappear from the market. Hey, read the
Windows 95 box: already, certain "perfectly good" computer systems are
excluded by virtue of the minimum processor, RAM, and HD specs. MAC? SPARC?
Amiga? Don't even ask. MS already *has* minimum "certification" specs; PC97
just tightens things up some more. Want a system that doesn't meet the PC97
spec? Go ahead; you can always choose OS/2, Solaris x86, or Linux; they all
have both excellent "native" applications as well as passable Windows
emulators.

In the '70's, if you had a working computer system and someone introduced a
new feature that you "had to have", you trashed your existing system
(hardware, software, the works) and got the new one. Every time. Just like
the old days with IBM mainframes; nothing was compatible, because everyone
was innovative and invented their own solutions to users' problems.
Nowadays, users expect to buy a computer from one place, a monitor somewhere
else, and later add a new sound card and extra RAM, maybe a bigger hard
drive and a faster CD-ROM. Then they want to plug it all in and have it
automatically work. Who cares about IRQ's? I don't want to have to fool with
all those floppy disks; why do I need  "driver" anyway? That's the lure of
"compatibility", and since the burden has fallen on MS to provide that
functionality, they have proposed a specification that would simplify the
job of getting that jumble of plastic, copper, and silicon to work.

As to your final comment, I'm not aware of anything in the PC97 proposal
that specifies an operating system. The idea is that the OS vendor(s) will
have the option of saying that the OS was designed to run on any system that
meets or exceeds the PC97 specs, and that the vendor will not promise any
results or provide support if you choose to attempt to run the OS on a
non-compliant machine. Hardware vendors can still make non-compliant
components, any OS vendor can opt to build a product that will (or at least
might) support non-compliant machines, and you can still have a choice of
OS, even if all hardware becomes compliant and all OS vendors restrict
support to only compliant machines.

As a counter-example, during its warranty period, I took my Ford mini-van to
the Ford dealer to have Ford-certified mechanics install Ford brand-name
parts. These days, I do most routine maintenance myself, using parts that I
pick up at the local auto parts store. The manufacturers of these parts
claim that they will work in my Ford van, but Ford does not "certify" Fram
filters or the brake pads that Sears installed; Ford would have voided my
warranty had I done that when it was new, because they could not guarantee
support for my vehicle since I had installed parts that they had not
certified themselves. Most people would agree that it is prudent to honor
that system; why is MS' attempt to do the same thing seen as such a crime?

Lucien
--

+----------------------------------------------------------+
|    If the purpose of feet is to walk and run, and the    |
|   purpose of painting is to make signs, then ballet and  |
|              Rembrandt are equally useless.              |
|    All art springs from the non-obvious use of tools.    |
+----------------------------------------------------------+

While we may roll our eyes and say that we think Microsnot is trying to keep
their greedy hands on their market share, I think that we have to remember that
even those of us who use "non-standard" OSs are more than happy to have
standards for C, C++ (maybe someday:), CD-ROM formats, and data transmission,
just to name a few.  Standards provide a way for many different companies to
provide hardware that runs "as-advertised" on many different machines with
varying configurations.  They also make sure we can write portable code, listen
to music, have bumpers which bump the other bumper (a problem in the EU), get
the correct tools, make phone calls, and all of those other everyday things
that we take for granted.  Our computer world, however, is changing so fast
that by the time we get one standard, it seems that the latest and greatest is
something else.  Wouldn't it be great if everything was PCI?  Isn't it nice
that your motherboard still has ISA slots?  Wouldn't it be great if Intel or
Alpha or Motorola completely redesigned their processor/bus architecture
instead of applying patch after patch?  Isn't it nice that you can just buy a
new motherboard/processor/add-on card to take advantage of the latest and
greatest?

At the same time we moan and complain that the IEEE/ISO standards commitees are
too slow, we grudgingly put up with this process because in the end, we expect
that the standards will increase our options, expand our capabilities, and make
the routine maintenence easier for us.  If we really want Microsnot to actually
do something that benefits all users, as I expect PC97 and any subsequent
standards will do, we need to help Microsnot to understand that they are not
the new Lords of Everything.  If we support whatever Microsnot does not and not
recognize something which can be beneficial not only to Windoze users, but to
ourselves as well, we make the mistake that too many companies have, and risk
becoming like Apple - easy prey to Microsnot's tactics.  Why not certify
vendors as Linux compliant if they help us to develop the OS by providing the
information that we need to write good code?  I don't know who all of these
vendors might be, but I do know that Matrox has been very helpful to the
XFree86 project.  Then, when you call up a vendor, you can ask them if their
products are certified to be Linux compliant.  If vendors respond the way they
are taught and trained to in marketing school, when they hear that question a
few times, they will start responding by finding out more about it.  We are and
will be the people who make purchasing decisions on technologies for major
corporations, or we are the ones who make decisions about the hardware and
software that will be produced.  We can't be snotty like Microsnot can and flip
Sun or Dell the bird if they don't agree 100% with us.  We may need to make
compromises in order to get vendors on our side, but we can do it.

The m{*filter*}of this story is that we don't have to beat Microsoft at their game,
and we don't have to join them by using their standards.  TCP/IP came from the
Unix world, after all, and it provides the "wrapper" of choice for the newest
technologies.  Maybe that will change someday, but do you see Microsnot using
anything else?  Evaluate the ideas on their own merits, not by who supports
it.  If they are not doing it in an open and fair manner (see I20 standard), we
need to let them know.  Every time we ignore or deride what Microsnot does,
that is money in Bill Gates' pocket.  By the way, have fun, too;).

-Chris

--
"So far as the laws of mathematics refer to reality, they are not certain.
And so far as they are certain, they do not refer to reality."
-- Albert Einstein, Geometry and Experience
***send mail to: mcki0...@itlabs.umn.edu***

This doesn't really answer the issue I raised as it isn't the
manufacturers that new users speak to - it is the resellers who often
have little real experience (the PC sales company I called are quite big
in the UK - TINY COMPUTERS - and their sales people were insisting that
the PC had to be PC97 compliant because MS is the only option.) so
getting people to put a 'Cool - it works with Linux' or other such logo
onto their kit wont address the problem which is companies who sell
systems where the sales staff and the customers do not see anything but
the made up system.

I think listing ket as Linux complaint and getting manufacturers to
display a Linux compatible logo is a great idea and it is certainly
necessary if Linux is to gain widespread acceptance but if the PC is
bought from a company that says that only an MS OS will run on it Linux
will never even be looked at by the customer.

Jason
--
Please note that my return email address is jason.cliff...@domgen.com

As a service I provide analysis of viruses and poor grammar to senders
of unsolicited commercial e-mail at a rate of $500.00 per hour. Delivery
of said correspondence constitutes a request for the aforementioned
services at said price. Supply billing address.