COME DOWN HERE FELLAS!

[JimC]

This should not be a difficult issue to reach agreement on:

Before and after upgrades from Win 98/ME to XP, many things
remain unchanged:
(1) the user.
(2) the CPU.
(3) the device in question.
(4) the hardware interface for the device.
(5) the O/S vendor.

Now it may be (as is often the case) that the device and
hardware interface offer a superset of functions and control
facilities in comparison to what gets implemented in any
given driver. In a sense, the driver defines the device.
In the mainframe era, it was not uncommon for users to order
a hardware upgrade for rented devices and then feel chagrin
when nothing changed except the monthly bill and the
installed drivers. For most PC devices. the user's
performance expectations will mirror whatever limitations
were imposed by the Win 98/ME driver.

Given these circumstances, how can you make a case for
blaming anyone but Microsoft when a device stops performing
as it did under Win 98/ME? The ONLY variable in this
problem is the Win XP software supplied by Microsoft.

Microsoft could have avoided alienating users by offering a
sensible installation/deployment scheme and friendly
marketing terms. There would be few complaints if users
could upgrade to XP on a "Try It And See" basis. Users who
are capable of resolving the problems posted here would do
so. Others would simply return XP and receive their refund.

I am totally pleased with my new XP system although it was
difficult to convince a major OEM vendor to sell me a nice
big tower system that was bare bones except for memory, HD,
FD, DVD, and the other stuff that has migrated to the
motherboard. As I installed devices from my Win 98 system,
XP never even blinked. In my case, every driver I needed
was included with my OEM version of Win XP PRO although none
of my devices came from the manufacturer of my new system.

The devices brought over from my Win 98 system included an
HP LaserJet, Creative sound card, Sony 19" monitor, TDK R/W
CD, Seagate SCSI TBU, 3 SCSI ZIPs, LinkSys NIC, Motorola
cable modem, MS keyboard, and MS mouse. All I had to leave
behind were my video card and my SCSI host controller. All
I had to buy was a new $70 Adaptek card.

I admit cheating by checking out the driver issues
beforehand, but it is precisely the success that I enjoyed
that induces me to sympathize with all of the
problem-plagued users that post here. Most of them didn't
buy their first PC before hard disks were introduced, and I
think they deserve a refund from Microsoft if they don't
want to cope with issues that Microsoft chooses not to
solve.

Jim
(-: a Seattle Computing copyright statement was probably
still embedded in my first DOS :-)

[Donald L McDaniel]

In article >, says...
> This should not be a difficult issue to reach agreement on:
>
> Before and after upgrades from Win 98/ME to XP, many things
> remain unchanged:
> (1) the user.
> (2) the CPU.
> (3) the device in question.
> (4) the hardware interface for the device.
> (5) the O/S vendor.
>
> Now it may be (as is often the case) that the device and
> hardware interface offer a superset of functions and control
> facilities in comparison to what gets implemented in any
> given driver. In a sense, the driver defines the device.
> In the mainframe era, it was not uncommon for users to order
> a hardware upgrade for rented devices and then feel chagrin
> when nothing changed except the monthly bill and the
> installed drivers. For most PC devices. the user's
> performance expectations will mirror whatever limitations
> were imposed by the Win 98/ME driver.
>
> Given these circumstances, how can you make a case for
> blaming anyone but Microsoft when a device stops performing
> as it did under Win 98/ME? The ONLY variable in this
> problem is the Win XP software supplied by Microsoft.
>
> Microsoft could have avoided alienating users by offering a
> sensible installation/deployment scheme and friendly
> marketing terms. There would be few complaints if users
> could upgrade to XP on a "Try It And See" basis. Users who
> are capable of resolving the problems posted here would do
> so. Others would simply return XP and receive their refund.
>
> I am totally pleased with my new XP system although it was
> difficult to convince a major OEM vendor to sell me a nice
> big tower system that was bare bones except for memory, HD,
> FD, DVD, and the other stuff that has migrated to the
> motherboard. As I installed devices from my Win 98 system,
> XP never even blinked. In my case, every driver I needed
> was included with my OEM version of Win XP PRO although none
> of my devices came from the manufacturer of my new system.
>
> The devices brought over from my Win 98 system included an
> HP LaserJet, Creative sound card, Sony 19" monitor, TDK R/W
> CD, Seagate SCSI TBU, 3 SCSI ZIPs, LinkSys NIC, Motorola
> cable modem, MS keyboard, and MS mouse. All I had to leave
> behind were my video card and my SCSI host controller. All
> I had to buy was a new $70 Adaptek card.
>
> I admit cheating by checking out the driver issues
I still dont understand why you expect Microsoft to be the patsy when it
comes to device drivers...

Don't you understand that when a new OS is released, memory addresses change?
The memory addresses in a certain device's chips are hard-wired to address
memory locations in an OS which exist only as long as the computer is turned
on. This is the reason device drivers must be upgraded with each OS
iteration. Again, Microsoft is responsible for the OS, while the device
manufacturers are responsible to write device drivers which interface to the
OS. These drivers MUST meet the memory requirements of the OS. Im sure
Microsoft tries to be as generic as possible, but, again, they can't write
code which perfectly connects to each of the 10's of thousands of devices
which are produced by 1000's of manufacturers. IF they tried to do that,
they would never get an OS out of the shop and onto the the shelves, on top
of which doing that would require millions of lines of extra code, making the
OS extrememly unwieldy and HUGE. So huge, in fact, that the OS would require
EXTEMELY HUGE amounts of memory and disk space to run in. Not to speak of
the extra money which would be required to pay a HUGE staff of programmers
who would write device drivers exclusively. I'm sure this would make the
price of the resulting OS so HIGH, only major corporations could afford to
buy it.

--
Donald L McDaniel
Post All replies to the Newsgroup, so that all may be informed
================================================== =================

[JimC]

"Donald L McDaniel" > wrote in
message
m....
>
> I still dont understand why you expect Microsoft to be the
patsy when it
> comes to device drivers...
>

Because the current patsies are the unhappy users who post
their problems in this newsgroup.

>
> Don't you understand that when a new OS is released,
memory addresses change?
> The memory addresses in a certain device's chips are
hard-wired to address
> memory locations in an OS which exist only as long as the
computer is turned
> on. This is the reason device drivers must be upgraded
with each OS
> iteration. Again, Microsoft is responsible for the OS,
while the device
> manufacturers are responsible to write device drivers
which interface to the
> OS. These drivers MUST meet the memory requirements of
the OS. Im sure
> Microsoft tries to be as generic as possible, but, again,
they can't write
> code which perfectly connects to each of the 10's of
thousands of devices
> which are produced by 1000's of manufacturers. IF they
tried to do that,
> they would never get an OS out of the shop and onto the
the shelves, on top
> of which doing that would require millions of lines of
extra code, making the
> OS extrememly unwieldy and HUGE. So huge, in fact, that
the OS would require
> EXTEMELY HUGE amounts of memory and disk space to run in.
Not to speak of
> the extra money which would be required to pay a HUGE
staff of programmers
> who would write device drivers exclusively. I'm sure this
would make the
> price of the resulting OS so HIGH, only major corporations
could afford to
> buy it.

Where in the world did you get these crazy notions about
"hard-wiring" and the like?

On various CPUs, specific memory regions may be reserved
because they are used to communicate between the CPU and
software, but such "hardware reserved memory" is defined by
the CPU, and not by any O/S. These regions may include
firmware that loads during a power-on cycle, but again,
there is no relationship to any O/S.

Beyond this, what you are calling "hard-wired" addresses are
not hard at all, they are simply software conventions.

Your references to "10's of thousands of devices" and
"1000's of manufacturers" creating a driver dilemma wherein
the O/S would require "EXTEMELY HUGE amounts of memory and
disk space to run in," sounds like:
(a) What we already have today with inadequate driver
support, and
(b) You watched too many appearances by Carl Sagan on the
Johnny Carson Show.

In reality, the big universe of device models and brands is
based on a much smaller population of core technologies, and
the problem is not as daunting as it may appear.

[Donald L McDaniel]

In article >, says...
> "Donald L McDaniel" > wrote in
> message
> m....
> >
> > I still dont understand why you expect Microsoft to be the
> patsy when it
> > comes to device drivers...
> >
>
> Because the current patsies are the unhappy users who post
> their problems in this newsgroup.
>
> >
> > Don't you understand that when a new OS is released,
> memory addresses change?
> > The memory addresses in a certain device's chips are
> hard-wired to address
> > memory locations in an OS which exist only as long as the
> computer is turned
> > on. This is the reason device drivers must be upgraded
> with each OS
> > iteration. Again, Microsoft is responsible for the OS,
> while the device
> > manufacturers are responsible to write device drivers
> which interface to the
> > OS. These drivers MUST meet the memory requirements of
> the OS. Im sure
> > Microsoft tries to be as generic as possible, but, again,
> they can't write
> > code which perfectly connects to each of the 10's of
> thousands of devices
> > which are produced by 1000's of manufacturers. IF they
> tried to do that,
> > they would never get an OS out of the shop and onto the
> the shelves, on top
> > of which doing that would require millions of lines of
> extra code, making the
> > OS extrememly unwieldy and HUGE. So huge, in fact, that
> the OS would require
> > EXTEMELY HUGE amounts of memory and disk space to run in.
> Not to speak of
> > the extra money which would be required to pay a HUGE
> staff of programmers
> > who would write device drivers exclusively. I'm sure this
> would make the
> > price of the resulting OS so HIGH, only major corporations
> could afford to
> > buy it.
>
> Where in the world did you get these crazy notions about
The fact remains, my friend, that when a new OS is brought in, memory
addresses change. This is one reason why new device drivers have to be
written. Addresses contained within a device driver which formerly worked
with one OS no longer work with a new one.

--
Donald L McDaniel
Post All replies to the Newsgroup, so that all may be informed
================================================== =================