Failure of keyboard and mouse on a Compaq DeskPro EN.

In a Compaq DeskPro EN, 350 MHz, I've replaced the
hdd, partitioned it, run the latest ROMPaq diskette,
installed DOS 5.0 on the first partition and installed
Native Oberon on the second partition. Nothing in
any of this appears abnormal. The keyboard and mouse
work properly for the installations. The installed
DOS works with no problem.

When Oberon runs from the hdd, the display is normal
except for absence of the mouse pointer. The keyboard
and mouse do not work.

Any ideas about this?

Thanks, ... Peter E.
peasthope .at. shaw .dot. ca

Peter wrote:
In a Compaq DeskPro EN, 350 MHz, I've replaced the
hdd, partitioned it, run the latest ROMPaq diskette,
installed DOS 5.0 on the first partition and installed
Native Oberon on the second partition. Nothing in
any of this appears abnormal. The keyboard and mouse
work properly for the installations. The installed
DOS works with no problem.

When Oberon runs from the hdd, the display is normal
except for absence of the mouse pointer. The keyboard
and mouse do not work.

Any ideas about this?

Thanks, ... Peter E.
peasthope .at. shaw .dot. ca


Knowing nothing about Native Oberon except for what I read just now on
the web, I would have to guess that the DeskPro EN has a keyboard-mouse
controller chip that is incompatible with the Native Oberon.

Otherwise, I would suggest that you direct an inquiry through the Native
Oberon web site. But don't expect a quick response or a quick solution.
The last version of the software was "Latest PC Native Release (2.3.6)
[next release early 2003]", so the project appears to be somewhat
dormant... Ben Myers

Ben & others,

On Feb 17, 8:09*pm, Ben Myers wrote:
... I would have to guess that the DeskPro EN has a keyboard-mouse
controller chip that is incompatible with the Native Oberon.

But the NO installer diskette with Oberon0 works
properly. The keyboard and mouse code it installs
to the hdd is exactly the same as the code it runs.

... [next release early 2003]", so the project appears to be somewhat
dormant...

NO evolved to A2.

I noticed a similar keyboard problem with Linux on
a DeskPro EN was reported in another group. The
user happened to find that disabling the onboard
USB let the keyboard work. This suggests a bug in
the Compaq ROM code. I had hoped that someone
would know more about that. If a bug was found,
the repair didn't necessarily propagate into all
affected SoftPaqs.

This weekend I can try shutting off USB. Also
might install A2 and find whether the failure is
there also.

Thanks for the comments, ... Peter E.

Peter wrote:
Ben & others,

On Feb 17, 8:09 pm, Ben Myers wrote:
... I would have to guess that the DeskPro EN has a keyboard-mouse
controller chip that is incompatible with the Native Oberon.

But the NO installer diskette with Oberon0 works
properly. The keyboard and mouse code it installs
to the hdd is exactly the same as the code it runs.

... [next release early 2003]", so the project appears to be somewhat
dormant...

NO evolved to A2.

I noticed a similar keyboard problem with Linux on
a DeskPro EN was reported in another group. The
user happened to find that disabling the onboard
USB let the keyboard work. This suggests a bug in
the Compaq ROM code. I had hoped that someone
would know more about that. If a bug was found,
the repair didn't necessarily propagate into all
affected SoftPaqs.

This weekend I can try shutting off USB. Also
might install A2 and find whether the failure is
there also.

Thanks for the comments, ... Peter E.

Another thought: Can you try a USB keyboard and a USB mouse instead?

.... Ben Myers

Hi!

But the NO installer diskette with Oberon0 works properly. The keyboard
and
mouse code it installs to the hdd is exactly the same as the code it runs.

Does it run a keyboard and mouse driver at setup time? I think it would have
to run a mouse driver, but it could rely on the BIOS to do the heavy lifting
for the keyboard related tasks. Some keyboard drivers are naughty (or poorly
written if you prefer) and attempt to "know better" than the system's own
BIOS by changing things such as the default scan code set. Sometimes that
doesn't go over too well.

The I/O controllers in these systems are highly integrated and packaged into
a component known as a Low Pin Count I/O controller. Compaq appears to have
used an SMSC part, although I think they customized it somewhat to add
functions for the case lock (if equipped) and "tattler switch" that
indicates case intrusion.

There will be emulation code in the BIOS to let a USB mouse and keyboard act
like they were connected to the PS/2 port. I think this emulation can be
turned off via a setting in the BIOS setup utility, but I'm not sure of
that. The USB ports can be disabled.

I have not examined the Compaq ROM for these systems too deeply, but from
what I've seen it appears to be well designed. These are some of the last
systems to use a "real" Compaq developed BIOS.

William

Well, the problem turned out to be mostly my lack
of perception.

Ben Myers wrote,
"Can you try a USB keyboard and a USB mouse instead?"

William R. Walsh wrote,
Does it run a keyboard and mouse driver at setup time?

I would have to study to see how the keyboard works in
setup. The mouse is not used until later.

In trying a USB keyboard and mouse I noticed that
Oberon.Text invoked DayTime.SetTime. Without
a network connection, there was no progress. With
that command commented, startup completed.
No problem with the keyboard and mouse.

I can't say that the installation is finished! I proceeded
to connect a CF card as slave on the primary IDE
and found nothing but problems. Now there is a
report that the first part of the master drive overlaps
the second. Next weekend I can fdisk the drive
again and reinstall the systems. Damage to the
part table is odd.

Thanks for the ideas, ... Peter E.

Hi!

I would have to study to see how the keyboard works in
setup. *The mouse is not used until later.

It wouldn't surprise me if they simply used the BIOS at first and
later loaded a driver to handle both mouse and keyboard. I've never
run Native Oberon, or programmed in it, so I can't say for sure.

It seems odd that the failure of a network based time/date
synchronization process to complete would not generate some kind of
error. (Again: maybe it does...I still haven't run Native Oberon to
see what it's about.)

I can't say that the installation is finished! *I proceeded to connect
a CF card as slave on the primary IDE and found nothing but
problems.

Is there a master device present? It's been my experience that the
later model Deskpro systems will tolerate a device configured only as
a slave on an IDE channel, but that's not "the right way" to do it.

Using CF cards over PATA/IDE can be a whole other can of worms. I've
been down that road with an IBM NetVista 8364. That system actually
has a CF card slot on the motherboard. It's intended for use only as a
means to back up the BIOS and CMOS settings, but a compatible CF card
plugged into it will show up as a primary master IDE device. It is
bootable. The idea was to use the CF card as a boot device and have a
mostly silent, low power usage and low profile computer.

I found that most cards worked fine, but a Kingston 64MB card did not.

(and if for any reason you'd be interested in seeing this setup:
http://greyghost.mooo.com/netvista98/ has pictures)

William