Where can I get replacement xenix software
Chris Weaver
xenix/unix software that also included all manuals etc.
Yesterday I finally tried to install the xenix OS but got
an *E* as soon as I entered at the boot with the 1st disk.
According to the manual this means the disk is corrupted
and that I should request a replacement:)
All the software is on 5.25" disks, which I was going to
backup to 1.44 floppies, however I can't install as the
first disk, and possibly others, are corrupted etc
I have never used unix before but have installed and
minimally used linux for basic internet etc, I will shortly
have a spare 486/DX66, 8Mb RAM, which I was hoping to install
the unix on for a little pleasurable self education :)
The only software I have tried to install so far was the
UNISYS XENIX 386. (On a 486/DX100)
How useful are all the other programs I have listed?
I have had a bit of a look around the internet but have not
been able to locate any replacement sources for this
software yet.
Can I get replacements, preferable in 1.44 format, for these?
Here is a list of the software I have -
10 Disk Set -
UNISYS
Title: XENIX 386 OPERATING SYSTEM
CLASS: B
Style: XNX-386-OS
Level: 2.3.2 GT
Dated: 7/14/89
Disks 1-10, being -
3 x Installation (1st disk corrupted, maybe others?)
2 x Basic Utilities
4 x Extended Utilities
1 x Games
3 Disk Set -
Information Technology International
DBQ
Op. System: XENIX 386
Version : 1.03.02
Issued On : 02/03/90
2 Disk Set -
SCO Professional
Type: n386
Release: 2.0
Media: 96dshd
Volumes 1 & 2
2 Disk Set -
SCO XENIX Tutor
Release: 2.0
Type: dosn86
Volume 1 - Master Disk, and,
Volume 2 - Lesson Disk
4 Disk set -
SCO CGI
Type: n286
Release: 1.1.0d
Media: 96dshd
Volumes 1-4
All help greatly appreciated,
Thanks,
Chris Weaver.
Stephen M. Dunn
$Some time ago I bought a collection of second-hand
$xenix/unix software that also included all manuals etc.
$Yesterday I finally tried to install the xenix OS but got
$an *E* as soon as I entered at the boot with the 1st disk.
$According to the manual this means the disk is corrupted
$and that I should request a replacement:)
Not necessarily - back in the old days (haven't installed Xenix
in many years, due to its antiquity :-) I recall having seen this with
known good Xenix installation diskettes, and it turned out to be a
weird hardware compatibility issue. Since it's so long ago, I don't
recall the cure, but it may have involved swapping floppy drives and/or
controllers, adjusting CPU speed, disabling cache and/or shadow
RAM, stuff like that.
Keep in mind that since the most recent version of Xenix is over
a decade old and was released when a 486/33 or thereabouts was the
fastest PC around, its support of more modern hardware is
lacking. Some fast 486es (e.g. 486/100s) had compatibility issues with
Xenix, and many people had problems getting it to work on
Pentium systems. If you're trying to get it to run on something
more modern than that, you're probably out of luck.
--
Stephen M. Dunn <ste...@stevedunn.ca>
>>>----------------> http://www.stevedunn.ca/ <----------------<<<
------------------------------------------------------------------
Say hi to my cat -- http://www.stevedunn.ca/photos/toby/
Scott McMillan
Well, Stephen Dunn has already noted a number of the problems you are
likely to encounter, but a read through
http://stage.caldera.com/cgi-bin/ssl_reference?104151 may give you
some ideas of the headaches you are attempting to unleash upon
yourself.
<Stephen wrote>
Keep in mind that since the most recent version of Xenix is over
a decade old and was released when a 486/33 or thereabouts was the
fastest PC around, its support of more modern hardware is
lacking.
<end>
I don't believe the above is entirely accurate - SCO claimed that
Xenix 2.3.4 would run on a 486/33, but was not stable on speeds
greater than that. I'm pretty sure - but have been wrong many times
before - that 2.3.4 was released well before the 486/33 was released.
I only bring this up to give you an idea of how very ancient it is.
We've (the company I work for) seen Xenix 2.3.4 run on Pentium
machines - nothing faster than a 120Mhz, if I recall correctly. That
was due to someone placing the existing drive from a 386/486 system
into a Pentium box and being lucky (??) enough to have it boot up. It
seemed to run just fine, much to our surprise.
Then we attempted to install the Y2K patch to the system (xnx427, I
believe) and relink the kernel. Say 'bye-bye' to system: We had
directories disappear, regular files appear to turn into directories,
vice-versa, etc. REALLY ugly. We do not know if the relink or the
application of the patch caused the problems, and really didn't care
as it was just a 'let's see how far we can push it' test.
There used to be a free-for-non-commercial-use version of SCO
Openserver available (Openserver 5.0.6 being the current release). I
don't know if that program is still around; I'm sure someone else here
will be able to provide that information. If it _is_ still available,
I would recommend playing with that. IMHO, Xenix is too far removed
from the current release to be of very much help to you.
Scott McMillan
James J
2.3.2, but it's a different release. The info you posted showed 2.3.2 GT,
which was a SCSI release. I've got 2.3.2 AT, which is a non-SCSI release.
Which brings up a point. I assume the SCSI release worked on non-SCSI
drives, but I don't know for sure. If 2.3.2 GT requires a SCSI hard
drive, you may have another problem, because Xenix will only recognize a
very limited set of very old SCSI adapters (eg. Adaptec 1542).
Also, if you are using an IDE drive, it will probably have to be small.
Xenix can't use more than 1 GB. It is possible to install Xenix on IDE
hard drives somewhat larger than that (and just have Xenix ignore the rest
of the hard drive), but I don't know what the maximum is. Any "modern"
hard drive may just be too big.
Somewhere in all those manuals, etc., do you have a card with a Xenix
serial number and activation key? The card would probably be pink. You
probably won't be able to finish the install without a serial number and
activation key. I say "probably" because they may only be required if
you're installing the link kit. I can't remember for sure. But even if
that's true, the link kit is required to make changes to the operating
system, like adding device drivers, so you'll probably need it.
The point is, if you don't have a serial number and activation key, there
may not be any point in trying to get a replacement floppy.
As for the rest of the software you listed, SCO Professional was a
spreadsheet program, as I recall. I don't know about the rest.
James
Chris Weaver
Thanks Stephen, Scott, and James, it appears that I
may have several problems here.
Firstly as Stephen said, being that trying to install on my
486DX/100 is looking for trouble as it is too advanced for
the XENIX 2.3.2 GT OS, sounds funny really ;-).
However I was using this PC as a test as it was the
only spare PC low spec'd enough at the moment.
The other 486 which is the PC I really want to install
on is actually a 433SX/D upgraded up to a 486DX/66, so
I can just "downgrade" this again and hopefully then it
isn't too advanced. I will have to wait a week or two
as this is my current firewall/router, and I am about to
upgrade/rearrange some PC's.
Secondly, as Scott points out, I may be better off trying
a newer version of Unix which I may do if I can't get the
"ancient one" to install, configure etc.
(I think sometimes I just like being a masochist!)
Also thanks for the URL, backs up Stephen's theory and more,
also ties in with James' thoughts on mine being for SCSI drives.
Thirdly as James says, if the GT version is a SCSI release it
probably isn't going to install on my IDE drive/s.
So, maybe I have good software and "dodgy" hardware :-)
Otherwise the HDD is under 1Gb (212 Mb), so size is OK.
As far as serials and activation keys go, well i am a lucky
chappy, I have them for the XENIX O/S, SCO Professional,
SCO Tutor and SCO CGI.
Actually it would be a pity not to try and get this going as
I have the manuals/user guides/admin guides for everything,
really I don't have an excuse ;-) (Except "dodgy" hardware!)
If I can't get this going then I wonder if I used the "AT version"
floppies instead of the 1-3 "GT version" installation floppies,
would the rest of the XENIX floppies 4-10 etc install properly?
I would assume only the first install floppy/ies would have to
be specifically for IDE or SCSI? As the remainder of the software
should be identical for either?
Would this assumption be correct?
Thanks,
Chris Weaver.
Tony Lawrence
> If I can't get this going then I wonder if I used the "AT version"
> floppies instead of the 1-3 "GT version" installation floppies,
> would the rest of the XENIX floppies 4-10 etc install properly?
AT version.. was that the one that worked on MicroChannel hardware?
I really do not remember.
You know, there is absolutely NOTHING there that has any educational
value. It's of historical interest, but that's all.
--
Tony Lawrence
Free SCO, Mac OS X and Linux Skills Tests:
http://aplawrence.com/skillstest.html
John DuBois
James J <SPAMMY...@netscape.net> wrote:
> I don't know if I can help with the bad floppy. I've got a copy of Xenix
>2.3.2, but it's a different release. The info you posted showed 2.3.2 GT,
>which was a SCSI release. I've got 2.3.2 AT, which is a non-SCSI release.
>
> Which brings up a point. I assume the SCSI release worked on non-SCSI
>drives, but I don't know for sure.
Yes, GT worked on ST506-interface (MFM/RLL) drives. It just added SCSI
support.
John
--
John DuBois spc...@armory.com KC6QKZ/AE http://www.armory.com/~spcecdt/
Bill Vermillion
>2.3.2, but it's a different release. The info you posted showed 2.3.2 GT,
>which was a SCSI release. I've got 2.3.2 AT, which is a non-SCSI release.
>
> Which brings up a point. I assume the SCSI release worked on non-SCSI
>drives, but I don't know for sure. If 2.3.2 GT requires a SCSI hard
>drive, you may have another problem, because Xenix will only recognize a
>very limited set of very old SCSI adapters (eg. Adaptec 1542).
> Also, if you are using an IDE drive, it will probably have to
>be small. Xenix can't use more than 1 GB.
I was never able to get it see more than 500MB. That was enough for
the client so I never checked to see if you could get more.
--
Bill Vermillion - bv @ wjv . com
Bill Vermillion
Tony Lawrence <to...@pcunix.com> wrote:
>Chris Weaver wrote:
>
>> If I can't get this going then I wonder if I used the "AT version"
>> floppies instead of the 1-3 "GT version" installation floppies,
>> would the rest of the XENIX floppies 4-10 etc install properly?
>AT version.. was that the one that worked on MicroChannel hardware?
>I really do not remember.
No - that is the MC version.
The label says 386MC. I'm reading that on an N1 floppy that I
used on my old Model 80 :-). I have 5 small gray SCO boxes with
that on it - all on 720K disks!
>You know, there is absolutely NOTHING there that has any educational
>value. It's of historical interest, but that's all.
You don't think seeing how thinks were so minimal in the past
compared to what we have today is educational :-)
Bill
Steve Wertz
<c_weaver@del_this_bit.telstra.com> wrote:
>I have had a bit of a look around the internet but have not
>been able to locate any replacement sources for this
>software yet.
>Can I get replacements, preferable in 1.44 format, for these?
At one time SCO Customer Service would send you replacement disks,
even for older unsupported prducts. You make a xerox of the disk and
they'll cut you a new one. Up until 2000 they still did this, at
least, and for Xenix disks. Be advised, as others mentioned, the
letter you're seeing may not mean it's a bad floppy and can be elusive
as the 'bad magic' error in later releases. You should technically be
able to dd the disk and run a checksum on it, if somebody from SCO
could provide that checksum (which I think they stored in a database
somewhere...)
-sw
Tony Lawrence
If it's just the boot disk,
http://www.aplawrence.com/SCOFAQ/scotec1.html#downloadboot
lists some Xenix imahes..
Scott McMillan
Chris,
If you're _really_ bent on getting this particular package running,
check Tony's download boot page posted earlier.
One question: Does 'a little pleasurable self education' include any
TCP/IP functionality? Xenix did not include TCP/IP as a standard - it
was a separate package (? I know it was separate in 3.2v4.2, but
included in ODT3). Masochist or not, that could certainly put the
halt on your tinkering.
The hardware you describe should handle the Xenix version you have.
GT included some embedded SCSI drivers, but would load on IDE as well.
Best of luck should you choose to press on!
Scott McMillan
Chris Weaver
> In article <RLs_9.106755$6G4.13491@sccrnsc02>,
> Tony Lawrence <to...@pcunix.com> wrote:
>
>>Chris Weaver wrote:
>>
>>
>>>If I can't get this going then I wonder if I used the "AT version"
>>>floppies instead of the 1-3 "GT version" installation floppies,
>>>would the rest of the XENIX floppies 4-10 etc install properly?
>
>>You know, there is absolutely NOTHING there that has any educational
>>value. It's of historical interest, but that's all.
>
>
> You don't think seeing how thinks were so minimal in the past
> compared to what we have today is educational :-)
Well I just sort of thought that if I was serious about learning
Linux/FreeBSD/Unix then I should probably see how it started,
besides this software etc has been looking at me for some time,
it's getting kind of hard to ignore :-)
So I suppose that would be educational and definately interesting!
Chris Weaver.
Chris Weaver
> If it's just the boot disk,
> http://www.aplawrence.com/SCOFAQ/scotec1.html#downloadboot
>
> lists some Xenix imahes..
Thanks Tony, I downloaded a few just in case :)
Chris Weaver.
Chris Weaver
<brave tale snipped :)>
> Chris,
>
> If you're _really_ bent on getting this particular package running,
> check Tony's download boot page posted earlier.
Thanks, I did that.
> One question: Does 'a little pleasurable self education' include any
> TCP/IP functionality? Xenix did not include TCP/IP as a standard - it
> was a separate package (? I know it was separate in 3.2v4.2, but
> included in ODT3). Masochist or not, that could certainly put the
> halt on your tinkering.
I got a supply of disks from Tony's URL, and all I have to do now is
wait until the old 486 is free. If I get this installed then I will
worry about networking, internet etc.
Probably worry about a lot of other things too :)
>
> The hardware you describe should handle the Xenix version you have.
> GT included some embedded SCSI drivers, but would load on IDE as well.
Yes, I think you are right, my first attempt was obviously thwarted by
the sheer speed of my 100Mhz cpu, good job it wasn't a P4 3.0G :)
>
> Best of luck should you choose to press on!
Thanks, I will accept any donations of good luck!
> Scott McMillan
Chris Weaver.
Tony Lawrence
Best of luck.
If by some miracle you get it working and would like to share the fun
with the rest of the world, see http://aplawrence.com/publish.html
Bill Vermillion
I spent years using Xenix. But your line about learning where it
started leads me to this comment.
The most popular Unix incantation on iNTEL processors for years
was SCO's Xenix - but that was not where it started, but it was one
of the first in iNTEL platforms, however in those days SCO
was alos a 'porting house' and they sold Xenix for other
systems too including Apple's Lisa.
>So I suppose that would be educational and definately interesting!
That it would be. I know how it about old SW just sitting there.
I keep meaning to bring my NeXTStep on iNTEL up again soon. I do
have to put it in a machine with older hardware as it requires
specficic SCSI contollers and specific video cards. The old Xenix
is a bit more tolerant.
James J
> In article <Xns9313EE3C83FD...@205.188.138.161>,
> James J <SPAMMY...@netscape.net> wrote:
> ...
>> Also, if you are using an IDE drive, it will probably have to be
>> small. Xenix can't use more than 1 GB.
>
> I was never able to get it see more than 500MB. That was enough for
> the client so I never checked to see if you could get more.
Xenix could see up to 1 GB (2048 cylinders x 16 heads x 63 sectors/track)
on IDE drives. Of course you had to remember to keep the root filesystem
within the first 1024 cylinders (about 500 MB). I'd set up the rest of
the 1 GB as a seperate filesystem.
I've actually installed Xenix on IDE drives somewhat larger than that. 2
GB. Maybe a bit larger. You just had to make sure the BIOS/drive wasn't
using LBA. I would also change the drive parameters during Xenix
installation to 2048 x 16 x 63, although I'm not 100 % sure that that was
necessary. The important thing was to make sure that Xenix stayed within
2048 cylinders (changing the drive parameters was an easy way to do that),
and make sure the root filesystem was within the first 1024 cylinders.
The rest of the disk (after the first 1 GB) was unused, at least by Xenix.
Technically, I think the limit in Xenix for IDE drives may have been 2047
cylinders or something like that. I seem to recall getting errors from
badtrk right at the end if I tried to use all 2048 cylinders. FWIW, I
would always have badtrk do at least a quick non-destructive scan on these
larger drives to make sure Xenix could read the whole area.
James
James J
news:3e3a50cb$0$231$5a62...@freenews.iinet.net.au:
> Firstly as Stephen said, being that trying to install on my
> 486DX/100 is looking for trouble as it is too advanced for
> the XENIX 2.3.2 GT OS, sounds funny really ;-).
That CPU is definitely not too fast. I've installed Xenix on a few
Pentium systems. Actually, at my work we used Xenix 2.3.2 on a Pentium
166 for years.
If there's a problem installing on that system, it's not simply that the
CPU is too fast. If the BIOS has any cache options, you might try
disabling them during the Xenix installation. And if there are any
"speed" options (eg. there used to be buttons for "normal" and "turbo"
clock speeds), you could try selecting the slower mode during Xenix
installation.
For some reason the installation seems to be particularly sensitive to
the speed of the system. In my experience, after installation you can
usually switch back to the "fast" settings (cache enabled, etc.).
If all else fails, there's no harm in trying to install on the slower
system. If it works there, you know the boot floppy works. Actually, you
could try just booting from that floppy on the slower system. Nothing
will be done to the hard drive until you actually create partitions, etc.,
so simply booting from the Xenix floppy won't do any harm.
Which brings up another thought. If there's a problem reading the
floppy, the problem could be the floppy drive, not the floppy. Have you
tried cleaning it? Maybe you could swap it with a floppy drive from
another system?
> Secondly, as Scott points out, I may be better off trying
> a newer version of Unix which I may do if I can't get the
> "ancient one" to install, configure etc.
That's a valid point. Xenix is pretty old. A lot of the commands, how
to use them, basic shell scripting, etc. can be applied to newer versions
of *nix. But there are a lot of differences too, like the advent of
graphic user interfaces. Xenix had nothing like that (that I'm aware of).
Of course if the plan is to use old hardware like a 486, you probably
wouldn't want to (be able to?) run a GUI anyway. <g> But I would imagine
that it could run a basic character-based Linux system reasonably well.
> As far as serials and activation keys go, well i am a lucky
> chappy, I have them for the XENIX O/S, SCO Professional,
> SCO Tutor and SCO CGI.
Good. Just making sure.
> If I can't get this going then I wonder if I used the "AT version"
> floppies instead of the 1-3 "GT version" installation floppies,
> would the rest of the XENIX floppies 4-10 etc install properly?
> I would assume only the first install floppy/ies would have to
> be specifically for IDE or SCSI? As the remainder of the software
> should be identical for either?
> Would this assumption be correct?
Hard to say. For example, there could be device drivers on the later
floppies and they could be different. However, from what others have
said, the GT version supports non-SCSI drives as well, so you shouldn't
have to resort to that.
Good luck.
James
Jeff Liebermann
<SPAMMY...@netscape.net> wrote:
>Chris Weaver <c_weaver@del_this_bit.telstra.com> expounded in
>news:3e3a50cb$0$231$5a62...@freenews.iinet.net.au:
>
>> Firstly as Stephen said, being that trying to install on my
>> 486DX/100 is looking for trouble as it is too advanced for
>> the XENIX 2.3.2 GT OS, sounds funny really ;-).
>
> That CPU is definitely not too fast. I've installed Xenix on a few
>Pentium systems. Actually, at my work we used Xenix 2.3.2 on a Pentium
>166 for years.
(trimmed...)
The AMD 486DX4/100 is a rather fragile beast. It runs at about as
fast as the motherboards and memory of the time could handle. The
problems isn't that the processor is too fast, but that it requires
60ns SIMM's to operate correctly. I had a customers office full of
these things on some model of MSI motherboard that drove me insane
with mysterious hangs and reboots. Wholesale RAM replacement was too
expensive so I opted to replace most of the CPU's with Intel 486DX2/66
chips. The hangs and reboots disappeared.
I dunno how much of this applies to Xenix 2.3.2 as this was on DOS 6.x
and Windoze 3.1. Just check the speed of the RAM. Also, there's a
16MB RAM limit for Xenix. The system will recognize more, but won't
use it. I had weird problems on one experiment that was traced to too
much RAM. Bela assured me that it should not be a problem, but I was
in a hurry and just reduced the RAM to 16MB. Fixed.
I've run Xenix 2.3.4 on a variety of semi-modern systems. I think the
fastest was a Pentium 75 with the cache disabled. There are several
data loggers I've blundered across running Xenix 2.3.3 and 2.3.4 on
similar speed systems. No problems. However, when Xenix try it on
fast systems, I can usually get it to load, but find it screws up
trying to talk to ancient ISA tape controllers and smart serial cards.
Moral: Better than stock is not necessarily better. Try to recreate
the system as was fashionable in its day.
--
Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
(831)421-6491 pgr (831)336-2558 home
http://www.LearnByDestroying.com WB6SSY
je...@comix.santa-cruz.ca.us je...@cruzio.com
James J
news:hbqq3vke3612oloqp...@4ax.com:
> ...
> Also, there's a 16MB RAM limit for Xenix. The system will
> recognize more, but won't use it. I had weird problems on
> one experiment that was traced to too much RAM. Bela assured
> me that it should not be a problem, but I was in a hurry and
> just reduced the RAM to 16MB. Fixed.
> ...
Good point about the 16 MB RAM limit. I meant to mention that and
forgot.
Interesting that you had a problem with more RAM. I, too, was under the
impression that the additional RAM would just be ignored by Xenix, but I
don't think I ever actually tried it.
James
Jeff Liebermann
Well, before I was informed that there was a 16MB limit, I built a
Xenix machine with more RAM and had problems with stability. It would
crash, hang, panic, and belch weird error messages at random. I
thought I was dealing with defective hardware, so I transplanted the
hard disk to a different box. Same problem. So, I removed the hard
disk from my highly reliable office server (486DX2/66 with 8MB running
3.2v4.2) and tried again. It was stable for a day or two, and then
just hung. At that point, Bela pointed out the 16MB limit. I reduced
the ram to 8MB, and Xenix lived happily ever after. It's not suppose
to screw up, but it certainly did so in my hardware.
John DuBois
Are you referring to the same incidents? In one case you're talking about
reducing memory to 16MB; in another to 8MB. Having your problems go away when
you reduce memory to 8MB is a lot less mysterious than having them go away when
you reduce memory to 16MB, and doesn't really say anything about the
probability of encountering problems with >16MB memory on a XENIX system.
Jeff Liebermann
wrote:
>Are you referring to the same incidents?
Yes. As I vaguely recall, I initially reduced it to 8MB because I had
a few 4MB SIMM's and I wanted to replace *ALL* the RAM. I later
juggled the chips around and I think (not sure) that I delivered the
system with 16MB of RAM. Both configurations were quite stable on the
original 486DX2/66 system.
I don't recall the original memory configuration that failed. I think
it was 24MB, with 2ea 8MB SIMM's, and 2ea 4MB SIMM's.
>In one case you're talking about
>reducing memory to 16MB; in another to 8MB.
Sorry.
>Having your problems go away when
>you reduce memory to 8MB is a lot less mysterious than having them go away when
>you reduce memory to 16MB, and doesn't really say anything about the
>probability of encountering problems with >16MB memory on a XENIX system.
Huh? Why should 8MB work any differently than 16MB? Any problems
with address space wrap above 16MB as caused by the additional 64KB of
RAM that was added as a result of UMB (upper memory block)?
Here's Bela's original comments on a similar problem.
<http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&selm=199806181828.aa06467%40rubidium.pdev.sco.com>
Bela claims that extra RAM is not a problem. My anecdotal experience
with one system is quite different.
I couldn't find the article where I complained about the Xenix memory
problem earlier.
--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831.336.2558 voice http://www.LearnByDestroying.com
# je...@comix.santa-cruz.ca.us
# 831.421.6491 digital_pager
John DuBois
Jeff Liebermann <je...@comix.santa-cruz.ca.us> wrote:
>On 06 Feb 2003 16:40:50 GMT, spc...@deeptht.armory.com (John DuBois)
>wrote:
>>you reduce memory to 8MB is a lot less mysterious than having them go away when
>>you reduce memory to 16MB, and doesn't really say anything about the
>>probability of encountering problems with >16MB memory on a XENIX system.
>
>Huh? Why should 8MB work any differently than 16MB?
My recollection is that there were problems with ISA boards putting their
memory-mapped IO in the just-below-16M range, only experienced by machines that
had the full 16MB. So, without further details, I wouldn't be surprised to
hear of anyone having problems that appeared when the 8M-16M range was filled
with memory. Having memory beyond 16M, OTOH, isn't a known source of problems.
John
(btw, contratulations on your upgrade :)
Jeff Liebermann
wrote:
>In article <1ut54vgdqc5t5277v...@4ax.com>,
>Jeff Liebermann <je...@comix.santa-cruz.ca.us> wrote:
>>On 06 Feb 2003 16:40:50 GMT, spc...@deeptht.armory.com (John DuBois)
>>wrote:
>>>Having your problems go away when
>>>you reduce memory to 8MB is a lot less mysterious than having them go away when
>>>you reduce memory to 16MB, and doesn't really say anything about the
>>>probability of encountering problems with >16MB memory on a XENIX system.
>>
>>Huh? Why should 8MB work any differently than 16MB?
>
>My recollection is that there were problems with ISA boards putting their
>memory-mapped IO in the just-below-16M range, only experienced by machines that
>had the full 16MB. So, without further details, I wouldn't be surprised to
>hear of anyone having problems that appeared when the 8M-16M range was filled
>with memory. Having memory beyond 16M, OTOH, isn't a known source of problems.
> John
Yep. However, you had to specifically enable the hole between 15MB
and 16MB in the BIOS for it to screwup. I don't think this was the
problem because these machines did not have a smart serial card that
required such a hole. I may have "accidentally" enabled the 15MB-16MB
hole on one board, but not on three of them. I'll send this incident
to the Journal of UnReproduceable Results and see if they'll publish
it.
>(btw, contratulations on your upgrade :)
Thanks. I didn't mean to do it. Of course, everyone is now trying to
bury me in new HF radio projects.
--
Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060