Installation problems ("free" UW 701)
Hervé PARISSI
1) On my PentiumII box, it doesn't recognize the partition pattern
When I select "Install from cd-rom" it ends up blinking red screens that
state that I have partition problems (well, why dos, win98, linux, BeOS R4
are all working with it, wonder why...). Well I got 2 IDE drives, one 6.4Gb
the other 8.4Gb, and I want to install it on the 1st one.
2) On my older Pentium, it's just that it doesn't see my 3rd drive,
the 1st is an 1Gb IDE, the 2nd is 1Gb SCSI, the 3rd is 3.2 Gb SCSI.
On the list, I only have the 2 first ones, which of course are already loaded
with windows...
Whereas it's not a problem that a free, opensource unix clone doesn't install
(I didn't manage yet to install OpenBSD 2.4, but it's a question of time),
it's really a big big problem for a commercial product. I think I won't recommend
SCO (remember, SCO was build from M$...), but rather the only unix that actually
installs and works smoothly : Linux !
Gary R. Schmidt
>
> 1) On my PentiumII box, it doesn't recognize the partition pattern
> When I select "Install from cd-rom" it ends up blinking red screens that
> state that I have partition problems (well, why dos, win98, linux, BeOS R4
> are all working with it, wonder why...). Well I got 2 IDE drives, one 6.4Gb
> the other 8.4Gb, and I want to install it on the 1st one.
> 2) On my older Pentium, it's just that it doesn't see my 3rd drive,
> the 1st is an 1Gb IDE, the 2nd is 1Gb SCSI, the 3rd is 3.2 Gb SCSI.
> On the list, I only have the 2 first ones, which of course are already loaded
> with windows...
Interesting, stuff usually only installs to the primary-master IDE or a
SCSI ID of 0.
> Whereas it's not a problem that a free, opensource unix clone doesn't install
> (I didn't manage yet to install OpenBSD 2.4, but it's a question of time),
> it's really a big big problem for a commercial product. I think I won't recommend
> SCO (remember, SCO was build from M$...), but rather the only unix that
SCO (remember, SCO was build from M$...), but rather the only unix that
actually
> installs and works smoothly : Linux !
Here we go again, a troll who thinks that SCO came from Microsoft.
Please refer to DejaNews for the last bout of it, not more than 2-3
weeks ago.
Cheers,
Gary B-)
--
---------------------------------------------------------------------------
Gary R. Schmidt at Work: g...@pfxcorp.com
POWERflex Corporation Developers of PFXplus
Vox: +61 3 9888 5833 Fax: +61 3 9888 5451
Visit our Web Site: www.pfxcorp.com
NOTE: All opinions are mine, and bear no necessary relation to those
held by my employers.
---------------------------------------------------------------------------
Gary R. Schmidt
> Gary R. Schmidt <g...@pfxcorp.com> a écrit dans le message : > > Interesting, stuff usually only installs to the primary-master IDE or a
> > SCSI ID of 0.
>
> Well that's what is called a "misfeature" (another M$ invention I suppose)
As far as I now, this a _feature_ of the BIOS, so blame IBM.
> So, remind me as a troll, isn't Xenix a M$ stuff at first hand, hmm ?
No. There are/were some MS parts to it, and copyright stuff, but UW is
not, and never has been, an MS product. The path to SCO of UW is Bell
Labs to Novell to SCO, no MS in that path.
> the shame behing SCO (see, they're even SELLING the old kernel sources even Unix v7,
> well I have Unix v7 kernel sources since 1991, really not interesting). V7 was from Bell Labs.
> I used Xenix 286 8 years ago, it was painful to install, ugly, just a kind a unix with all
> the bad things from dos (segmentation, M$ buggy compiler/debugger and the like)
Again, nothing to do with DOS, that was the 80286 chip you were dealing
with, which had those limitations. I used it for many years, and
developed quite a few large and expensive (for the end buyers) products
on it. Of course, the 80386 was a great improvement.
> Anyway in 1999, a commercial OS that doesn't install well, I really don't
I wonder how it installs on a clean machine?
> more market. Anyway Linux is free (for all the meanings of the word) and Linux is working
> better & better, improving all days.
Yes and no. I just saw an install of Linux on a machine that had Win98
already installed wipe out the partition table, and several other
things. It seems that it doesn't like stuff being more than 4Gb into
the disk, just like SCO UNIX, SCO UW, NT, and all the rest.
BTW - I use Linux, NTWS, NTS, NTTS, Netware, SCO, VMS, AIX, OSF-1,
Win9x, HP-UX, and others.
Hervé PARISSI
> > Gary R. Schmidt <g...@pfxcorp.com> a écrit dans le message :
> > Interesting, stuff usually only installs to the primary-master IDE or a
> > > SCSI ID of 0.
> >
> > Well that's what is called a "misfeature" (another M$ invention I
suppose)
> As far as I now, this a _feature_ of the BIOS, so blame IBM.
So, why do I have Linux & BeOS on my 2nd HD ???
Any 'decent' OS (you know, not DOS, not M$) should install anywhere on
any disk, that's the case for Linux and BeOS, NT and OS/2 still need some
stuff
to be on C:
Anyway I do blame IBM, Intel & M$ all day for having won the game, I always
prefered
68k vs 8086 (I had an Atari ST). I'm not the boss of 'em, I can't go to the
past, and say
"hey the PC will last for decades, make something decent, open, spend more
time on
design...". We have tools to come around these problems.
>
> > So, remind me as a troll, isn't Xenix a M$ stuff at first hand, hmm ?
> No. There are/were some MS parts to it, and copyright stuff, but UW is
> not, and never has been, an MS product. The path to SCO of UW is Bell
> Labs to Novell to SCO, no MS in that path.
Just I was saying that the partition stuff is SO bad it looks like Xenix.
Anyway, I think there's (c) M$ in all PC unices except free ones of course,
maybe it's
why they're so successful ;)
> > the shame behing SCO (see, they're even SELLING the old kernel sources
even Unix v7,
> > well I have Unix v7 kernel sources since 1991, really not interesting).
>V7 was from Bell Labs.
Yes I know, but SCO now owns all the sources from Bell Labs, and they're
selling'em (even V7),
and I can tell you, V7 sources are really really ugly (and it's small for
the bucks they're asking), check www.sco.com
> > I used Xenix 286 8 years ago, it was painful to install, ugly, just a
kind a unix with all
> > the bad things from dos (segmentation, M$ buggy compiler/debugger and
the like)
> Again, nothing to do with DOS, that was the 80286 chip you were dealing
> with, which had those limitations. I used it for many years, and
> developed quite a few large and expensive (for the end buyers) products
> on it. Of course, the 80386 was a great improvement.
Well just the idea of making a 286 unix toy always made me laugh.
Minix was far cheaper for that.
> > Anyway in 1999, a commercial OS that doesn't install well, I really
don't
> I wonder how it installs on a clean machine?
Like others, as it can take all the HD, it works (well, I'm not sure many
OSes can install perfectly
on really huge disk...), but hey, my primary usage is learning SCO UW
specificities at home, not
to build a server. Anyway, it's no excuse for an OS not to account for
what's already on disk.
>
> > more market. Anyway Linux is free (for all the meanings of the word) and
Linux is working
> > better & better, improving all days.
> Yes and no. I just saw an install of Linux on a machine that had Win98
> already installed wipe out the partition table, and several other
> things. It seems that it doesn't like stuff being more than 4Gb into
> the disk, just like SCO UNIX, SCO UW, NT, and all the rest.
>
> BTW - I use Linux, NTWS, NTS, NTTS, Netware, SCO, VMS, AIX, OSF-1,
> Win9x, HP-UX, and others.
Lucky, I wish I had enough money for that...
Anyway, I'm really upset I can't have UW installed.
> Cheers,
> Gary B-)
Guess what :
?>$... (USA are gonna lose!)
Bill Vermillion
<g...@pfxcorp.com> wrote:
>Hervé PARISSI wrote:
>
>> Gary R. Schmidt <g...@pfxcorp.com> a écrit dans le message : > >
>> Interesting, stuff usually only installs to the primary-master
>> IDE or a SCSI ID of 0.
>> Well that's what is called a "misfeature" (another M$ invention I
>> suppose) As far as I now, this a _feature_ of the BIOS, so blame
>> IBM.
And it's also an urban legended, only true if you have a totally
weirdo SCSI card.
You can boot SCO from almost any SCSI ID you wish. Many of the
modern controllers have a setting to specify the boot drive,
however that's not a major limitation if you have something
as recent as an Adpatec 1542C (I've never tried from others).
If you try to install on another drive - say 1 for example -
and it is the first drive found in the system - and you don't
change the boot command on install, you will have problems.
The first drive seen is the boot drive. So if you had
an HD (removeable perhaps) at ID0, another at ID1, and a Zip at
ID6, the systsem would try to boot from ID1.
However the OS 'thinks' it is drive 0. At the install time
used the defbootstr option to set Sdsk to you adaptor
code (blc,ad,??) followed by the parameters in the form
of (0,0,0,0), which give the adator card number, the scsi bus, the
SCSI id, and the LUN.
Once you do that, anytime this drive appears as the first drive, it
will be used for boot.
On SCSI adpators that do not have the 'boot from drive xx' option,
and using the above caveat that you have something like a 1542c or
newer it's just as easy.
Go into the SCSI card set up and turn Scan for BIOS OFF for all
devices prior to the drive you wish to boot from. It's not black
magic, and there is no truth that SCO can ONLY boot from ID0.
I have a removeable HD (with power switch) as HD0. When I turn
it's power off, the next ID - in this case 1, is listed
as the first hard drive, and ID6 - a zip drive - is listed as the
second hard drive.
If I turn SCAN for BIOS off on 1 and 2. the first drive found is
the ZIP at ID6, and I boot into MSDOS. (I did it to try, not than
I run from it).
SCSI is far too flexible to be constrained as most people seem to
think it is.
You can boot from any drive that you can make see as the boot
PROVIDING you also supply the Sdsk parameters to boot: during the
initial install phase. Then you never have to do that again.
--
Bill Vermillion bv @ wjv.com
Bill Vermillion
Hervé PARISSI <hpar...@NOSPAM.mail.dotcom.fr> wrote:
>past, and say
>"hey the PC will last for decades, make something decent, open, spend more
>> > the shame behing SCO (see, they're even SELLING the old kernel
>> > sources even Unix v7, well I have Unix v7 kernel sources since
>> > 1991, really not interesting). V7 was from Bell Labs.
>Yes I know, but SCO now owns all the sources from Bell Labs, and
>they're selling'em (even V7), and I can tell you, V7 sources are
>really really ugly (and it's small for the bucks they're asking),
>check www.sco.com
Well it's a license from SCO - but the sources come from the
PDP preservation society.
The Lions Commentary was also legally published for the first time
thanks to SCO, as the other owners of the code refused to let that
happen, and AT&T even baned the import of the books which could be
confiscated at customs, if found.
It may be ugly, but it's an OS in about 9000 lines of C code
(including comments). Does anybody think NT will be under
30Million lines ? :-)
Peter da Silva
I used the Microsoft version of UNIX before SCO was a gleam in anyone's
eye. I'm still running executables from *that* version on OpenServer
5.0.4. I think that's pretty convincing evidence.
--
In hoc signo hack, Peter da Silva <pe...@baileynm.com>
`-_-' "Heb jij vandaag je wolf al geaaid?"
'U`
"Tell init(8) to lock-n-load, we're goin' zombie slaying!"
Peter da Silva
>not, and never has been, an MS product. The path to SCO of UW is Bell
>Labs to Novell to SCO, no MS in that path.
Um, actually, no. It's Bell Labs to AT&T's UNIX Support Group to Microsoft
to SCO. Meanwhile USG kept developing theur non-bell-labs version of UNIX
with input from Berkeley and SCO... there were several re-merges on both
sides all the way along.
Both OpenServer 5.0.4 (SVR3.2v5) and Unixware (SVR4.2) run Microsoft
Xenix-286 binaries (I can telnet into both from here and demonstrate),
showing Microsoft input on both sides of the path that lead to Unixware 7.
For a 1984 version of UNIX, Xenix-286 wasn't half bad.
Jean-Pierre Radley
|
| Both OpenServer 5.0.4 (SVR3.2v5) and Unixware (SVR4.2) run Microsoft
| Xenix-286 binaries (I can telnet into both from here and demonstrate),
| showing Microsoft input on both sides of the path that lead to Unixware 7.
Except that starting with UnixWare 7.01, you no longer can run those old
Xenix binaries.
--
Jean-Pierre Radley <j...@jpr.com> XC/XT Custodian Sysop, CompuServe SCOForum
Peter da Silva
Jean-Pierre Radley <j...@jpr.com> wrote:
>Peter da Silva averred (on Sat, Apr 10, 1999 at 12:17:18AM +0000):
>| Both OpenServer 5.0.4 (SVR3.2v5) and Unixware (SVR4.2) run Microsoft
>| Xenix-286 binaries (I can telnet into both from here and demonstrate),
>| showing Microsoft input on both sides of the path that lead to Unixware 7.
>Except that starting with UnixWare 7.01, you no longer can run those old
>Xenix binaries.
Because they no longer ship x286emul, so they can get out from under a
licensing deal with Microsoft. Unixware 7 is obviously a merge of OpenServer
and the original Novell Unixware (what they renamed SVR4.2 to after buying
USL), and it would be strange if they dropped more code than they had to to
satisfy their contract with their parent company.
(Parent company? Yep. Last I checked Microsoft owned over 30% of SCO, so
it's a reasonable description of their relationship.)
Jean-Pierre Radley
| In article <1999040921...@jpradley.jpr.com>,
| Jean-Pierre Radley <j...@jpr.com> wrote:
| >Peter da Silva averred (on Sat, Apr 10, 1999 at 12:17:18AM +0000):
| >| Both OpenServer 5.0.4 (SVR3.2v5) and Unixware (SVR4.2) run Microsoft
| >| Xenix-286 binaries (I can telnet into both from here and demonstrate),
| >| showing Microsoft input on both sides of the path that lead to Unixware 7.
|
| >Except that starting with UnixWare 7.01, you no longer can run those old
| >Xenix binaries.
|
| Because they no longer ship x286emul, so they can get out from under a
| licensing deal with Microsoft. Unixware 7 is obviously a merge of OpenServer
| and the original Novell Unixware (what they renamed SVR4.2 to after buying
| USL), and it would be strange if they dropped more code than they had to to
| satisfy their contract with their parent company.
|
| (Parent company? Yep. Last I checked Microsoft owned over 30% of SCO, so
| it's a reasonable description of their relationship.)
Microsoft NEVER owned anywhere near 30% of SCO, and these days, it's 12.3%.
Gary R. Schmidt
>
> In article <370C101B...@pfxcorp.com>,
> Gary R. Schmidt <g...@pfxcorp.com> wrote:
> >No. There are/were some MS parts to it, and copyright stuff, but UW is
> >not, and never has been, an MS product. The path to SCO of UW is Bell
> >Labs to Novell to SCO, no MS in that path.
>
> Um, actually, no. It's Bell Labs to AT&T's UNIX Support Group to Microsoft
> to SCO. Meanwhile USG kept developing theur non-bell-labs version of UNIX
> with input from Berkeley and SCO... there were several re-merges on both
> sides all the way along.
> Both OpenServer 5.0.4 (SVR3.2v5) and Unixware (SVR4.2) run Microsoft
> Xenix-286 binaries (I can telnet into both from here and demonstrate),
> showing Microsoft input on both sides of the path that lead to Unixware 7.
So, if I run a Z80 emulator on my RS/6000, does that mean that
"Sinclair" owned some part of it? Or if a Microport binary runs on SCO
UNIX, Microport "own" it?
> For a 1984 version of UNIX, Xenix-286 wasn't half bad.
Correct. It was bloody brilliant, and extremely cost-effective at the
time. (Remember, network cards cost a bomb!)
Bill Vermillion
>> In article <370C101B...@pfxcorp.com>, Gary R. Schmidt
>> <g...@pfxcorp.com> wrote:
>> Both OpenServer 5.0.4 (SVR3.2v5) and Unixware (SVR4.2) run
>> Microsoft Xenix-286 binaries (I can telnet into both from here
>> and demonstrate), showing Microsoft input on both sides of the
>> path that lead to Unixware 7.
>So, if I run a Z80 emulator on my RS/6000, does that mean that
>"Sinclair" owned some part of it? Or if a Microport binary runs on
>SCO UNIX, Microport "own" it?
Don't confuse a Z80 emulator with ZX80 emulation. The former is a
Zilog(Exxon) chip the latter is "Sinclair" machine. (I hope I
haven't screwed that up - it's been 18 years since I touched one
and didn't like the membrane keyboard, but liked some of their
other ideas - such as the self-completing lines in BASIC.
Peter da Silva
Gary R. Schmidt <g...@pfxcorp.com> wrote:
>Peter da Silva wrote:
>> In article <370C101B...@pfxcorp.com>,
>> Gary R. Schmidt <g...@pfxcorp.com> wrote:
>> >not, and never has been, an MS product. The path to SCO of UW is Bell
>> >Labs to Novell to SCO, no MS in that path.
>> Um, actually, no. It's Bell Labs to AT&T's UNIX Support Group to Microsoft
>> to SCO. Meanwhile USG kept developing theur non-bell-labs version of UNIX
>> with input from Berkeley and SCO... there were several re-merges on both
>> sides all the way along.
>No. Microsoft _never_ owned in fee-simple the UNIX source code.
I didn't say they had. I'm talking about who worked on it, not who owned it.
>> Both OpenServer 5.0.4 (SVR3.2v5) and Unixware (SVR4.2) run Microsoft
>> Xenix-286 binaries (I can telnet into both from here and demonstrate),
>> showing Microsoft input on both sides of the path that lead to Unixware 7.
>So, if I run a Z80 emulator on my RS/6000, does that mean that
>"Sinclair" owned some part of it? Or if a Microport binary runs on SCO
>UNIX, Microport "own" it?
If your emulator needs kernel support from Sinclair or Microport, then I'd say
there was Sinclair or Microport heritage there. We both know it doesn't, not
the way the Xenix-286 emulation needed some really nasty stuff (we had to hack
x286emul to get it to work reliably on Unixware. SCO's x286 support was much
better: it was a matter of how they handled requests for <64K segments in the
kernel... Unixware returned a 64K segment regardless, which made the PL/M-286
compiler and the 286 linker chew up gigabytes of swap before dying).
t...@my-dejanews.com
> If your emulator needs kernel support from Sinclair or Microport, then
I'd say
> there was Sinclair or Microport heritage there. We both know it
doesn't, not
> the way the Xenix-286 emulation needed some really nasty stuff (we had
to hack
> x286emul to get it to work reliably on Unixware. SCO's x286 support
was much
> better: it was a matter of how they handled requests for <64K segments
in the
> kernel... Unixware returned a 64K segment regardless, which made the
PL/M-286
> compiler and the 286 linker chew up gigabytes of swap before dying).
Interesting--why does the kernel care about the size of 286 segments on
Unixware? On the x286emul Darryl Richman and I wrote at Interactive
Systems Corp. for AT&T and Microsoft, the kernel didn't care about 286
segment sizes, if I recall correctly. If the Xenix program wanted a
segment of N bytes, we just malloc'ed N bytes, and then told the kernel
to set up an LDT entry refering to that N bytes.
The major nastiness we ran into was small vs. large model. Our x286emul
was based on i286emul, which was for running 286 System V binaries on
386/ix. Those binaries were sensible--the C and system call libraries
converted everything to large model before calling the kernel. E.g., if
a small model program did read(0,buf,4096), the library would change buf
to a far pointer before actually doing the system call. Xenix binaries,
on the other hand, did not, so the kernel (and, therefore, x286emul) had
to deal with this.
--Tim Smith
--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---