New NeXT RISC Workstation photos

Started by Nitro, January 21, 2008, 10:02:58 AM

Previous topic - Next topic

itomato

That would have been an awesome machine to sit in front of.  Even the subtle stuff like the integrated soundbox, the additional power, the expansion slots...  Even the difference in the sound of the fan, blowing air through the entire power supply vs. the arrangement on the slab.

Does anybody know the specs on the expansion slots?
-itomato

Nitro

The original post has been edited to show additional pictures that Tom sent in.  Thanks Tom!
Nitro

nextchef

I wonder if our new forum member Marc has some insights and stories concerning this aborted piece of hardware.

Chef

mdriftmeyer

Yes it works.

It even had a Custom NS 3.3 with SMP compiled support since it's a Dual Processor PowerPC system.

It used a CPU prior to the PowerPC 600 series.

It's called The Brick in NeXT now Apple Engineering.

krfkeith

So, there is still a working version eh?  How far did the Motorola 88k version get?
Newton MessagePad 2100, Original MessagePad (OMP), 2 DataRover 840s
Gateway Laptop w/ OPENSTEP 4.2
NeXTstation Color & Turbo Color

mdriftmeyer

It's a beautiful piece of machinery.

Fiber optic ports included.

NS 3.3 with SMP custom build was never released because they cancelled hardware for Black and the support moving forward with Openstep broke down between Sun and Apple Executives.

Let's just say SUN had Openstep 4 ready to role on Sun hardware. They had everything ported.

The relationship broke down at the negotiating table on money.

Surprised? I wasn't.

HP Gecko were cool hardware. Too bad HP screwed themselves out of some kick ass hardware.

Unless Apple Engineering got rid of their Museum of NeXTSTEP to Openstep to Rhapsody to OS X release builds running on different machines you should be able to see the Brick running smoothly and it makes you just wish Apple would release such a sleak design with Dual 10Gb Ethernet, Fiber Optic Ports, eSata, etc., on their latest hardware and offer this to the Engineering Markets and much more.

Not everyone wants a Tower.

You'd think a Brick would be a big seller where people could have them slotted underneath desks, but not on the floor and another option to sell high quality hardware for the professional markets.

I hope they do it some day.

krfkeith

Now, to clone it!! Haha, just kidding.  Get your hand grenades and machine guns and your ninja outfits.  We're breaking in!! Operation: RISCy has commenced!
Newton MessagePad 2100, Original MessagePad (OMP), 2 DataRover 840s
Gateway Laptop w/ OPENSTEP 4.2
NeXTstation Color & Turbo Color

pentium

Quote from: "krfkeith"Operation: RISCy has commenced!
:lol:  That was a horrible pun.
Just horrible.

-NeXT 68040 Cube with NS 3.3 and 64Mb ram

nfjanette

Quote from: "mdriftmeyer"
Let's just say SUN had Openstep 4 ready to role on Sun hardware. They had everything ported.

The relationship broke down at the negotiating table on money.

Surprised? I wasn't.

Anyone that knew the history of NeXT's contentious relationship with other partners would not be surprised.  Here's a story to consider - it's completely made up, of course.

"Rumor has it that": IBM had spent considerable effort porting the (then) current version of NeXTstep (I can't recall if it was 1.x or 2.x) onto some of their hardware, only to have the surprise of a lifetime when NeXT - unbeknownst to them - publicly released the next major version, making much of the effort useless.   I recall even seeing the printed IBM literature advertising what instantly became an obsolete version of the OS.

That's a part of why you never saw that port of NeXTstep on IBM hardware at a time way before the Sun and HP ports were ever considered.

End of story.

mdriftmeyer

Quote from: "nfjanette"
Quote from: "mdriftmeyer"
Let's just say SUN had Openstep 4 ready to role on Sun hardware. They had everything ported.

The relationship broke down at the negotiating table on money.

Surprised? I wasn't.

Anyone that knew the history of NeXT's contentious relationship with other partners would not be surprised.  Here's a story to consider - it's completely made up, of course.

"Rumor has it that": IBM had spent considerable effort porting the (then) current version of NeXTstep (I can't recall if it was 1.x or 2.x) onto some of their hardware, only to have the surprise of a lifetime when NeXT - unbeknownst to them - publicly released the next major version, making much of the effort useless.   I recall even seeing the printed IBM literature advertising what instantly became an obsolete version of the OS.

That's a part of why you never saw that port of NeXTstep on IBM hardware at a time way before the Sun and HP ports were ever considered.

End of story.

It's no story.

IBM had a license to run NeXTSTEP.

IBM AIX team was competing against NeXT for being the choice of OS on IBM's hardware.

The AIX Team was quite confident until it was shown that NeXTSTEP ran circles around it and that just wouldn't do.

Solution: IBM AIX added an Emulator layer requirement and then installed NeXTSTEP on top of that and during the Executive Briefing all the execs saw was NeXTSTEP running with poorer performance benchmarks compared to AIX.

Sad, but one "non-fictional account" that left a permanent bad taste in Steve's mouth and we didn't even deal with future partnerships after that bit of politics inside IBM.

nfjanette

Quote from: "mdriftmeyer"
Solution: IBM AIX added an Emulator layer requirement and then installed NeXTSTEP on top of that and during the Executive Briefing all the execs saw was NeXTSTEP running with poorer performance benchmarks compared to AIX.

Sad, but one "non-fictional account" that left a permanent bad taste in Steve's mouth and we didn't even deal with future partnerships after that bit of politics inside IBM.

There is more than one way to spin any story, and the version I heard - from someone close to the action - painted NeXT as less the victim and more a player in the fiasco.  I seem to recall OS/2 was involved in the situation as well, but my memory is a bit fizzy on the issue.  This situation was unlike, for example, the ND board compression chip fiasco, in which NeXT was very much the blameless victim.

But, I wasn't there, so it's just a story to me, and I can believe the spin I received (from a huge supporter of NeXT) is incorrect.  I was, however, closely involved in some other business at important NeXT customer sites; those experiences lead me to believe it's more than possible that management arrogance at NeXT was more harmful than helpful in some cases.  It can be a fine line between confidence and arrogance.

mdriftmeyer

Quote from: "nfjanette"
Quote from: "mdriftmeyer"
Solution: IBM AIX added an Emulator layer requirement and then installed NeXTSTEP on top of that and during the Executive Briefing all the execs saw was NeXTSTEP running with poorer performance benchmarks compared to AIX.

Sad, but one "non-fictional account" that left a permanent bad taste in Steve's mouth and we didn't even deal with future partnerships after that bit of politics inside IBM.

There is more than one way to spin any story, and the version I heard - from someone close to the action - painted NeXT as less the victim and more a player in the fiasco.  I seem to recall OS/2 was involved in the situation as well, but my memory is a bit fizzy on the issue.  This situation was unlike, for example, the ND board compression chip fiasco, in which NeXT was very much the blameless victim.

But, I wasn't there, so it's just a story to me, and I can believe the spin I received (from a huge supporter of NeXT) is incorrect.  I was, however, closely involved in some other business at important NeXT customer sites; those experiences lead me to believe it's more than possible that management arrogance at NeXT was more harmful than helpful in some cases.  It can be a fine line between confidence and arrogance.

Without a doubt the NeXTDimension board was not ready for primetime. It definitely was buggy and needed more research before it became a reality. That's all on NeXT.

blakespot

Quote from: "mdriftmeyer"...
HP Gecko were cool hardware. Too bad HP screwed themselves out of some kick ass hardware.
...
Not to derail the thread, but here's some pics of my Gecko.

And a post I made about it's unique display hardware, HP Color Recovery.




blakespot
:::My NeXTs:
- NeXTstation Turbo Color w/ 21" NeXT CRT (NS3.3)
- HP 712/100 "Gecko" (NS3.3)
http://bytecellar.com

neozeed

Quote from: "krfkeith"So, there is still a working version eh?  How far did the Motorola 88k version get?

You know I think it must have worked...


if (! $?MACHINE) then
   if (-d /NextApps) then
       set MACHINE=`hostinfo | awk '/MC680x0/ { printf("m68k") } /MC880x0/ { pr
intf("m88k") }'`
   endif
endif


You can find stuff like that in xnu...

% egrep -rl m88k *
EXTERNAL_HEADERS/stdarg.h
bsd/conf/machine.awk
bsd/conf/tools/doconf/doconf.csh
bsd/ppc/signal.h
bsd/vm/vm_unix.c
iokit/conf/tools/doconf/doconf.csh
libkern/conf/tools/doconf/doconf.csh
libsa/conf/tools/doconf/doconf.csh
osfmk/conf/tools/doconf/doconf.csh
osfmk/libsa/ppc/stdarg_apple.h
pexpert/conf/tools/doconf/doconf.csh


Even the m68k stuff:

% egrep -rl m68k *
bsd/conf/machine.awk
bsd/conf/tools/doconf/doconf.csh
bsd/dev/i386/km.c
bsd/dev/ppc/km.c
bsd/hfs/hfs_macos_defs.h
bsd/kern/init_sysent.c
bsd/kern/kern_core.c
bsd/kern/mach_header.c
bsd/kern/mach_header.h
bsd/kern/subr_prf.c
bsd/netinet/bootp.h
bsd/netinet/ip_auth.c
iokit/conf/tools/doconf/doconf.csh
iokit/mach-o/mach_header.h
libkern/conf/tools/doconf/doconf.csh
libkern/mach-o/mach_header.h
libsa/conf/tools/doconf/doconf.csh
osfmk/conf/tools/doconf/doconf.csh
osfmk/mach-o/mach_header.c
osfmk/mach-o/mach_header.h
pexpert/conf/tools/doconf/doconf.csh


Yet HPPA is practically gone:

% egrep -rl hppa *
EXTERNAL_HEADERS/stdarg.h
bsd/ppc/signal.h
iokit/Drivers/network/AppleBPF/bpf_filter.c
osfmk/libsa/ppc/stdarg_apple.h
osfmk/profiling/ppc/profile-md.h



And finally the sparc stuff:

% egrep -rl sparc *
EXTERNAL_HEADERS/stdarg.h
bsd/crypto/blowfish/bf_locl.h
bsd/crypto/des/des_ecb.c
bsd/hfs/hfs_macos_defs.h
bsd/net/bpf_filter.c
bsd/netinet/ip_auth.c
bsd/netinet/ip_compat.h
bsd/netinet/ip_nat.c
bsd/netinet6/in6.h
bsd/sys/sysctl.h
iokit/Drivers/network/AppleBPF/bpf.c
iokit/Drivers/network/AppleBPF/bpf_filter.c
libkern/c++/OSUnserialize.cpp
libkern/c++/OSUnserializeXML.cpp
osfmk/libsa/ppc/stdarg_apple.h


But then for the heck of it, here is a quick search on the VAX:

% egrep -rl vax *
EXTERNAL_HEADERS/stdarg.h
bsd/i386/endian.h
bsd/include/stdio.h
bsd/kern/kern_subr.c
bsd/libkern/bcmp.c
bsd/netinet/ip_compat.h
bsd/netinet/ip_output.c
bsd/netiso/clnp_debug.c
bsd/netiso/esis.h
bsd/netiso/tp_tpdu.h
bsd/netns/ns.h
bsd/netns/ns_ip.c
bsd/netns/ns_output.c
bsd/ppc/endian.h
osfmk/ddb/db_run.c
osfmk/i386/endian.h
osfmk/kern/queue.c
osfmk/libsa/ppc/stdarg_apple.h
osfmk/mach/i386/exception.h
osfmk/mach/i386/syscall_sw.h
osfmk/mach/i386/vm_types.h
osfmk/mach/ppc/vm_types.h
osfmk/ppc/endian.h
osfmk/sys/varargs.h


Finally you may get a kick out of machine.h in the osfmk..


#define CPU_TYPE_ANY            ((cpu_type_t) -1)

#define CPU_TYPE_VAX            ((cpu_type_t) 1)
/* skip                         ((cpu_type_t) 2)        */
/* skip                         ((cpu_type_t) 3)        */
/* skip                         ((cpu_type_t) 4)        */
/* skip                         ((cpu_type_t) 5)        */
#define CPU_TYPE_MC680x0        ((cpu_type_t) 6)
#define CPU_TYPE_I386           ((cpu_type_t) 7)
/* skip CPU_TYPE_MIPS           ((cpu_type_t) 8)        */
/* skip                         ((cpu_type_t) 9)        */
#define CPU_TYPE_MC98000        ((cpu_type_t) 10)
#define CPU_TYPE_HPPA           ((cpu_type_t) 11)
/* skip CPU_TYPE_ARM            ((cpu_type_t) 12)       */
#define CPU_TYPE_MC88000        ((cpu_type_t) 13)
#define CPU_TYPE_SPARC          ((cpu_type_t) 14)
#define CPU_TYPE_I860           ((cpu_type_t) 15)
/* skip CPU_TYPE_ALPHA          ((cpu_type_t) 16)       */
/* skip                         ((cpu_type_t) 17)       */
#define CPU_TYPE_POWERPC                ((cpu_type_t) 18)


I wonder if they ever released all of the osfmk... it's amazing how many thing it seems to have run on.

And finally as someone pointed out the pre 60X PowerPC cpu's were supported they are mentioned in passing:

/*
*      MC98000 (PowerPC) subtypes
*/
#define CPU_SUBTYPE_MC98000_ALL ((cpu_subtype_t) 0)
#define CPU_SUBTYPE_MC98601     ((cpu_subtype_t) 1)


This is all from Xnu 123.5  (OS X 10.0)... Does anyone have anything older?
# include <wittycomment.h>

Nitro

/* Copyright (c) 1991 NeXT Computer, Inc.  All rights reserved.
*
* File: architecture/m88k/fp_regs.h
* Author: Mike DeMoney, NeXT Computer, Inc.
*
* This include file defines Motorola 88K architecturally defined
* floating point control and status registers.
*
* HISTORY
* 23-Jan-91  Mike DeMoney ([email protected])
* Created.
*/

http://www.opensource.apple.com/darwinsource/DevToolsJun2005/gas-590/include/architecture/m88k/fp_regs.h
Nitro