The clones that weren't

[jessenator]

Wanted to break this out into its own thread from the wonderful discussion happening in the L2 Sonnet thread.
Hopefully this can be a catch-all for any other Mac OS Compatible (nee "clone") and other PREP/CHRP discussion.


I'll start with the previous post from Coloruser

Met@box (originally Pios) started primarily as computer developer. They had a clone sublicense from Motorola and developed a CHRP Mac Clone (Maxxtrem) that was also planned (sans the Mac Chipset) as a new Amiga (many of the guys behind pios were ex Amiga/Commodore guys from germany and the US). I also played with the Tatung clone those days and that was planned to be used as basis for other Pios computers. In parallel they also had a sublicense from Umax. They (we) had clones based on UMAX S900 boards (Magna) in a unique housing and Tanzania I based units in the standard housing (Keenya) with just a special front plate. In the beginning they used daughtercards from a variety of suppliers but then went on to do our own. First a 604r with up to 300 Mhz (but without inline cache like Apple‘s Mach 5) and later G3 and G4 (even with 2mb BS Cache). When the clone license was basically useless, the company only focussed on Upgrades and developed a set-top box as the new main goal (Was not involved in that at all). Next came ZIF upgrades for G3s and then - to suit the Tanzania base - the Joecard T. We never really focussed on Alchemy/Gazelle and as far as i remember never got it running in the Alchemy testbeds. We also had a US subsidiary for a year or so. In the very late 90s, Metabox only focussed on the Set-top-box business and the Mac business slowly dwindled down. No more new developments like Sawtooth etc. made it past the drawing board. Finally in 2001 - as the set-top business failed - the company went bankrupt.

Pios ONE (Amiga Version without Mac Chipset - see the blank spaces) and single CPU daughtercard. Even a CPU existed.

Amiga people pay out even more than classic Mac people.

I went down a rabbit hole on this... I can sort of understand where the desire/hype comes from. Explains a lot about their absence in the Mac world. (that, and their numbers might be smaller than even the L2 G3 upgrades.)

Pios ONE (Amiga Version without Mac Chipset - see the blank spaces) and single CPU daughtercard. Even a CPU existed.

I had never heard of other CHRP-related designs/machines aside from IBM's offerings and the phantoms: Motorola Viper and Umax Basecamp prototypes... hyped right up to it all screeching to a halt as the licensing program was (mostly) dissolved.

Is there a relationship between the Basecamp and the Magna board you mentioned? (looks like an s900 derivative on the surface, but there's probably a lot of differences in the layers I'm guessing.)

 

[Coloruser]

Wanted to break this out into its own thread from the wonderful discussion happening in the L2 Sonnet thread.
Hopefully this can be a catch-all for any other Mac OS Compatible (nee "clone") and other PREP/CHRP discussion.


I'll start with the previous post from Coloruser

View attachment 37752

I went down a rabbit hole on this... I can sort of understand where the desire/hype comes from. Explains a lot about their absence in the Mac world. (that, and their numbers might be smaller than even the L2 G3 upgrades.)


I had never heard of other CHRP-related designs/machines aside from IBM's offerings and the phantoms: Motorola Viper and Umax Basecamp prototypes... hyped right up to it all screeching to a halt as the licensing program was (mostly) dissolved.

Is there a relationship between the Basecamp and the Magna board you mentioned? (looks like an s900 derivative on the surface, but there's probably a lot of differences in the layers I'm guessing.)

Hey,

Magna was simply a S900 Tsunami/Storm Surge Mainboard in a special housing. Only the pios one was a CHRP compliant development.

 

[jessenator]

Magna was simply a S900 Tsunami/Storm Surge Mainboard in a special housing.

Gotcha. Thanks for the clarification

 

[NJRoadfan]

When Dave Haynie spoke at VCF East a few years ago he mentioned his brief time at Pios/Metabox. I don't recall too much about the CHRP boards though!

I still find it amusing that these designs were all going to incorporate ISA and other legacy baggage from the PC platform. I guess they figured that if they are using the Intel PIIX southbridge, they might as well throw them in! CHRP logically should have been legacy free (all USB and PCI) though being a "clean sheet" platform. At least Apple went that direction with the iMac.

 

[Franklinstein]

This is an interesting bit of history isn't it?

WRT ISA and other PC oddities (including MCA for some reason), it was probably originally meant to facilitate transition from x86 to PPC and was folded into the CHRP standard from PReP. Remember that in the mid-90s there were multiple OSes available for PPC including AIX, Solaris, OS/2, and Windows NT 4 (up to and including an RC of Win2k). As long as someone wrote drivers (or if perhaps there was some sort of trickery to get existing x86 drivers to run on the PPC version of a given OS) there's no reason legacy PC peripherals wouldn't work in a PPC (or POWER) workstation. A little ambitious I'd say especially since there wasn't likely much demand for ISA or anything on PPC in the first place, and then CHRP and every other PPC-on-the-desktop project that wasn't a Mac basically died alongside (or perhaps because of) the discontinuation of the Mac clones: Motorola's PowerStacks, the BeBox (mostly this was caused by Gassee's arrogant mismanagement, same with the problems from his tenure at Apple in the late '80s), and this stuff, leaving only Macs and the occasional RS/6000 to limp along waving the PPC banner for desktop computing. There's speculation that Intel paid off Microsoft to drop the non-x86 versions of Win2k. Regardless of the truthfulness of this claim, Microsoft did drop support for non-Intel ISAs and the loss of Windows was the final nail in the coffin of not just PPC but also SPARC, Alpha, and MIPS in the consumer/business desktop space. Granted these had their own unique problems that led to their individual downfalls but the loss of Windows support was likely the common crippling blow.

 

[Unknown_K]

I think Microsoft just seen non X86 chips circling the drain and decided support was not worth the effort. How much software was available for non x86 systems under NT 4? So what if you can boot Windows when there are no apps that will work.

 

[jessenator]

There's speculation that Intel paid off Microsoft to drop the non-x86 versions of Win2k.

I've also seen similar speculation regarding the initial licensing talks: basically big daddy Redmond giving the likes of Compaq a look like:

 

[Coloruser]

The Pios Maxxtrem was based on the Pios One board, completely unrelated to other CHRP/PREP designs. IBM and Motorola both offered their CHRP reference designs like Yellowknife (Motorola) or Long Trail (IBM).

info on IBMˋs approach can be found here: https://www.applefritter.com/files/lt_tour.pdf

info on Motorola’s approach can be found here:

Yellowknife X4 Evaluation System (Discontinued)

The Tatung TPC 6600 was another CHRP design based on IBM’s reference, just like the Umax unit. It is interesting that those use up to two CPU ZIF sockets (the BeBox and the Pios One also being able to have at least two 603 CPUs),

 

[Franklinstein]

Yes the support aspect for continued Windows development on non-Intel platforms would've been huge (for example, this is why Sierra, who had a Half-Life port ready to go for Mac back in 1998, didn't release it: they didn't want to deal with supporting the platform). I'm sure there's a cost/benefit analysis in a Microsoft vault that may shed light on what they had experienced doing so up through NT4, the results of which may have driven their abandonment of the other platforms. The whole multi-platform thing was likely mostly Microsoft hedging their bets and also trying to squeeze out potential competition, so when it looked like they had succeeded and/or it was no longer worth the effort, they stopped.
Or it could've been Intel with a fat check. Maybe both; we'll never know for certain.

What's certain is that there are too many unknowns that may have changed the outcome: if DEC had survived, would x86 performance still be where it is today or would the fact that AMD and Intel couldn't carve up and devour the Alpha's IP like a ham mean that we're still stuck with NetBurst-style trash in the consumer space? If Mac clones had continued would there be more PPC options available, Mac or otherwise, or would CHRP and related projects still have collapsed? In a multi-platform world, would Java have stepped up to become the dominant platform-agnostic application environment it always longed to be, bringing the ability to run the same code with the same experience on any platform, or would it be the same painfully slow, security-hole-filled, bug-riddled mess that it always has been? Tune in next week for Alternate Reality Theatre.

 

[Unknown_K]

Nothing can compete with x86/x64 because of the scale of use.

When the internet came about for most people in the late 90's they needed a computer to use it and the massive expansion of computers in the home that were x86 based (remember there was Nexgen, Cyrix, AMD, Intel, IDT, and probably a few others competing). This is when Apple was taking a dive because they were more expensive and slower and all other platforms were dying as well because they could not compete with the massive X86 gains plus the start of 3D graphics for the PC coming out.

To be honest the only reason we have ARM PCs today is because of the massive amount of smartphones that came out using that chip architecture allowed production and R&D to make them viable for home computing as well. If anything PPC had a resurgence because of game consoles.

 

[Coloruser]

Interesting side note: the IBM Long Trail documentation at one point mentions the compatibility to both, PPC and Cyrix. Would have been cool to have a mainboard that could do x86 and PowerPC just by switching the CPU - of course this another alternative reality.

With Apple’s x86 shift in 2005, we got both Mac and Windows without the hassle to change CPUs. That’s reality - or at least it was until Apple switched again.

 

[Franklinstein]

In the mid '90s IBM was developing a CPU (PPC 615) that could execute both x86 and PPC code natively. It was apparently never fully developed but it was intended to go into a mostly standard Super 7 board which would basically turn it into a CHRP-compliant machine if selected to operate in PPC mode (or mixed code mode), or tell it to run as a Pentium and run regular DOS stuff.

 

[Coloruser]

Right you are. It is not clear from the doc whether it is meant for a combo CPU or PowerPC. As the docs talk about Cyrix and the way the cache slot will be addressed differently then, I guess it is really meant for two different CPUs.

 

[finkmac]

I found this old thread on Macrumors where someone found a Viper machine.

Note the special version of "MacOS on CHRP". Don't think that was ever archived.