IBM's Processor Plans after the 970

[ffakr]

Re: Re: Re: Doubtful

Originally posted by Jon the Heretic
Yes, these existed and probably still do somewhere. They weren't clones, but RS/6000s (PPC 601 or PPC 604e) which used the CHRP architecture. They could run a CHRP version of MacOS 8.x. A buddy of mine was at IBM Research and played with one of these a few years ago.

These never made it out of IBM research then. Apple never shipped a chrp machine, and the public versions of OS 8 could never boot without Apple firmware.

Although they did exist, running the MacOS on a multiprocessor $10,000+ RS/6000 isn't exactly cost effective, nor did the traditional MacOS make very good use of the multiple processor 604/604e systems. IBM still makes good use of the 604e, if you must know. It is a great multiprocessor chip ...

The Mac OS (8) actually couldn't take any advantage of a SMP system. SMP support was shimmed in with a system library that was mostly designed by Daystar systems (and licensed by Apple). The library allows aware applications to use SMP on the classic Mac OS (Photoshop for example), but the OS never took advantage of this, at least not the core of the OS.

Clones: Not CHRP per se---they used Apple-designed motherboards and parts! CHRP was intended to allieve the reliance on Apple components and the need for Apple to design reference motherboards for their own Mac-cloning competitors. As far as I know, only Apple and IBM ever released anything based on CHRP (all of Apple's New World systems were heavily influenced by the CHRP design). I can't recall if the BeBox was CHRP but that is pretty moribund, may the BeOS rest in peace.

I actually can't comment on how much New World was based on CHRP. NW does use less ROM than previous systems, the bulk of the proprietary ROM was moved into the System Folder, though as far as I understand about 600k still remains in a PROM, even today.
I think it may be more fair to say that IBM designed CHRP not for Apple, but as another way to advance the uptake of the PowerPC platform. The idea was, why design your own systems when IBM will give you a reference design for free. They actually claimed that CHRP would lead to cheaper systems as the PPC was smaller than x86 processors in transistor count.
As I recall, CHRP seemed to be targeted at other potential PPC customers like Linux vendors, Amiga, maybe Be....

At least that's how I remember it... but memory is often colored isn't it?

 

[Abstract]

Originally posted by settledown
i think some od you are missing the piont here. the fact that they ARE different cultures, is good. IBM doesnt make appealing consumer desktops because of their stuffed shirts, they should let apple do that for them. Apple doesnt make big servers well, because they are too busy being cool....they should let ibm do that for them.

the key is for ibms servers to work well with apples desktops for the business market. the windows stronghold (business workstations) would get a good run for its money from the appeal and solid osX of apples computers, but only if they have the server backoffice of good ol reliable and trustworthy, stuffed shirt IBM.

again not a merger, but let each company do what they do well, but with the other company in mind, and as a compliment

Okay, I digress. Great idea. 😉

 

[Phil Of Mac]

Re: Re: Re: Re: Doubtful

Originally posted by ffakr
These never made it out of IBM research then. Apple never shipped a chrp machine, and the public versions of OS 8 could never boot without Apple firmware.

The one and only CHRP system was the Motorola StarMax 6000 🙂

The iMac was based largely on CHRP technology as well.

 

[Jon the Heretic]

Re: Re: Re: Re: Re: Doubtful

Originally posted by Phil Of Mac
The one and only CHRP system was the Motorola StarMax 6000 🙂

The iMac was based largely on CHRP technology as well.

I can't comment on the StarMax, but yes, CHRP is the foundation upon which all newer Macs since the iMac have been built ("New World"). Granted, we don't call it "CHRP" anymore, and I doubt they comply with that specification anymore, but Apple has leveraged that technology considerably.

There is a misperception that CHRP and even Copland "failed". It depends on how you defined success! They failed as "products"---without a doubt---but the underlying technology has been very successful and have been included in many Apple systems. CHRP technology has been used to make cheaper Apple-branded Macs---using cheaper components---and to reduce reliance on the classic ROM. Just because it is no longer called "CHRP" doesn't mean it hasn't been included to a significant degree in every Mac since 1997 or so... Businesses aren't quite THAT stupid to throw away perfectly good intellectual property.

Same goes for Copland. Yes, the underlying OS constructs are "real dead", things like NuKernal, etc. I doubt the NeXT crew at Apple even bothered to look to see if there was anything at the OS level to utilize. They are too entranced with the NeXT technologies to look at their Apple predecessors's ideas from before NeXT acquired Apple [sic]. But other Copland technologies have been very successful in Apple products.

The GUIs of MacOS 8 - 9 have been dramatically enhanced by inclusion of Copland UI elements, from HFS+ to Sherlock to pop-up windows, etc. Almost every new UI element since System 7.6 until (but not including) MacOS X was released was based on Copland UI design concepts. Apple *mined* that intellectual property and milked it for **years**. I have a hunch that Piles and Expose are probably Copland-era concepts, too, so new Copland-era UI concepts may be entering X directly rather than by copying of Classic features which already incorporated them.

Apple has *NO* significant UI research component, but is sitting on a mountain of great UI patents and ideas from Copland-era UI scientists who now work elsewhere. Eventually, they will run out of these old but never implemented concepts and not be able to replace them --- Apple doesn't employ real human factors scientists any longer, just boutique artists --- but fortunately the Copland legacy left them a mountain of great ideas. Pretty good for a "failure".

CHRP and Copland were the most successful "failed products" Apple ever had. The products failed but the technology, well, we've been using it for years.

 

[Jon the Heretic]

Re: Re: Re: Re: Doubtful

Originally posted by ffakr
These never made it out of IBM research then. Apple never shipped a chrp machine, and the public versions of OS 8 could never boot without Apple firmware.

Actually, no.

Because he was in research he had access to big expensive *shipping* RS/6000 systems that ran a CHRP version of MacOS 8.x. No one in his area worked on those systems, they just had access to them. They weren't dedicated Mac boxes, just a curiousity. Too expensive. He is a big Mac fan and thought it was pretty cool way back when. (He worked on things like foldable keyboards, if you must know. Remember that ThinkPad that had a keyboard which unfolded into a full length keyboard? That was an example.)

CHRP wasn't just to run the MacOS, you know, but to run a number of OSes on a common reference platform, including the MacOS. IBM delivered on this platform, though not for the express purpose of running the MacOS, it could. Apple delivered too----there were CHRP-supported builds of MacOS 8.x. I know even the shrinkwrap version of 8.x had at least some CHRP support in it but can't recall how much. I wouldn't be surprised if it could boot a CHRP machine right out of the box. (CHRP required OF support, if you recall. A hardware ROM was not required).

Amiga and BeBox: I had heard they were going CHRP, too, but don't know if they ever made it.

 

[granex]

IBM and NeXT

Many of you may not know that IBM and NeXT once had a pretty close relationship that didn't really end up working too well. When NextStep first came around, IBM licensed it to serve as the OS on some of thier higher end workstations. Apparently this was a far bit of work, because it took some time. By the time IBM was ready to release v1.0 of their offering, NeXT released v2.0 of NextStep, and IBM decided to kill the deal. I don't exactly why NeXT didn't try to help them mature the code as they were working on it, but I assume that they didn't want the competition.

Anyway, IBM might be a little gun shy of Jobs and using the OS X on IBM hardware (don't want to be second class citizens), but I think that it would be amazing for them to port it to noncompeting highend workstations and servers. Servers might not actually make sense, but highend workstations would help Apple to continue to penetrate scientific and video production applications.

<I>Capitalization note<\I>: NeXT and NextStep exist in about ten different capitalization forms (especially the latter), so don't nitpick. I've always prefered the "NeXT" form. NEXTSTEP and NextSTEP were the worst, IMHO.

 

[IJ Reilly]

I put my hands on the first IBM CHRP box at MW-SF. Must have been '97. For some reason, it was not drawing a huge crowd on the floor.

Anyway, why Linux and why not Darwin?

 

[Phil Of Mac]

Yeah, I read that the work on CHRP reduced iMac development time remarkably. (So did building it out of PowerBook parts.)

There was also a CHRP version of Windows NT, incidentally.

 

[MisterMe]

Re: Re: Re: Re: Re: Doubtful

Originally posted by Jon the Heretic
Actually, no.

Because he was in research he had access to big expensive *shipping* RS/6000 systems that ran a CHRP version of MacOS 8.x. No one in his area worked on those systems, they just had access to them. They weren't dedicated Mac boxes, just a curiousity. Too expensive. He is a big Mac fan and thought it was pretty cool way back when. (He worked on things like foldable keyboards, if you must know. Remember that ThinkPad that had a keyboard which unfolded into a full length keyboard? That was an example.)

CHRP wasn't just to run the MacOS, you know, but to run a number of OSes on a common reference platform, including the MacOS. IBM delivered on this platform, though not for the express purpose of running the MacOS, it could. Apple delivered too----there were CHRP-supported builds of MacOS 8.x. I know even the shrinkwrap version of 8.x had at least some CHRP support in it but can't recall how much. I wouldn't be surprised if it could boot a CHRP machine right out of the box. (CHRP required OF support, if you recall. A hardware ROM was not required).

Amiga and BeBox: I had heard they were going CHRP, too, but don't know if they ever made it.

Start up your MacOS X computer in single user mode. You will find that Darwin is targeted to CHRP.

 

[billyboy]

Brokering deals with IBM, VW, the Big Five in the Music Industry. This makes great business sense for Apple to consolidate and grow. I wonder where Apple is heading next? A deal with Bill Gates to take over of M$´ GUI department...

 

[Kermit]

Re: Re: Re: Re: Re: Re: Doubtful

Originally posted by Jon the Heretic
Apple has *NO* significant UI research component, but is sitting on a mountain of great UI patents and ideas from Copland-era UI scientists who now work elsewhere. Eventually, they will run out of these old but never implemented concepts and not be able to replace them --- Apple doesn't employ real human factors scientists any longer, just boutique artists --- but fortunately the Copland legacy left them a mountain of great ideas. Pretty good for a "failure"

If this is true then I am a little worried. Why would Apple not put some of all that cash-reserve they have on some possible revolutionary UI research? The Macs selling point in my mind is its ease-of-use, and good UI design is the very essence of that concept. I read somewhere that Apple had disbanded their Human Interface Group a while back and that that is the reason why Mac OS X and Apples other software suffers from UI inconsistencies. Could all this really be true and if so could someone please try to explain to me what the plan is behind all this. Also if someone could point me in the direction of more articles to read on the subject I would be very thankful.