What If It Wasn't PPC

[dammy]

Which magic in ia32 and amd64 are you thinking about ?

Does quad core with direct connect architecture for under $200 USD count?

Dammy

[weirdami]

@dammy

I remember there being some chip that McEwen was pushing that was some kind of really low temperature thing that would "emulate" something. I think it was a mobile chip, maybe. I just remember him saying that it would "emulate", but I can't remember what it was emulating. In any case, he was on Dvorak's TV show years ago and said something about how when the chip came out it didn't work like they said it would so they didn't go with it.

Anyhoo, AI's been trying to do the mobile thing for years now. Wasn't the whole PPC thing AT's deal instead anyway?

[itix]

@dammy

For a mobile device, it ECS/AGA compatibility wouldn't have been needed, at least not initially. If they had gotten a base model out and released a decent SDK, we would have many ports as we see to day for OS4/MOS/AROS.

Which companies would have developed software for new Amiga? Companies which didnt have sentimental ties to Amiga dropped it right after collapse of Commodore. For the world Amiga was dead. There is no way tiny Amiga Technologies could have convinced those companies to join in.

Please also note that most games ported to NG Amiga systems are just Linux/SDL ports. Some of them are really cool but this option was not available at that time... Linux, just like Internet, was only starting to shape up.

And not to forget... HP released their OS 3.5 in 1999. That was quite late, wasnt it? Now how long it would have taken to port the OS to another CPU? We have got real life example...

With a jump to the more modern DragonBall MX, EUAE integration would have been option.

DragonBall MX was released when, in 2001 or so? If Amiga would have survided (and sort of florished) from year 1995 to 2001 there would be no need for UAE integration.

Anyway Amiga Technologies/Amiga Inc. was never controlling Amiga and its direction. Petro was only a marionette trying to hang along.

[itix]

@the_leander

Nowhere. Without full OCS or AGA compatibility it would have been yet another fiasco for Amiga Technologies. It was the custom chips which made Commodore-Amiga successful.

Tbh I thought it was the fact that you could get the games in just about any playground in the UK that made it such a hit over here... Might be wrong though...

Maybe. But when Amiga Technologies stepped in that industry was already dead.

[cantido]

corto wrote:
Which magic in ia32 and amd64 are you thinking about ? PPC is a modern and clean processor. If it was more popular, processors and compilers will be even better.

What makes the PPC any cleaner than ARM, MIPS, SPARC?,... Maybe ia32 is a mess but most of negative comments about it come from people that have no idea what the hell they're talking about. There are broken bits but most of the time that stuff is handled in the OS and simply doesn't matter to desktop users. Anyhow it's probably better to listen to people like Theo de Raadt et al. on what's wrong with ia32 opposed to passers by on forums that have uneducated axes to grind.

Whats good about ia32 and now amd64.... Number 1; hellava cheap. Number 2; Performance/Dollar, Pound etc ratio very high; Number 3; Lots of modern (read as fast) hardware around. Number 4: Did I mention its cheap?

-- Edit, forgot to mention what ia32/amd64 has that most don't

- Decent'ish hardware virtualisation support without breaking the bank.
- Hardware support for marking no executable pages. All ia32 makers seem to have done this including Via and Transmeta. In Linux at least this seems to be emulated for PPC and SPARC.
- Highspeed memory and processor interconnects within the same die.

blah etc blah

oh did I say.. it's cheap?

[dammy]

by itix on 2008/11/29 15:48:24

And not to forget... HP released their OS 3.5 in 1999. That was quite late, wasnt it? Now how long it would have taken to port the OS to another CPU? We have got real life example...

We certainly do. It took Dr.Schulz about seven months, doing it part time (full time job, wife with newborn plus two very young children) to port AROS to SAM440, and that was a crap load of stuff being ported or fresh code.  Had he been full time at it, I bet he could have done it in less then 12 weeks.

Dammy

[AmiDelf]

Its not about PPC beeing any better than ARM or x86. All I say is that Amiga alike operating systems should stay on PPC. In this way AmigaOS 4 or MorphOS can create communities which promotes for what they use to others. On x86 MorphOS or AmigaOS 4 would just vanish into nothing like BeOS.

Amiga doesnt need commercial forces to get known. The scene have served Amiga advert since day one etc.

With a scene and community which Amiga/MorphOS got now. Users get software made with quality in mind and not in commercial share holders minds.

Example to this is AmiNetRadio for MorphOS and AmigaOS 3.x (yes I write in MorphOS favour now as I dont have a AmigaOS 4 machine). This is just awesome piece of software. It competes bravely with iTunes etc. With some adjustments, this piece of software shows how the scene makes the amigaworld go around.

Another one is Titler for MorphOS. Its just awesome to be able to do simple 3D text/obj animations for video. If I wanted such software for Windows or MacOS X I had either to pay a lot or try the endless search engines etc. MorphOS gives the oportunity for the developer to evolve.

I really like to see PPC as an alternative platform. No monopoly is healthy, and thats why we as Amigans should help PPC in right direction. We might not be million users, but look what Amiga have done! Aminet is a proof of that. Its really amazing to see Amiga alive after so many year. Proof is that this is something that should be kept on.


Its night, and Ive seen James Bond at the cinema ;P But my words makes sense somehow. Conclusion is just that PPC is a great alternative to x86. If we loose PPC platform, I just dont want to see the end result. I quit computing stuff then for sure,...!

[quarkx]

I haven't been keeping up on PPC, so forgive this if it sounds dumb, but isn't the cell processor multiple PPC cores?

[dammy]

Its not about PPC beeing any better than ARM or x86. All I say is that Amiga alike operating systems should stay on PPC. In this way AmigaOS 4 or MorphOS can create communities which promotes for what they use to others. On x86 MorphOS or AmigaOS 4 would just vanish into nothing like BeOS.

IIRC, BeOS had the biggest increase in number of users and developers when they went to x86.  BeOS management, OTOH,  :roll: Arch is meaningless TBH, it's about the OS that matters.

Dammy

[AmiDelf]

OS yes, but for Amiga to survive. A skip over PPC is not a good way to go. Atleast with Amiga on the PPC side. Those companies which tries to do something for the community like Genesi or Acube. They can earn because of this community. They wouldnt had a chance on x86.

You see, OS is not all about OS... its also about making it be alive

[dammy]

I haven't been keeping up on PPC, so forgive this if it sounds dumb, but isn't the cell processor multiple PPC cores?

Single PPC core with SPEs, see Wiki. Not very impressive for a desktop CPU (no OoO processing) and using a PS3 would be gimped by Hypervisor.  Which is why the thread if AOS went to another arch, where would we be now?  aPhone?

Dammy

[itix]

And not to forget... HP released their OS 3.5 in 1999. That was quite late, wasnt it? Now how long it would have taken to port the OS to another CPU? We have got real life example...

We certainly do. It took Dr.Schulz about seven months, doing it part time (full time job, wife with newborn plus two very young children) to port AROS to SAM440, and that was a crap load of stuff being ported or fresh code. Had he been full time at it, I bet he could have done it in less then 12 weeks.

AROS is written portability in mind. OS 3.1 wasnt and according to an interview of Olaf Barthel it seems that it was quite mess:

While I was still doing consulting work for Amiga Technologies GmbH in 1995/1996 I had started to rework the AmigaOS source code to build on a single Amiga (the original code required more than one computer, one of which had to be a Sun/3 workstation).

Andy Finkel had prepared an operating system build for Amiga Technologies, which I tried to get to work on my computer system at home. At that time the R&D effort undertaken by the company was picking up speed (only to be wound down not much later when the parent company, ESCOM, filed for bankruptcy) and I thought that it would be helpful to have a complete working build available for future development work. There was no R&D lead in the company handing out tasks, I picked this job myself.

It took a while to become familiar with the build environment and the tools. My goal was to reduce the number of different compilers and assemblers required to build the operating system. Ideally, you'd just have to say "make" and a couple of hours later you'd have the Kickstart ROM files and the Workbench distribution sitting on your hard disk. In total, it must have taken 1-2 years to bring the build to this level of functionality. I worked on this on and off whenever there was time.

Which again implies that Amiga Technologies was not able to carry Amiga development on. Not at least after collapse of ESCOM.

I also found this gem (dated 07/11/95):

During his key note address held in Los Angeles at the Video Toaster Expo, Petro Tyschtschenko, CEO and President of Amiga Technologies officially announced the Power PC to be the processor used in the future generation of Amiga computers.

 The first POWER AMIGA will be available 1st quarter 1997 and will feature the Power PC 604 RISC CPU. Further models will be available later in the entry-level, as well as in the mid-range.

 The Power Amigas will be backwards compatible with current models and will also feature a new and more powerful chipset.

 "Our pre-emptive multitasking Operating System AmigaOS will be ported to the Power PC platform first. Our goal is to make our OS hardware independent to allow further ports on other platforms", said Petro Tyschtschenko.

 He also added: "We have a clear business plan: We do things consequently, step by step. First, we ramped up the production and set up an organisation to handle the Amiga market and satisfy the demand. Now that we have achieved this successfully, we focus on research & development to bring new and better products on the market. One of the mistakes the former Commodore made, was to do too many things at a time, too many promises and therefore losing focus on important aspects of its business. We have learned from these mistakes and won't repeat them. Commitments and promises are nice but facts are better."

 The developments will also focus on including more features in the AmigaOS, especially regarding network abilities and memory management.

 The development of the native RISC AmigaOS will be made internally at Amiga Technologies. An R&D department is currently being set up in Bensheim with sufficient engineers to meet the announced schedules. Former well known Commodore engineers as well as new competencies will join the team in Bensheim this year.

 This development project will also be involving a dozen companies in close partnership with Amiga Technologies GmbH.

 More good news for all Amiga users: The Power PC technology will not only be available for new Power Amigas. Thanks to a close co-operation between Amiga Technologies and Phase V, a German turbo board manufacturer, a full range of Power PC boards will also be available for the A1200, A3000 and A4000 series.

 This will allow a general migration of the Amiga platform towards Power PC in a short time, also for current models.

ESCOM declared itself insolvent in July 1996. Ten years later Amiga Inc announced AmigaOS 5.0 :-P

[AeroMan]

taunusand wrote:

BillE wrote:
I wouldn't buy a mobile phone no matter who the manufacturer was. I don't want one and I don't need one.

Who can live without a phone theese days? :-?

I gave my cell phone to my grandmother in 2003. My life has been so much better since I did it...
It is just like quit smoking.  :-o

[AeroMan]

dammy wrote:

Which magic in ia32 and amd64 are you thinking about ?

Does quad core with direct connect architecture for under $200 USD count?

Dammy

Freescale has an 8 core chip running @ 1.5GHz. The QorIQ P4080.
It also has loads of on chip peripherals that ia´s and amd´s doesn´t.

But the focus for PPC market is not desktop PCs anymore. So we should not expect these beasts in our desktops unless somebody at IBM have the crazy idea of selling a POWER6 or a Cell based machine. I don´t believe they will do that.

To be honest, I think ARM and MIPS are better designs than PPC and x86. ARMs can do so much better with less silicon and power consumption than x86. They are cheap also, and sooner or later portable devices will take them to the same performance level that we have today with x86.

An ARM based Amiga could have all those benefits, but back in the late nineties, we had a scenario where PPCs were awesome processors, and ARMs were just quite nice chips. Who would bet they would have gone that far and that PPCs were almost dropped? It would be just like declaring that x86 will be at second place in sales by 2020

[persia]

What if Amiga Inc actually produced a product?

Nah, even someone with a good imagination would have trouble with that...

[Tron2k2]

Transmeta.  They had a chip which could be a chamelion, supposedly, and be reconfigured to emulate x86, PPC, etc.  It never worked right :-(  At least not at any respectable speed (and this was 'respectable' in 2002, mind you..)