@Derfs

Why I blame Hyperion? Because as far as I remeber they are the ones who made OS4 possible. That is why they are the ones who have the knowledge to do it ;)

Its their OS4

_________________

@Crumb

Quote:

Crumb wrote: MorphOS had clear plans in the right direction, as they designed it in a way that all AmigaOS compatible software would run in a sandbox and future applications and games would run in the unveiled "Q-Box". Unfortunately the Q-Box API didn't evolve enough and being unfinished, the public docs didn't were published

There are no docs about Q available, because it is virtually not present yet. Currently Q is Quark plus some servers. More of Q has not arrived and it's features are not graved into stone yet.
Currently Q is a good idea, but not much more.

_________________
My programs: via.bckrs.de
MorphOS user since V0.4 (2001)

@Hagbard_Celine

Quote:

Hagbard_Celine wrote: @Derfs Why I blame Hyperion? Because as far as I remeber they are the ones who made OS4 possible. That is why they are the ones who have the knowledge to do it ;) Its their OS4

i think you are confusing them with amiga inc, as in hyperion dont own everything to do with os4 being ported to any system they like. they can make os4 for ppc, and thats it.

do you want hyperion to delay finishing the ppc version to get a contract for x86, even though they have said over and over they would never do that anyway.

there is such a thing as putting the cart before the horse you know..

however this is getting off topic, as carl-s is talking about where we see this in the future, not putting the blame game on companys and people now.

*edit* sic

Last edited by Derfs on 24-Oct-2006 at 04:47 PM.
Last edited by Derfs on 24-Oct-2006 at 12:32 PM.

_________________

@olsen

Quote:

Last, but not least, the overall PowerPC architecture wasn't so ugly either. Compare this to what the i386 legacy brings with it today and weep.

Which is why I noted in particular: Intel CPUs still carry a lot of legacy around with them. AMD CPUs a lot less so. The biggest considerations of "legacy" are that both still start in real mode and are switched into protected mode/long mode when the OS starts. None have been so bold as of yet to drop those modes completely.

Intel dropped a lot of legacy "cruft" from the southbridge with the ICH8 - saying goodbye to the legacy devices such as floppy disk controllers, serial and parallel ports.

I'd say the x86/x86_64 architecture is the cleanest it has ever been.

The PowerPC had no baggage to carry along with it. It was a new architecture, and still is to some extent.

Quote:

In total, I'd say that going to the PowerPC made sense then and still makes sense today. The reasons for choosing this platform and no other are not as simple as you may have concluded.

I'm less convinced. Don't get me wrong; I'm not a single-arch fascist, but if I had to sit down tomorrow and start picking a hardware architecture for the framework of an OS I had, I'd have a hard time picking PowerPC over x86. And two good reasons:

1. The CPUs have to be purchased in bulk volumes and are expensive.
2. Supporting chips are expensive and each one is lacking in one or more feature (e.g. Marvell Discovery II - supports common southbridge controllers, but lacks AGP).

Compare to x86:

1. There has been a variety of CPUs for each socket architecture with a clear price/performance graph meaning you can buy a single, cheap CPU if you so desire or an expensive high performance CPU if you desire (c.f. Celeron/Pentium, Sempron/Athlon).
2. VIA, Intel, SIS, ATI, nVidia, etc. produce northbridge/southbridge combinations, some can be mixed and matched, and are cheap. Downside is they have to be purchased in bulk, but this is less of an issue than it is with the comparatively expensive CPU.

Quote:

At the time this decision was made Motorola was very supportive of Amiga's intention to go with the PowerPC.

IBM are very supportive of the PowerPC right now. I don't know if Freescale are. So why don't we see more PowerPC boards being produced?

@Carl-S
This may be the most sensible definition/vision of Amiga I have ever seen in the recent years...

@Carl-S

@Carl-S

Quote:

I'm an old CPU architecture person myself... and I understand your points and agree. Yes, the x86 is a Frankenstein. It's horrible. But, here's the rub: Fast Cheap Motherboards (Tiger Direct Link). That is what economies of scale and serious competition make possible. For me, I reached a point where I just imagine the x86 instruction set to be a form of microcode... which is always ugly and difficult to read, but no one really cares.

Interesting thoughts. I have nothing against x86. Rather the opposite. In my opinion, it is *the* desktop architecture, and I am totally dependent on what it offers (when it comes to performance, OS and applications), both on a professional and personal level.

But for many reasons, I don't think this architecture is relevant for *OS4*.

For real and true desktop use, I see no possibility whatsoever that Amiga can go head to head with Windows and Mac (all questions in that post must be answered, and it will still only be a start). The way I see it, an Amiga Desktop is only relevant to developers and nostalgic hobbyusers, not the serious/professional desktop user.

OS4 is PPC. That's the way it is, like it or not. This won't change by the press of a button, this whole time since OS4 was announced is how long time the 68k to PPC migration took. As you say Carl, the key is being cheap and making the technology accessable for the broad masses, the underlying architecture is of less importance. Couldn't agree more. But x86 is not necessarily a must for that, PPC does have things to offer as well. Here, here, here, here, and here you can see what an efficient OS like Amiga can make out of really cheap PPC hardware like this (more info about that demo here). I think that is quite amazing actually. This is exactly what is needed! Oh, and one more here. As these video clips show, it's on devices like this that Amiga can really make a difference. It makes it possible to build really cheap and low power (as in performance, watts and heat), yet have a completely usable system.

If desktop *really* is the goal, I would advocate a little creativity, think out of the box in order to find alternative ways of gaining momentum. Maybe this could be a start? I'm not talking about end-goals here, but more like a mean to gain momentum, to add both users and developers (the creativity you mentioned in your initial post) that nobody ever thought of, and to make the platform grow that way? (How well wouldn't REBOL fit into this picture, BTW? )

(Note - while I write OS4, I also mean MorphOS, since they are in a much similar position)

@takemehomegrandma

will you please stop plying your wares. plus it still isnt that cheap.

*edit*

removed pondering

Last edited by Derfs on 24-Oct-2006 at 04:47 PM.
Last edited by Derfs on 24-Oct-2006 at 04:31 PM.

_________________

@nine

Quote:

nine wrote: @olsen Quote: Last, but not least, the overall PowerPC architecture wasn't so ugly either. Compare this to what the i386 legacy brings with it today and weep. Which is why I noted in particular: Intel CPUs still carry a lot of legacy around with them. AMD CPUs a lot less so. The biggest considerations of "legacy" are that both still start in real mode and are switched into protected mode/long mode when the OS starts. None have been so bold as of yet to drop those modes completely. Intel dropped a lot of legacy "cruft" from the southbridge with the ICH8 - saying goodbye to the legacy devices such as floppy disk controllers, serial and parallel ports. I'd say the x86/x86_64 architecture is the cleanest it has ever been.

Given where it came from and what it went through, that's easy to say

But, at least for me, the biggest bit of baggage the x86 architecture carries around with itself is the instruction set architecture. No matter if what goes into CPU is unlike what the processing unit(s) eventually work with. From what I've learned and learned to appreciate, the less effort that goes into telling the processing unit what it should do, the better. Given a choice, I'd rather tell it as PowerPC instructions than in x86/x86_64 instructions, having to make assumptions about how the register renaming may take place, how the pipelines will be utilized and how the cache hierarchy may come into play. I would prefer as much transparency in the instruction set architecture as possible. And that simply isn't there in the x86 world.

Quote:

The PowerPC had no baggage to carry along with it. It was a new architecture, and still is to some extent. Quote: In total, I'd say that going to the PowerPC made sense then and still makes sense today. The reasons for choosing this platform and no other are not as simple as you may have concluded. I'm less convinced. Don't get me wrong; I'm not a single-arch fascist, but if I had to sit down tomorrow and start picking a hardware architecture for the framework of an OS I had, I'd have a hard time picking PowerPC over x86. And two good reasons: 1. The CPUs have to be purchased in bulk volumes and are expensive. 2. Supporting chips are expensive and each one is lacking in one or more feature (e.g. Marvell Discovery II - supports common southbridge controllers, but lacks AGP).

It's a very different market today. Back in the mid-1990'ies the chip makers were still actively occupying their niches and looking to find other markets. These days you seem to have rather rigid market segments, as in: simpler microcontrollers, embedded systems, general purpose processors, special purpose processors.

And in this market segmentation, the PowerPC has fallen from the ranks of being a general purpose processor. It's being made for applications which do not sell in the same volumes as simpler microcontrollers or general purpose processors. It's somewhere in between, and we, who are trying to use the PowerPC in the Amiga, are paying the price for it.

Quote:

Compare to x86: 1. There has been a variety of CPUs for each socket architecture with a clear price/performance graph meaning you can buy a single, cheap CPU if you so desire or an expensive high performance CPU if you desire (c.f. Celeron/Pentium, Sempron/Athlon). 2. VIA, Intel, SIS, ATI, nVidia, etc. produce northbridge/southbridge combinations, some can be mixed and matched, and are cheap. Downside is they have to be purchased in bulk, but this is less of an issue than it is with the comparatively expensive CPU.

Different markets, different applications. Apples and pears

Quote:

Quote: At the time this decision was made Motorola was very supportive of Amiga's intention to go with the PowerPC. IBM are very supportive of the PowerPC right now. I don't know if Freescale are. So why don't we see more PowerPC boards being produced?

If the product you're making is not tied to a specific processor architecture, these days you pick the most cost-effective solution you can. And the most cost-effective solution (other limiting factors excluded) is not necessarily the PowerPC. We don't see more PowerPC boards because they cannot necessarily compete with the alternatives in terms of cost-effectiveness and (these days) processing power delivered. For example, the R&D work done on making the x86 architecture processors faster and cheaper certainly does pay off.

@nine

Quote:

Which is why I noted in particular: Intel CPUs still carry a lot of legacy around with them. AMD CPUs a lot less so. The biggest considerations of "legacy" are that both still start in real mode and are switched into protected mode/long mode when the OS starts

Factor in EFI equiped Intel Macs... The PC market defined legacy support as a strength not a weakness.

Note the lesson on why all of the 68K PC platforms are all (commercially) dead i.e.
AMD/Intel upgrade their CPU line while existing X86 software automatically gains faster computation performance.

Quote:

The PowerPC had no baggage to carry along with it. It was a new architecture, and still is to some extent.

Note that PowerPC ISA is not the "cleanest" RISC ISA.

Last edited by Hammer on 24-Oct-2006 at 01:57 PM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@Hammer

Quote:

Hammer wrote: @nine Quote: [..] [quote] The PowerPC had no baggage to carry along with it. It was a new architecture, and still is to some extent. Note that PowerPC ISA is not the "cleanest" RISC ISA.

Definitely. But the "cleaner" architectures tend to be less suitable for humans to program them directly, without the help of a compiler. Not that I'm into assembly language programming, but the PowerPC instruction set architecture at least allows you to make a mental model of how/what you're trying to accomplish with your program. I can't say the same about the more RISC-ish instruction set architectures I've had to work with.

@Carl-S

Wow. Carl. Hi!

Good to see you here. Sorry everything's such a mess... Not quite how you left it.

I agree with everything you've said although I do believe (as dull and quantifiable as hardware is) that killer hardware was key to the Amiga's original success.

It needs it again to make a full comeback.

Quote:

would love to see the Amiga return someday as a force in the computing arena. I actually think it is possible, as insane as that may sound. But, it's going to require a clear vision and a grassroots unity of purpose.

...and a massive, enormous pile of cash.

_________________
Sam

@olsen

Quote:

Given a choice, I'd rather tell it as PowerPC instructions than in x86/x86_64 instructions, having to make assumptions about how the register renaming may take place, how the pipelines will be utilized and how the cache hierarchy may come into play. I would prefer as much transparency in the instruction set architecture as possible. And that simply isn't there in the x86 world.

Factor in Itanium (X86/IA-32 HW compatible) i.e. Itanium exposes it’s internal ISA while still retaining cut-down X86 front-end (in-order).

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@olsen

Quote:

Definitely. But the "cleaner" architectures tend to be less suitable for humans to program them directly, without the help of a compiler. Not that I'm into assembly language programming, but the PowerPC instruction set architecture at least allows you to make a mental model of how/what you're trying to accomplish with your program.

What’s your target market?

For most cases, C/C++ or any higher level language with a competent IDE/compiler would suffice.

Last edited by Hammer on 24-Oct-2006 at 02:26 PM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@Hammer

Quote:

Hammer wrote: @olsen Quote: Definitely. But the "cleaner" architectures tend to be less suitable for humans to program them directly, without the help of a compiler. Not that I'm into assembly language programming, but the PowerPC instruction set architecture at least allows you to make a mental model of how/what you're trying to accomplish with your program. What’s your target market?

In that particular case I was wrestling with a DSP. The DSP won, I went away with the insides of my head all screwed up "Manual" VLIW programming is not for everyone.

Quote:

For most cases, C/C++ or any higher level language with a competent IDE/compiler would suffice.

It ought to suffice. But today's languages are not necessarily well-matched with the hardware the code they produce is supposed to run on. Where this mismatch is slight, good for you, but where it's not so slight you ought to be able to let assembly language code help you. And that language ought to give you better control over what the instructions produce than the high-level language does.

@Derfs

Feel free to respond to what I'm saying, instead of posting your standard reply as soon as you see a post signed by me ....

@Carl-S

It was with delight that I watched/listened to you at AmiWest and what a joy to see you posting here. I voted YES.

You are one of the original visionary people and that is exactly what is needed now. A BIG PICTURE view as seen through the eyes of a visionary! I hope you can “juice” this effort again and make your “baby” thrive once more!

OS4 is a nice piece of work but we have bogged down ...again!

I run the Linux distro of Ubuntu on my old AMD system purchased in the “way back” for the new Amiga system slanted for x86....sigh! I'm not much of a geek and Ubuntu is amazing, it “just works” but when I try to look under the hood it is "still dark and mysterious" there.

Your approach would have wide appeal and I hope this is the start of something new and exciting.

_________________

@Hammer

Quote:

Hammer wrote: Note that PowerPC ISA is not the "cleanest" RISC ISA.

Perhaps not, but who gives a damn? PPC is what OS4 runs on, and that defines the prerequisites for any discussion that includes both of the words "Amiga" and "Future".

The underlying architecture doesn't really matter; the key thing is to make Amiga available *cheap* and in *broad scale*.

You could add together Carl's two last points into this single word: "Community". That is key, not CPU. It's *the people* that will make a difference ...

Amiga is (perhaps forever) a niche platform.

IMHO it is not very suitable for desktop use when you have in mind what people expect from the OS today in terms of security, multiuser etc.

A complete OS4 will exel in ease of use to all Amigans and newcomers, my parents were able to learn how to use the Amiga much faster than WinXP for example. So thinking about what we have now (OS4.0) perhaps we should target existing Amiga users and people new to computers.

IMHO Takemehomegrandma is right if you read the comment he links to in his last post. Link is called: "this might be a start".
A very interesting read if you don't mind the Efika advertising

_________________

@takemehomegrandma

Quote:

and that defines the prerequisites for any discussion that includes both of the words "Amiga" and "Future".

No, in this thread it doesn't. We're not specifically talking about the current AmigaOS nor the related intellectual property, but rather an hypothetical system which would fall into the definition of "Amiga" given by Carl.

_________________
“It is well enough that people of the nation do not understand our banking and monetary
system, for if they did, I believe there would be a revolution before tomorrow morning.”
~~ Henry Ford

@Derfs

Quote:

Derfs wrote: @takemehomegrandma will you please stop plying your wares .. its getting so old. plus it still isnt that cheap. why does tmhg remind me of a phones 4 u salesperson?? /me ponders

If you have something to contribute, please do so. Otherwise, leave the personal attacks out of the thread. The only non-abusive comment you made above was simply, 'It still isn't that cheap.' The rest was pure flaming/baiting.

Quote:

Flame/Attacks: Do not flame! Flaming refers to derogatory, abusive, threatening, sarcastic, rude, or otherwise mean-spirited messages directed at members/users. Be cautious when using sarcasm and humour. Without facial expressions and tone of voice, they do not translate easily over the Internet in posts and may be perceived as flaming. Report the flame posts immediately to an Amigaworld Team Member so that the situation can be dealt with immediately. Repeat offenders will be banned.

In the event you feel TMHG, or any other user, has violated the TOS of the site, please use the Abuse Report button and Staff will investigate and make a determination.

Last edited by wegster on 24-Oct-2006 at 04:28 PM.

_________________
Are we not done with the same silly arguments and flames yet??!