@all

I haven't read the whole thread but it lookslike Dave Haynie does spilt the Amiga camp! Pro X86 and Pro PPC.

For classic Amigas it was only logical to go the PPC way. Even after C= went bankrupt all new attempts to build new Amigas based on PPC.

If only one developer or manager (ESCOM, DELL, AMIGA, EYETECH...) would have said we will go X86, the Amiga community would not have supported them!

So the only logical for Amiga, Eyetech, Trika and Hyperion was the PPC way!

Now that Dave said X86 would have given us a brighter future (where he's probably right) and that Apple clearly showed they achieved the swicth in less than 6 months and the Apple community first weren't happy with the switch, now after the new presentation, don't want to go back, suddenly seem to open many of the Amiga community they eyes and they all want to go X86 too.
(BTW the PPC-Intel-switch-in-6-months was only possible because Apple and Intel did have over 1000 engineers working on it!!!!)

Why do you all behave like a flock of Sheeps? Someone cries "Go X86" and they are all running after!
I'm sure Hyperion did check the X86 situation and came to the conclusion that only the PPC would be viable for them!
With PPC they do have the Amiga community behind them (until this thread about Dave Haynie telling his opinion and Apple achieving the PPC-Intel switch)
AOS4 on X86 would be one of the hundreds other OS for X86! And these OS are only used by some freaks! The same would happen to AOS4 on X86, only we and some used-to- remember-Amiga users would try it on their PC.

Also forget the gameconsoles with PPC, even if AOS4 will run on one of them it won't be sold!

But PPC are nearly build in every other device out there: coffemaschines, cars, TV, settop boxes, well million devices!

If Hyperion can get AOS4 running on only one of these devices (some million TV settop boxes) they will spread AOS4, but not like most of us want to see it!

But for them it's the only way to survive!

For me I will wait for update 4 and final version and my A1XE will mailny be used for number crunching, PageStream DTP, email system. And I'm hapy with it. If it's PPC or X86 or Alpha or whatever based is really of no importance to me!
For all other applications I will use my Macs or PDAs!

_________________
A computerworld without MS products and Windows!
Connect your Amigas ...
...The Red ONE-A1XE G4 - A3000T- A3000 - A4000 - A2500- A1000 - A600 - CDTV - CD32...
and your PDAs and laptops ...
Psion 5mx Pro - Psion NetBook - Apple iPhone - MacBook Pro

@minator

Quote:

Aside from having to learn multithreading (which everyone will have to do anyway) programming the SPEs should be pretty much like using AltiVec, you can also use them as scalar (normal, non-vector) processors though.

Additionally the programmer has to take care of SPU’s memory model.

In my question(to you) about the number Intel’s multi-core processor designs. You forgotten to mention that Intel has at least 15 multi-core processor designs in R&D i.e. one of them was a co-processor arrays (i.e. Intel is still spec’ing SSE4).

Last edited by Hammer on 14-Jan-2006 at 10:53 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)

@mbilla

I'm pro-both. It's easy to distribute binaries for both processors...

@mbilla

Bandwagon jumping... flock of sheep...?

You'd think this was the first time any of us have thought about this. Some people (wisely) keep quiet most of the time to avoid having the discussion.

@Hammer

perhaps not 7 years, but an advantage none the less. Technology moves faster forward now than 20 years ago, so it would be silly to think you could get 7-10 years lead today.
Of course the x86 world isn't standing still and will eventually catch up with the CELL in some way, but neither is CELL standing still, the CELL 2 is already in the pipeline with the renewed deal between the STI companies. And of course it would be possible to put many CELLs in the same computer, as I understand it they have an exponental factor when using many chips. This could be more or less like the planned Transputer in the late 80's.

Perhaps the CELL isn't the best allround chip, but it should outperform other current chips easily in multimedia and f.ex. 3D-rendering applications. This would put Amiga back to square one, where it all started from.
I can see many companies interested in making 3D-products for the CELL for obvious reasons... speed. This could be Amigas ticket back into the TV and Hollywood business. The games side would be a lot trickier, but perhaps it would be enough if it could play PS3 games, even on a desktop version. Sony makes money on games sold, not harware so this shouldn't be a problem.

Amiga could then also become the ultimate develompement environment for PS3 games... that would also bring developers for the Amiga-side of things, wouldn't the gamesmakers avery now and then have the urge to make some new tools or commodities for the chosen operatingsystem? :)

As I said earlier, if I were Hyperion, this would be my primary objective. Surely, embedded products could also become a cashflow, but it doesn't really promote the OS used.

-Vikke

@mbilla

Quote:

mbilla wrote: With PPC they do have the Amiga community behind them (until this thread about Dave Haynie telling his opinion and Apple achieving the PPC-Intel switch) AOS4 on X86 would be one of the hundreds other OS for X86! And these OS are only used by some freaks! The same would happen to AOS4 on X86, only we and some used-to- remember-Amiga users would try it on their PC.

Without commenting how easy it would be to make the actual port, I would speculate that - based on what the remaining community has already been through - people would follow AOS to X86, no problems. Hell, people can just put an 'Intel outside' sticker on the box if it eases them . And even better, the PPC could co-exist with X86, if I've understood correctly the comments from more technically advanced komrads.

So instead of loosing users, the community would gain some more (as you said yourself) - first freaks (developers!) as you say, then more aware end users who are tired with Windows. I would count myself somewhere there.

These freaks (with no prior Amiga history) will not buy PPC hardware just to try AOS, they continue fiddling happily with other alternative OS's on X86.

If the strategy is to develop AOS only for embedded markets**, I believe the community will faid away gradually. No need to be enthusiastic about a washing machine, right?

If the price of X86 conversion is not too high for the understandably small resources of Hyperion, I would say there is more to win then to loose in X86 world. The possible licensing issues with Amiga Inc could be ironed out with them with pure business case logic?

It is not too late, Apple will show to the rest of the world (outside planet Linux) that there are options available for Windows. It will be an eye opener to the Joe Average. Amiga would survive with a fraction of the X86 user base.


** Actually I believe the embedded markets is only the first step.

Pasi

Ok. That's enough with speculation on my part.

_________________
Huh?

@mbilla

Quote:

mbilla wrote: @all I haven't read the whole thread but it lookslike Dave Haynie does spilt the Amiga camp! Pro X86 and Pro PPC. For classic Amigas it was only logical to go the PPC way. Even after C= went bankrupt all new attempts to build new Amigas based on PPC. If only one developer or manager (ESCOM, DELL, AMIGA, EYETECH...) would have said we will go X86, the Amiga community would not have supported them! So the only logical for Amiga, Eyetech, Trika and Hyperion was the PPC way! Now that Dave said X86 would have given us a brighter future (where he's probably right) and that Apple clearly showed they achieved the swicth in less than 6 months and the Apple community first weren't happy with the switch, now after the new presentation, don't want to go back, suddenly seem to open many of the Amiga community they eyes and they all want to go X86 too. (BTW the PPC-Intel-switch-in-6-months was only possible because Apple and Intel did have over 1000 engineers working on it!!!!)

And we seem to need 1000 engineers to work on hardware so that we get
AOS released. So what is better ?

Quote:

Why do you all behave like a flock of Sheeps? Someone cries "Go X86" and they are all running after!

That's what it looks like to you. I would like to ask why are we not allowed to
bring this into discussion (this time via Daves old statments) without beeing spit upon ? Why do you have to take it so serious?
Hyperion is not going to change their mind so what's to worry about. Just enjoy the disucssion.

Quote:

I'm sure Hyperion did check the X86 situation and came to the conclusion that only the PPC would be viable for them! With PPC they do have the Amiga community behind them (until this thread about Dave Haynie telling his opinion and Apple achieving the PPC-Intel switch)

But if the amiga community is 2000 users what good is it going to do hyperion ?

Quote:

AOS4 on X86 would be one of the hundreds other OS for X86! And these OS are only used by some freaks! The same would happen to AOS4 on X86, only we and some used-to- remember-Amiga users would try it on their PC.

So then it would be 2000 + ...

Quote:

Also forget the gameconsoles with PPC, even if AOS4 will run on one of them it won't be sold! But PPC are nearly build in every other device out there: coffemaschines, cars, TV, settop boxes, well million devices! If Hyperion can get AOS4 running on only one of these devices (some million TV settop boxes) they will spread AOS4, but not like most of us want to see it! But for them it's the only way to survive!

Then I am here for the wrong reasons.

Quote:

For me I will wait for update 4 and final version and my A1XE will mailny be used for number crunching, PageStream DTP, email system. And I'm hapy with it. If it's PPC or X86 or Alpha or whatever based is really of no importance to me! For all other applications I will use my Macs or PDAs!

And I don't mind using XP so I probably stick to that allthough I rather would use
Amiga again one day.

_________________
AmigaOS or MorphOS on x86 would sell orders of magnitude more than the current,
hardware-intensive solutions. And they'd go faster.-- D.Haynie

@minator

Quote:

Cell is already on the market, see Mercury's announcement last week.

A blade? What's the cost?

Quote:

By the time x86 can go quad core you will be able to put 2 Cells on a chip and they'll scale better.

Intel's quad-core Kentsfield is already been taped out in 2005.

_________________
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)

@Vikke

Quote:

perhaps not 7 years, but an advantage none the less. Technology moves faster forward now than 20 years ago, so it would be silly to think you could get 7-10 years lead today. Of course the x86 world isn't standing still and will eventually catch up with the CELL in some way, but neither is CELL standing still, the CELL 2 is already in the pipeline with the renewed deal between the STI companies. And of course it would be possible to put many CELLs in the same computer, as I understand it they have an exponental factor when using many chips.

Both AMD and Intel already stated that Cell is a good idea, except for compromised control CPU. AMD wouldn’t have licensed Rambus technology (e.g. XDR) if that was not the case.

Quote:

Perhaps the CELL isn't the best allround chip, but it should outperform other current chips easily in multimedia and f.ex. 3D-rendering applications. This would put Amiga back to square one, where it all started from.

RT Raytracing is already reachable in 2005 era G70.

_________________
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)

@mbilla
Quote:

I haven't read the whole thread but it lookslike Dave Haynie does spilt the Amiga camp! Pro X86 and Pro PPC.

1) You should read the whole thread (worthwhile reading, at least most parts of it)
2) Unfortunately, the Amiga camp is already split
3) There is no "Pro X86" vs. "Pro PPC". Most of us just regard x86 as an additional option. Nobody has to throw way their PPCs. (hint: better sell them to me )

Quote:

If only one developer or manager (ESCOM, DELL, AMIGA, EYETECH...) would have said we will go X86, the Amiga community would not have supported them!

Maybe. At least there would have been a little turmoil since the "x86=evil" was even more prevalent at that time. Look at Apple. Some geeks and hardcore users felt insulted because its against their religious principles but the others just don't care I guess all those running Winuae, AROS, Amithlon also don't really care about it.

Quote:

Now that Dave said X86 would have given us a brighter future (where he's probably right) and that Apple clearly showed they achieved the swicth in less than 6 months and the Apple community first weren't happy with the switch, now after the new presentation, don't want to go back, suddenly seem to open many of the Amiga community they eyes and they all want to go X86 too.

You seem to make it appear as if supporters for "Amiga on x86" were non-existant before - there were lots of them (again: Amithlon, Winuae, AROS). I agree that they are much more perceivable on AWN at the moment. I don't wonder why. What do you think would have been the reaction when someone has opened a thread here 1 month ago (before the IRC chat) asking "hey, why don't they just port it to x86?". Probably something like "Hyperion have stated their reasons why it will never happen. And it will never happen. Stop trolling. Period."
There is not really a "bandwagon" people are jumping on. I only see people sitting inside the warm house and people waiting outside for anybody opening the door. Since there is nobody opening that door (hardware, if you can follow), they are looking for a loop-hole to get in. It's not about which house it is, it's just about getting inside (ok, and not paying too much entry I must add ). So what I'm trying to say is, ahem, I mainly agree with you. I won't care on which platform it runs. If we see a PPC mainboard soon, I will likely buyt it, just because I am desperate. If there is OS4 on x86, I will likely prefer it (I guess there is no arguing why).

Quote:

But PPC are nearly build in every other device out there: coffemaschines, cars, TV, settop boxes, well million devices!

Personally, I am not looking for OS4 on either one of those devices, rather I simply want a desktop (I know, so "old-fashioned" at the moment). Why do you think that these "million-device solutions" for OS4 will be more likely to happen than getting a simple, non-buggy, mainboard for it? The whole PDA/Embedded-market topic has not been looked at from the start. There is not even Java for OS4, so which chances would it have on these devices, realistically?!

Quote:

But for them it's the only way to survive!

Honestly speaking, if the embedded and PDA market is the only way to survive for them, then why wouldn't they have done it in the first place and skipped the desktop stuff (which admittably did not work too well until now)? Why would they have developed OS4 for a desktop market, which doesn't supply enough revenues due to hardware constraints/unavailability? I just don't know what their business plan is, so I can't argue about their decision for a proprietary hardware platform. Also note that it was a decision taken a long time ago; they might question it themselves, but at a certain point it's too late for CTRL+A+A - you will finish what you have started - and you will defend your decision no matter what the reasons are. What else can you do?

Last edited by hatschi on 14-Jan-2006 at 10:47 PM.

Wow this post has grown.

True, there is a good point about those of us that have hardware already. I have more then a couple friends who want the new Amiga yesterday. I totally can see you guys point on this! Someone needs to step up to the plate and fast. I just hope you guys buy the new motherboards when they become available.

Second, I am in the camp that believe the cell is the way to go for us. Could you imagine OS 4 on that thing, wow! It would bring us back to where we should be. Honestly, I think the current eyetech machines are developer machines or a stepping stone to the next generation. If I were to put my money where my mouth is, I would bet on cell! Probably wouldn't take a whole lot of time getting os 4 running on it.

Last but not least, I plan to buy the new motherboards when they are available to replace the eyetech machines. As mentioned with the Amy, I like the loss of legacy ports personally.

Yogi

@Hammer

Quote:

Notice the performance cloth simulation on PPE's VMX vs Pentium IV i.e. less than 10 percent. The G5?s VMX is quite competitive against Pentium IV. Clock speed alone doesn?t equal performance.

Do you really think a 800MHz G3 or G4 with a slow memory bus is going be in the same ballpark as a 3GHz+ PPE with a 25GB per second bus?

...and why do you keep talking about the cloth simulation, to quote the first line:

"This technology demonstration shows a prototype of a next generation cloth solver algorithm. It is a proof-of-concept..."

Hardly a good reference for judging performance.

The paper can be found here if anyone's interested.

They don't say anything about the PPE version so I don't know what's going on there.

Quote:

SPU?s main weakness is double precision floats and IEEE 754 compliancy.

Weakness is something of a relative term given it's Linpack result is 10GFLOPS for double precision, dual core AMD64 should be around the same.

Quote:

You can't run general purpose code on SPU.

Yes you can, you can run whatever you like. It wont like some sorts of algorithm but there are several ways you can solve that, not least using a different one.

Quote:

X64 cores is directly competition with PPE, while GpGPU is against SPU. PS; Both Opteron and GpGPU also scales nicely.

A Cell system will likely have a GPU so the CPU will be against a CPU.

Quote:

Additionally the programmer has to take care of SPU?s memory model.

You do something very similar in AltiVec and most likely SSE if you want good performance, they don't have a local store so you have to indirectly manage the cache instead.

Getting maximum performance on *any* processor will involve the same techniques you'd use on Cell.

_________________
Whyzzat?

@hatschi

"1) You should read the whole thread (worthwhile reading, at least most parts of it)"

The last 2+ hours I have done so, after posting my opinion and seeing that Dave Haynie is answering in this thread!

_________________
A computerworld without MS products and Windows!
Connect your Amigas ...
...The Red ONE-A1XE G4 - A3000T- A3000 - A4000 - A2500- A1000 - A600 - CDTV - CD32...
and your PDAs and laptops ...
Psion 5mx Pro - Psion NetBook - Apple iPhone - MacBook Pro

@mbilla
Quote:

he last 2+ hours I have done so, after posting my opinion and seeing that Dave Haynie is answering in this thread!

Yep, quite a surprise, heh? T_Bone was even offering his wife to him All in all it's an interesting read, I suggest to read it before the first ppl will start with something like "It was only x86 BS stated in it anyway. And then these [add term]-fanboys came in. Why not lock/delete it?"
Hmm, what I am trying to say is better read this one than the "Counting thread".

@mbilla
Like yourself I've now wasted some time reading most of this thread. Sigh....

I see Dave has joined the discussion and in "demi-god" like fashion believes he is above the rules of this sites and has no qualms about personal attacks, name calling and generally believing his view is correct and us users (cavemen ) know nothing.

Now as I'm a lazy (caveman ) I will cut and past a reply from another topic which is relevant to this topic too.

-----------------------

Just as some in the "other" thread don't seem to get it, currently AmigaOS doesn't stand ANY chance on the desktop no matter what CPU is used.
CPU and H/W platform does matter, and I'll give my reasons later.

Introducing AmigaOS only on the desktop currently is a dead end with no future prospects. It will of course be useful to us hard core users and some old users wishing to see what is going on. But in a very short time with not many up to date apps, interest for old users will wane, and they will leave. With this the market will gain a slight "Blip" in sales and then stagnate and die.

What I believe Hyperion is trying / hoping to do is release AmigaOS an a PPC platform to the hard core users and some old users so they have a desktop / DEV. system with which to create APPS. for a different market. i.e. PDA's, cell phones, STB, embedded devices, etc. The reason for this strategy is to again build up a brand name, and have some market penetration. NEW users of these devices will become used to the easy way AmigaOS works and then possibly be interested in a desktop in the FUTURE.

Why PPC CPU and "lock in"!
In Commodore - Amiga's days there was absolute "lock in", because one vendor provided the H/W and OS'es (AmigaOS & Unix). Money was made on the package. No one complained back then about H/W "lock in". Commodore made money on the H/W and OS, plus sales of OS upgrades. The OS upgrades were even "locked in" to PROM upgrades! (This I admit was over-kill, and a logistics night-mare)

If the product was good sales were guaranteed and profits were made with close to ZERO loss due to pirating. (pirating of Apps.& Games, that was a whole other story)
Commodore & Amiga failed due to bad management with no vision or direction, which in the end produced bad and unsupported product.
The failure was NOT due to "lock in".

H/W "lock in" is a way for a company to protect their IP and "product family" from been diluted by customers and opposing market segments.
Huh, what PR babble is this!
Look at the current example of the X-BOX. MS's box builder used almost standard x86 PC parts for the X-BOX. Hackers broke the protection (customers) which allowed pirating of S/W and the Linux (opposing market segments) crowed bought the H/W to run Linux at MS's expense and in direct opposition to MS.

I believe MS has learned from this and that is the reason they are using the PPC in the X-BOX 360. Better protection on the M/B and better "lock in" due to the obscurity of the PPC.

If Amiga wishes to enter the desktop market in the FUTURE it must have protection in the form of H/W "lock in", and as Apple has just stepped out of the PPC arena, this could be an uncontested market segment.

To those who run duel or triple boot systems, this might be good and fun for you and I, but to the companies providing OS's this is bad business and dilutes profits unless they are the company providing ALL of those Os's.

Firstly Hyperion's AmigaOS4.0 needs to make some money in OTHER markets before the desktop, then when they deem it safe and viable, enter the desktop market with a platform that provides "lock in" and protection.

Logically to be viable in ANY market the price must be {reasonably} competitive and performance adequate, but preferably market leading.
Ignoring the "lock in" issue, if x86 was used, price might match the competition, but performance will NEVER have the POSSIBILITY of besting the competition. (Apple will have to hire a new PR team)

So don't even think of selling your AmigaOne DEV. systems.
There is work to do NOW and in the FUTURE.


Cheers,
Tim

@mbilla

Quote:

The same would happen to AOS4 on X86, only we and some used-to- remember-Amiga users would try it on their PC.

And how is that a less desirable outcome than the current situation, which excludes all the used-to-be users, as well as a fair number of the "we" which just can't get or afford the hardware to "try it"?

Yes, OS4 probably has a 90+% market share on AmigaONEs (not the least because nothing else will run properly on the hardware...). But Asustek alone produced 30,000 to 40,000 times more motherboards in 2005 than Eyetech has over the lifetime of the project. ECS, Gigabyte and MSI between them sold more motherboards than Asustek.

Think of it in terms of salt. Grab a 5 kilo bag of salt, and let it stand for those 4 manufacturer's 2005 production of x86 motherboards. Now grab a teaspoon of salt, and drop it onto the kitchen bench. Divide it into two halves, and discard one half. Repeat for a total of 6 halvings. What's left now roughly represents the current lifetime hardware market for OS4 (It's actually a bit high, but I guess the CSPPCs come into it, too :).

@minator

Quote:

Do you really think a 800MHz G3 or G4 with a slow memory bus is going be in the same ballpark as a 3GHz+ PPE with a 25GB per second bus?

At least G4’s VMX performance is not less than 10 percent of Pentium IV.

Quote:

Weakness is something of a relative term given it's Linpack result is 10GFLOPS for double precision, dual core AMD64 should be around the same.

Estimated around 11.2GFlops for 2.8Ghz Athlon 64 FX-62 (assuming without improvments for stepping F.)

Linear algebra is not a complete system test. Linpack is usually not use in desktop benchmarks i.e. it's margin (in G5) doesn't reflect desktop applications.

Pure DP FP performance is basically a domain of Clearspeed.

Quote:

Yes you can, you can run whatever you like. It wont like some sorts of algorithm but there are several ways you can solve that, not least using a different one.

As mentioned in the cloth simulation test, the generalize code has to be re-structured.

Note that original code was scalar and optimization was only applied for Cell not for referenced Pentium IV (running original code) .

Refer
http://www.research.ibm.com/cell/whitepapers/alias_cloth.pdf

Using 8-way SPU’s 5 time over 1-way Pentium IV is unwised since this didn’t factor in Intel’s Tigerton.

In "Porting to the Cell Architecture"
Quote:

Porting to the Cell Architecture The most notable features of the Cell processor architecture are the IBM® Power-based RISC CPU core (PPE), the 8 SIMD vector units (SPEs) and the impressive bandwidth that can be leveraged to obtain maximum performance from the architecture. In order to port the prototype cloth simulator to the Cell processor, the code and data structures had to be restructured. First, changes had to be made to allow for parallelization of the algorithm(1). Even for more conventional multi-core targets, this is a challenging process. With the Cell architecture, there is the additional requirement to manage memory transfers between the main processor (the PPE) and the SPEs. Second, where feasible, scalar code was re-written to take advantage of the vector support in both the PPE and the SIMD units.

It seems that Pentium IV’s code was not optimized, since the original code needed to be restructured for vectorization and parallelization i.e.
such optimizations also benefit Pentium IV side (1) i.e. for SSE2/SSE3 and SMT.

Quote:

A Cell system will likely have a GPU so the CPU will be against a CPU.

Where’s the GPU array?

Quote:

Getting maximum performance on *any* processor will involve the same techniques you'd use on Cell.

There’s level of suitability in certain types code i.e. other “CPUs” doesn’t have to deal with massive threading models to gain maximum performance. A two thread model doesn’t automatically scale to four or eight.

Last edited by Hammer on 15-Jan-2006 at 07:46 AM.
Last edited by Hammer on 15-Jan-2006 at 06:21 AM.
Last edited by Hammer on 15-Jan-2006 at 02:20 AM.

_________________
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)

Don't forget the Apple have had OS X running on Intel since day one. The first public beta of OS X in 2000 or 2001 also was running on Intel boxes in Apple's skunkworks. Every release of OS X has also been updated on the intel box. And furthermore, what OS X is based on is NextStep (which became OpenStep). Originally on 68k black boxes, the hardware was dumped by Next and the OS ported to x86 in the 90s. All Apple needed from Intel were new logic boards to fit into an iMac or Powerbook (MacBook). That's why the transition only took 6 months (plus 1000 staff at Intel!).

I'm an Apple reseller technician, and dreading Apple releasing these mactels too early. Everyone other than Apple are still porting their apps, and Rosetta doesn't run a lot of things (especially apps that require altivec). Hopefully the early release will increase the speed that apps are being ported. I don't really trust Rosetta just yet!

Anybody noticed how quiet Hyperion are on this thread?

Also, thanks Dave for posting. You and Woz I think are they only two cool "big time" people left in the industry (Woz still posts on a lot of Apple forums). And the fact that both of you still work on cool projects, rather than have sold out. Not too many of those people left anymore!

@VidarL

Quote:

VidarL wrote: Bernd Meyer have posted a very interesting look at how he would implement 68k emulation in a x86 OS: http://flyingmice.com/cgi-bin/squidcgi/mbmessage.pl/amiga/125942.shtml Vidar

Yes, I saw that, and it's a good read. Some of the rest of the thread is sort of, umm, like some AW threads, but worth reading for those who are really interested instead of just the 'me too' type comments.

The idea of basically running under emulation for as many parts as are needed and chipping away as you go, but having a usable OS the whole time, is not so unlike what Hyperion did with the conversion from 68k to PPC native code..and as Bernd obviously knows the emulation overhead involved, ast least for 68k on x86, it's never a 'bad thing' to at least read his thoughts on a related subject.


@rest RE: Cell is different etc.

Yes, the Cell architecture is different than current desktop PCs. It's likely to be slower at some tasks, while excel at others. The 'desktop PC' today has effectively become a general purpose workhorse, rather than master of (m)any, while the Cell is more specialized in it's PPE/SPE design, but that does not mean it has no value.

Additionally, if IBM does manage to pull off their whole 'Cell architecture,' which is really an implementation of grid computing, where you can plug in additional nodes over the network (additional Cells), not only are OSes that work well with Cell going to be in demand, but developers that are writing code taking advantage of Cell will be as well. As Cell is not meant to be limited to consoles, but instead if successful, be nearly ubiquitous and be in everything, getting a 'head start' here just might not be a bad thing. Granted, Linux already runs on Cell, but looking at how long it's been for 64 bit Windows, and as large as the Windows code base is, there may very well be opportunites of some fashion there fro AOS, if ported and taken advantage of.

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

@T_Bone
An obvious (and often suggested/shot down, etc.)
solution for the very near term would be to try and
get OS 4 running on Mac PPC hardware as quickly
as humanly possible. I bet that Mac PPC HW is going
to become more affordable and much easier to come
by in the near term. The tougher question to ponder
would be will backward engineering Mac hardware be
more onerous than porting OS 4 to some equally
specific and unvaried X86 platform? I think it would
be a mistake to try and create an x86 specific OS 4
that does not have a very specific and narrow HW
target (it seems like there is to much HW variation
in the X86 world for very small developers to cope
with).

I think there could be an opportunity for some very
cool stuff in the realm of working with high definition
video on a cell based super Amiga, should there
ever be enough money and a HW platform to make
this happen. If you build it, perhaps they (new users)
WILL come.

@Hammer
Well, I think we only had stuff like VCD and Mpeg-1
in the last days of Commodore. NTSC DVDs were
still on the drawing board, if I'm not badly mistaken.
HD disc based content was not really on the radar
with regard to computers in 1994 - it was practically
sci-fi for most. Didn't Bill Gates once say something
about knowone ever needing more than 640k?