Amiga.org
Amiga computer related discussion => General chat about Amiga topics => Topic started by: dmac721 on November 19, 2004, 06:35:31 PM
-
Just thought I would ask "WHY" did they not just port Amiga OS to readily available architecture??? Wouldn't it have been much easier to sell software than thousand dollar hardware?
-
They did, it's called AROS.
-
You mean to PC´s ??? Like Linux tries to do ???
-
Which architecture? X86?
This has been gone over and over. Should be in the FAQ...
Well, from the base 68K Amigas we got faster 68K accelerators/CPU upgrades, then PPC accelerators. Then,well, PPC only software. To me, alot of the PPC accelerators being sold and in the works around 2000 seemed like almost complete motherboards in themselves. When you are seeing PPC accelerators that have their own memory slots and that can have a PCI board added. Put a video card and sound card in a setup like that and you've basically have two computers in one case. It seemed like a natural progression to just make a PPC board to replace that whole upgrade kludge mess. Not to mention on X86 you have to compete to some degree with Windows. I'm sure there are other issues.
-
Well I guess because the whole Amiga concept wouldn't be "different" enough to to warrant any sales, if it were to use x86 hardware.
-
Kinda agree (to a point) --
It would be great to see another operating system on X86 based platforms --
I have no love of MS -- and am tired of seeing everything feel that there is a Binary Choice (MS vs Unix/Linux)
Q is, the 68xxx codebase would still need emulation to run the other classic base of software and the development effort would still need to be there for applications. Sigh-- in a perfect world ....
BTW: Anyone else catch this --
Sun Microsystems is planning on Making Solaris 10 Open Source -- wahoo (can't wait to see if we can port a UAE to x86 running Solaris 10)
dB
-
KennyR on 2004/11/19 18:42:08
They did, it's called AROS.
Hey! that's my line :-D
I would like to point out that regardless of what CPU you OS runs on, you are always competing with windows. But by choosing the x86, you reduce the burden of cost on the user and allow them more options.
The choice of PPC was to lock the user into a specific and very expensive platform.
-
Don't listen to what anyone else tries to tell you, the real reason they didnt port it is so we could have threads like this. It's a global conspiracy.
-
Hum,
And here is the AROS Webpage (http://www.aros.org/)
@bloodline
are you having problems with your server?
-
On a positive note, OS4.0 will be available to Classic Amiga owners (A1200 and A3000/A4000) with Blizzard/Cyberstorm PPC cards. I wonder how well is this end going to be supported, as far as drivers for all available classic hardware go? E.g. Elbox products, Zorro cards, etc...
-
Hum,
And here is the AROS Webpage
@bloodline
are you having problems with your server?
No, it seems to be fine for me... but I am actually sitting in Starbucks reading (and writing) this on my pda :crazy:
-Edit- I would note that by going PPC one, in theory at least, can support the PPC cards fron Phase5 for the classic amigas... but only Karlos gets to enjoy that ;-)
-
bloodline wrote:
The choice of PPC was to lock the user into a specific and very expensive platform.
But also to avoid endian problems, the same problems that make it nigh impossible for AROS as it is to have integrated 68k emulation. Phase5's and so later Amiga's decision to use PPC wasn't just based on greed for a hardware market or anti-x86 religion.
-
That's just it, being a programmer I know that endian issues aren't really a justification. The next generation Amiga from Commodore was to be based on an a HP PA RISC chip. My biggest guess here is that it was done to keep us locked in to a motorola architecture and a lot of the hardware people involved were trying to stay competitive with Mac and following Apple's lead..
Anyone who believes the choice to stay away from intel was an endian issue obviously wasn't aware of the hardware politics at the time. The PC at the time wasn't advanced enough to give me Amiga competition, it's competition was the Mac.. This is not the case today..
The Amiga OS needs procesor independence to survive and thrive today.. Especially with 64 bit processors out their now (LOL okay the powerPC is 64 bit, but is the OS that supports it on the Amiga??)
KennyR wrote:
bloodline wrote:
The choice of PPC was to lock the user into a specific and very expensive platform.
But also to avoid endian problems, the same problems that make it nigh impossible for AROS as it is to have integrated 68k emulation. Phase5's and so later Amiga's decision to use PPC wasn't just based on greed for a hardware market or anti-x86 religion.
-
DonnyEMU wrote:
That's just it, being a programmer I know that endian issues aren't really a justification.
They're not a problem until you try to integrate legacy support with your new OS, like loading 68k shared libraries and using them with x86 applications.
-
Didn't the AROS guys say they were going to integrate UAE for running classic software?
If they can get far enough, be somewhat compatable with OS4s new APIs, and add enough presentation polish, they could potentially give OS4 a run for its money. Without costing anything.
I want OS4 to succeed, but its a LOT of money. The fastest PC running AROS could kick an A1s ass at the moment, and you know it.
Ok. Scampering back to my happy place now. (http://amigaworld.net/images/smilies/icon_spy.gif)
-
Sure, but integrating UAE isn't good enough to match OS4 and MOS in terms of compatibility. Imagine having to install a mirror copy of all your system files for 68k - MUI libs, datatypes, etc, just because they can't be loaded by native executables. And native libraries couldn't be loaded by 68k ones. The two systems aren't really "integrated", just attached.
You could for example run IBrowse on an integrated UAE, but without taking advantage of native Zune libraries. And you couldn't get it online anyway, because it wouldn't be able to load the native TCP/IP stack library (unless you use a wrapper like WinUAE does). And you'd have to emulate everything else like graphics functions too. It'd probably be better using WinUAE on its own if that's what you want. And this would not kick OS4's ass, not even on the fastest PC.
-
@BigBenAussie
Integrating UAE into AROS is not a trivial task. I'm not saying it is impossible but it sure would not be easy. Just integrating a 680x0 engine is tricky enough as you have all kinds of endian issues. The 680x0 emulation would not readily be able to easily share system structures with the native x86 code as it can on MorphOS and OS4.
Don't underestimate the power of the legacy emulation in OS4/MOS and the benefit of the native OS. I have tried UAE on some serious x86 kit and have ran system friendly 680x0 apps in OS4/Petunia on a mere 800MHz G4 with 133MHz FSB. I am absolutely not joking when I tell you it was the AthlonXP 2800 limping away clutching its' arse. I'm sure that other users can confirm that MorphOS' 68K JIT is similarly more than capable of giving UAE a good thrashing.
Even my lowly 603e 240 MHz can run 680x0 code on OS4 faster than an 060/66MHz (the old H&P 680x0 voxelspace demo for instance).
-
dmac721 wrote:
Just thought I would ask "WHY" did they not just port Amiga OS to readily available architecture??? Wouldn't it have been much easier to sell software than thousand dollar hardware?
So the Amiga OS didn't suffer the same fate as the BeOS?
You know, install it, boot it, play with it, remark "how quaint". Then reboot into windows and forget about it.
Where are you getting this "$1000" figure from anyway?
-
From the price of the A1 motherboards.
AROS shows what could have been, and it would have been a great OS (AROS is great, even though it's nowhere near complete). I think modern x86 architecture deserves a decent lightweight OS, and that's not mutually exclusive with OS4 or MOS (however that ends up).
-
Whilst an unashamed OS4 afficionado, I think it is fair for me to say that AROS is presently the most fun you can have with any x86 system natively :-)
-
Dudes, I said AROS(implying native) would kick OS4s ass, not UAE emulation. I'm just passing on what I read somewhere, and wasn't assessing its technical merits.
Besides, if you know something is 16bit you'd have to install it under UAE running on AROS, instead of it doing it automatically, and accept the limitations. My hope would be that they integrate the window system somehow, so it at least looks like its one entity.
I figure if AROS gets far enough, people could create ports for it, like they do for MOS, classic and OS4. Maybe they are already, and I'm just not hearing about it.
-
*moan* :oops:
windows cross bread with AmigaOs :-x
...........yeah bill gates would love that!!
he would have access to actual advancement,
but what a horrible cut 'n paste patch kludge afiar that would be!!! ewwwww!!! :-o
-
terminator wrote:
So the Amiga OS didn't suffer the same fate as the BeOS?
You know, install it, boot it, play with it, remark "how quaint". Then reboot into windows and forget about it.
As opposed to suffering the same BeOS fate of a proprietrary PPC machine that no one ever heard of, let alone got to use?
Also, the problem you describe is of the free version of BeOS - no one is suggesting that AmigaOS should be made free. If loads of people install it then forget about it, if they've all paid for a copy, who cares? That's still more revenue.
-
Argo wrote:
This has been gone over and over. Should be in the FAQ...
I can't wait for that FAQ to be ready. I guess everyone who posts a question that is covered by it will get the reply "RTFF!". Maybe there could be an icon for it like: :rtfm: or (http://www.forum4free.info/grafik/smilies/rtfm.gif)
:-D
-
From the to be written FAQ:
Q1283: "why didn't they port Amiga-OS to x86 ?"
Answer: "cos it would have been common sense"
:shocked:
-
Kronos wrote:
From the to be written FAQ:
Q1283: "why didn't they port Amiga-OS to x86 ?"
Answer: "cos it would have been common sense"
:shocked:
Well without getting into any kind of technical debate I have to point out that there are plenty of OSs available for x86 (http://dmoz.org/Computers/Software/Operating_Systems/x86/).
Give PPC a chance ;-)
-
Not again this.
1) Hyperion doesn't have a license for that.
2) Decision what hardware will get supported.
3) Probably a lot worse piracy problem.
I hope that you know Hyperion is working on AOS4 with very low budget. To support dozens of hardware options in x86 world would be...impossible?
-
Well AI cries "We need to have a cashflow" It would still be alot easier to sell a $99 software package than a thousand dollar system to people that aren't sure their money is a worthwhile investment.
-
X86 codebase ...?
How would you like to be infected today ?
Spyware ?
Adware?
Trojans?
virii ?
Please take your pick.
Me ? I'll stay codebase incompatible with all that Shyte?
If they wanna infect me .. They'll have to burn the midnight oil and re code their anal crap.
-
by Mr_Capehill on 2004/11/20 14:01:14
Not again this.
1) Hyperion doesn't have a license for that.
2) Decision what hardware will get supported.
3) Probably a lot worse piracy problem.
I hope that you know Hyperion is working on AOS4 with very low budget. To support dozens of hardware options in x86 world would be...impossible?
THANKS Mr_Capehill . . .thanks a LOT :-)
i couldn't agree more
-
>1) Hyperion doesn't have a license for that.
Which is probraly only so because they never asked for one.
>2) Decision what hardware will get supported.
You just pick one completly outdated Mobo and sell it at an extreme price .... oh wait, I got that one wrong ....
Amithlon prooved that one can support a great portion of x86 without major hassle.
>3) Probably a lot worse piracy problem.
Ah, so selling 1 copy is better than selling 2 copies + 5 pirate-copies ?
It was claimed that Amithlon sold ~2700 times in the short time it was on sale, thats about 3 times more than OS4 sofar. And inspite of being fudded and demonized by self-proclaimed "savers of the community"
Just imagine how many copies could have been sold if all forces had gathered behind, how many users and developers wouldn't have left....
And for the piracy, there are allways dongles, just as safe, but less painfull than restricting ourself to one line of medicore (you wish) mobos.
>I hope that you know Hyperion is working on AOS4 with very low budget.
Duh, they knew that when they started, and they should have thought of the consequnces instead of forcing another severe shift/split onto the community.
> To support dozens of hardware options in x86 world would be...impossible?
Did I allready mention Amithlon ?
-
:-(
aahhaaah!! sorry . . .bored . . .
-
>I hope that you know Hyperion is working on AOS4 with very low budget.
Duh, they knew that when they started, and they should have thought of the consequnces instead of forcing another severe shift/split onto the community.
I remember that Garry Hare(CEO of KMOS) mentioned that Hyperion have not asked for more money to support development. Its pure speculation, but I wonder if Hyperion has not asked for more money because KMOS would have tried to water down their agreement somehow, cutting them out of some potential profit on sales. Hyperion keep compaining that they haven't enough money to support faster development, yet they don't ask for money from KMOS, who obviously has some money to throw around if they go around aquiring companies. It can only be that they like the deal they have now and are looking at the big picture. Hyperion have mentioned that they are not able to port to different PPC architectures and that licenses have to be purchased from AmigaInc/KMOS. So ultimately their hands are tied when it comes to an x86 port, even if it was possible.
Ok. Scampering back to my happy place now. (http://amigaworld.net/images/smilies/icon_spy.gif)
-
I think it ultimately came down to this.....Sorry if there are historical innacuracies, but feel free to correct me nicely.
AmigaInc(or an earlier incarnation) well after the fall of Commodore and all the changes of ownership, wanted to produce a new OS, ultimately based around an AmigaDE hardware agnostic system. Amigans went ballistic and understandably believed an Amiga to be about hardware, and the original OS, not some new fangled software. So AmigaInc reluctantly agreed, but due to the animosity of Amigans to the x86 platform, probably because of years of platform warfare, they couldn't take the x86 route. AmigaInc were berated when even the dev platforms for AmigaDE came out on x86. So AmigaInc went PPC and contracted Hyperion to produce a PPC AmigaOS. Of course they then made claims that it would be done very quickly....but I digress.
Also, with users crying out for speed increases on classic hardware, various companies created speed-up boards based on the PPC chip, because in those days they seemed like the natural progression to the Motorola 68k line. Also the whole simplicity of RISC computing seemed in line with the architectual proficiency of the AmigaOS. The x86 hardware, considered CISC(Although its apparently more RISC these days) was/is considered complicated bloat. So it made sense from a lot of people's view to go PPC and AmigaInc went so far as to support the Amiga community that had purchased these PPC boards as well, as OS4 will do.
Add to this what was going on in the tech industry at the time the decision was made. The IBM/Motorola/Apple alliance were going to create a common platform to rival the Intel platform, but later had a falling out as they couldn't agree on a common platform, which could have been leveraged by a new Amiga. It probably couldn't be forseen that Apple's stake in the PPC meant they get all the newest chips first, and due to the chip shortage would means a PPC Amiga would always be a generation behind Apple, and thus a generation behind x86 too.
The problem, as I see it, is not a matter of cost perse as the big box Amigas were always fairly expensive, but rather one of not being able to keep up with x86 hardware advancement, especially in the games advancement. Games sold the Amiga. Amigans wanted kick-ass hardware to run games and we simply haven't got it. AmigaOS is fine, and the hardware is fast enough to run it nicely, and we'll have a great user experience. But I believe the true essense of the Amiga was its advanced hardware, as it was designed as an advanced games console, that could be used as a computer.
The problem with the current line, like the A1s is that it is much like a PC, even coming in mobo form, and underpowered at that. Unfortunately it takes vast amounts of R&D to produce a kick-ass games console, that would be an Amiga computer. I think that is what we've all signed up for, with the PPC direction, but we never got the kick-ass hardware. With kick-ass games playing hardware you could justify the price.
The only way we can get there is if KMOS licenses a console manufacturer's hardware and ports AOS to it, rebranding it as an Amiga. Failing that, they could make a deal with a 3d card manufacturer to produce that killer integrated 3d games playing Amiga system. I know a lot of people don't care much for 3d, but that is what is the most impressive thing at the moment, and the only real way to Amiga's rebirth. Hardware standardisation was the Amiga's strength and creating an Amiga without custom chips just seems wrong, even if we don't hit the hardware any more.
Ok. Scampering back to my happy place now. (http://amigaworld.net/images/smilies/icon_spy.gif)
-
@bigbenaussie
Well, you totally forgot one major part of the story. In 1997, Phase5 had already produced PPC boards for the Amiga and proved they could supply a viable hardware PPC platform for migration from 68k without having to drop legacy totally and suddently (which would be fatal for such a small software base). Other PPC solution announcements - amijoe, Brainstormer G3, the original Escena AmigaONE - turned out to be vapor and were never produced. The SharkPPC remains in limbo.
Amiga Inc was nothing really to do with the decision to go PPC. They chose to announce OS4 on PPC. This was more as an appeasement measure to hold their possible new market base together than any serious announcement, but they did have a reason in that bPlan (what remained of Phase5) did have software and hardware ready to become the new Amiga and OS4. This later became Pegasos and MorphOS after negotiations between AInc and bPlan broke down. Certain figures then decided it was better to fatally split the community than allow these non-Amiga branded solutions to become the new Amiga, even if in doing so it meant virtually no profit for anyone. Think about it - they did this really for Amiga Inc, who have done NOTHING for the community. Sad.
And on the related subject Amiga Inc. probably didn't choose Amithlon or AmigaOS XL to go on to become OS4, simply because they could never have licenced x86 boards and got their cut the same way they could with proprietary hardware. That's all, really.
-
KennyR wrote:
@bigbenaussie
Well, you totally forgot one major part of the story. In 1997, Phase5 had already produced PPC boards for the Amiga and proved they could supply a viable hardware PPC platform for migration from 68k without having to drop legacy totally and suddently (which would be fatal for such a small software base).
yeah and I've ALWAYS (even tho I own a PPC+Bvision) wondered WHY they went for the dodgie cache flushing context switching 68k/PPC :-? when they could have saved money + buyer cost + a faster board by having just a PPC chip with a 68k emulator????
you could have held down ...say F8 to disable the card for old games (non os friendly etc.) I dont think this would have been a bad thing :-D
-
yeah and I've ALWAYS (even tho I own a PPC+Bvision) wondered WHY they went for the dodgie cache flushing context switching 68k/PPC when they could have saved money + buyer cost + a faster board by having just a PPC chip with a 68k emulator????
That is pretty much MorphOS now.
But just 68k emulator is not adequate. Emulating 68k on BPPC with static emulator is pretty slow and without PPC native OS even worse (emulated gfx, mui, ahi...? no way).
-
AmigaDE is the new AmigaOS!
-
The_Editor wrote:
X86 codebase ...?
How would you like to be infected today ?
Spyware ?
Adware?
Trojans?
virii ?
Please take your pick.
Please, do not make demagoguery.
Keep in mind that those issues does not affect to *all* x86 OSes.
x86 is not Windows (as B.Gates pretends to tell us). Take the example from an OpenBSD/i386:
"Only one remote hole in the default install, in more than 8 years!"
Please, don't get confused and don't confuse the people.
-
Hagbar wrote:
The_Editor wrote:
X86 codebase ...?
How would you like to be infected today ?
Spyware ?
Adware?
Trojans?
virii ?
Please take your pick.
Please, do not make demagoguery.
Keep in mind that those issues does not affect to *all* x86 OSes.
x86 is not Windows (as B.Gates pretends to tell us). Take the example from an OpenBSD/i386:
"Only one remote hole in the default install, in more than 8 years!"
Please, don't get confused and don't confuse the people.
Right. Also, The_Editor should realise that spyware exploits holes in scripting languages like JavaScript, or via holes in ActiveX or even Java applets. None of these are CPU-specific. A security hole in any of these is still a security hole on x86, PPC, StrongARM, 68k... It has nothing to do with the CPU at all.
-
So the reason they went ppc is because some halfast company built and sold like a dozen add ons, and they are basing the fate of the entire Amiga community's future on an add on half working, half useful piece of hardware that nobody except for a select few could afford in the first place???? Right????
-
-
That about sums it up... there are technical merits for using a byte order agnostic CPU like the PPC... but they are outweighed by the cost benefits of the x86.
Guess it took me a while to get it, but once I realized that an _emulated_ '040 on a dirt-cheap XP mobile chip crunches numbers pretty much identical to a Pegasos G4, the advantage of something like AROS became pretty apparent.
(Talking from a strict "cost v performance" aspect here, I still miss my Pegasos and would own another if I won the lottery tomorrow, right after I bought my CLK-GTR..)
-
That "halfass" company was there for 10 years, making addons for the Amiga and Mac. They produced a fully
working (yet kludgy) solution. There were about 10000
boards sold according to Phase5 iirc.
-
the advantage of something like AROS became pretty apparent
ok, now imagine AROS running in 64 bit, at 4GHz ...
-
10,000? Somebody lied. If they HAD sold 10,000 there would still be 10,000 in circulation. What happened to them, did they all break down or did people just throw them away? Of course they didn't throw them away, not when they could have EASILY sold them to another amigan on ebay for a couple hundred dollars.
-
Argo: not to mention on X86 you have to compete to some degree with Windows.
You're competing with Windows if you do Windows-like things. The hardware is irrelevant. Microsoft makes many standards used on x86 boards, but they certainly don't own the architecture.
I've said it a million times: there really is nothing to stop Amiga from choosing a single, good x86 board, like an Abit, and using that as the official platform. Why people keep thinking you either have to make proprietary hardware for hundreds of dollars or sell software off-the-shelf is beyond me. Both extremist plans are doomed.
My impression is that most of the x86 Amiga fans have just left after the PPC announcement. Once you go PPC, you're stuck with it.
Bloodline: I would like to point out that regardless of what CPU you OS runs on, you are always competing with windows. But by choosing the x86, you reduce the burden of cost on the user and allow them more options.
Yeah. What ever happened to DE and Java? It's the 21st century... haven't we gone beyond specific hardware platforms? I wanted x86 for flexibility (for Amiga Inc., at least), and not for technical supiriority.
Remember why the sucky PC architecture became so popular in the first place. Choice, features, and value are more important to end-users than raw performance and tech specs, especially in a market where performance changes so rapidly.
Terminator: So the Amiga OS didn't suffer the same fate as the BeOS?
Be made lots of critical marketing errors, and were way too dependent on Apple clones. After Apple squashed the clones, Be really had no choice but to go x86 OEM. Frankly, I'm surprised they lasted as long as they did.
DonnyEMU: That's just it, being a programmer I know that endian issues aren't really a justification.
Not for data, at least. Many graphics and audio formats require a decoder to convert the endian order. Older formats like TIFF require you to manually set the endian order when you save files (!), but most modern formats do it automatically.
Many embedded CPUs don't care about endian order, and can work in any mode. x86 and PPC are still trying to fight it out for some stupid reason.
PS - Is it true that 64-bit float math is easier with little endian?
Mr Capehill: Probably a lot worse piracy problem.
At least it would *have* software to be pirated. How anyone expected a software industry, or even a close-knit community, to blossom from a few thousand hugely expensive boards is beyond me.
The Editor: X86 codebase ...?
How would you like to be infected today?
It's sad how Intel and AMD are actually adding new instructions to their CPUs to overcome Microsoft's terrible programming practices -- and people like yourself blame Intel for security issues.
The Editor: If they wanna infect me .. They'll have to burn the midnight oil and re code their anal crap.
Most infections/trojans happen because people are damned stupid and *LET* that crap on their machines. There's really no difference between installing an XPI plugin on Mozilla and ActiveX on IE if people are too dumb to realize that a pop-up that says, "Your Internet connection is not optimal!" is a scam.
I've been using Windows for years and have never had a virus, and the only "spyware" I get on my machine are cookies, which plague all web browsers and must be regularly (and easily) cleared out. How do people end up with five viruses and two dozen spamware programs on their computers? Stupidity, and the inability to read before they click "Yes".
I should offer computer classes for paranoid home users. I'd bet I'd get a hefty number of students with one newspaper ad.
Anyway, I just hope, if anything else, AmigaOS doesn't go out of its way to protect people from themselves. It makes real administration very frustrating. If Ben's "security through obscurity" policies are true, then I guess hardcore Amiga geeks shouldn't worry. :-)
I remember that Garry Hare(CEO of KMOS)...
You mean the Invisible Man, running an IT company without a website?
BigBenAussie: It probably couldn't be forseen that Apple's stake in the PPC meant they get all the newest chips first, and due to the chip shortage would means a PPC Amiga would always be a generation behind Apple, and thus a generation behind x86 too.
Right. Low-end embedded PPC chips are a dime a dozen, but if we want high-end chips, we get Apple's leftovers. Amigans should be more worried about competition from Apple than from Microsoft. Microsoft didn't get rich playing the hardware market -- they specifically avoided it.
Frankly, of all computers, being a generation behind *Apple* is a real embarrassment.
Games sold the Amiga. Amigans wanted kick-ass hardware to run games and we simply haven't got it.
Well, at this point, kickass games are really not possible. It costs millions to make top-tier 3D drivers, and consoles are always hard-coded so expandability is out of the question. I still remember those days when I didn't want to buy a CPU upgrade because all my older software (including applicaions) wouldn't work anymore. I don't want the Amiga to be another game machine, really. There's nothing really wrong with the current fleet of consoles, and they have little to no use for Amiga with tools like XNA over the horizon.
Now, getting rid of the GUI cruft that clogs a typical Linux system... that's what Amiga should be doing! I really wanted a QNX powered Amiga, especially since that company put a lot of effort and money into a new Amiga before Gateway screwed them over.
Oh well. Someone will get it right, eventually.
But I believe the true essense of the Amiga was its advanced hardware, as it was designed as an advanced games console, that could be used as a computer.
Hopefully, Sony's new "Grid" processor will be a real wake-up call to the industry. Game machines have always been far more advanced than PC hardware in all respects except I/O and CPUs -- and even that is quickly becoming debatable these days.
It's sad how a typical GPU has more transistors and performs more calculations than a P4 (by a longshot), and yet the P4 needs a cooler capable of handling 100+ watts of dissipation. My dad and I just put together a 3.2Ghz P4 with that damned "LGA" socket and a monsterous, fully exposed 90mm fan, and it was a sheer nightmare. A far cry from the Northwood P4s, which were truly amazing -- super silent and very easy to install.
BigBenAussie: The problem with the current line, like the A1s is that it is much like a PC, even coming in mobo form, and underpowered at that.
Correct. Take away the PPC, and you have a really expensive four-year-old vanilla PC. It pains me to see Amigans bash PC hardware so violently, because all these motherboards really just use the same standards. You can't shun the engineering genius and value in a typical Abit, Gigabyte, or MSI board with a staight face unless you're still living in the 80's.
-
9,995 are sat in a box waiting to be 'officially' fixed ;-)
-
@Waccoon
:lol: Heheh I'm glad somebody see's the world the way I do, what's wrong with people anyway?
@holley
rotflmao! That's what I said too :-P
-
Nobody lied. It is a very sensible number if you keep in mind that the first boards were made in 1997.
There *were* many Amigans back then. There are boards still in circulations, many dead boards (mostly
BlizzardPPC boards, due to the complexity of the board it's "easy" to kill it). Almost all the Greek Amiga users I
know still have PPC boards. Example is one who killed his board (a wierd story... He was fixing his tower and
accidently pulled a cap with it's trace.) and managed to find another board quite quickly. Also consider that both AmigaOS4 and MorphOS started their development cycle on these boards.
-
@Wacoon
Nice comment, I agree with most of it but there is a little problem in part of it. The endian issues are not that simple. The PPC *can* work in both modes, the x86 can't. On a little endian machine, you wouldn't be able to integrate native and emulated libraries the way it is done in MOS and OS4 right now, ie. you can just drop a 68k lib and it works in both 68k and PPC apps, without using endianess switching everywhere. All legacy code is big endian while native x86 code is little endian.
To everyone else: Don't start lecturing people about x86 hardware, most of us *ARE* PC users apart from Amiga users. I might only have a Pegasos here (running Linux at the moment, for Uni stuff) but I've got 3 PCs back home, all of which built by myself.
-
It's not a matter of which hardware is better. I agree PPC beats x86 in most things hands down. It's not a matter of which is better, but a matter of target audience. A Corvette is better than a Geo. If I was trying to bring back Edsels, I couldn't charge corvette prices.
It's just a matter of costs and if you make something cost less, more people are willing to buy. I'm sorry but you can buy a 1.6 ghz dvd-rw 512mb ram 80gb e machine for $399 at office depot right down the street. thats without a keyboard mouse and monitor, but how can you keep up with that? PPC is not THAT much better.
-
10,000? Somebody lied. If they HAD sold 10,000 there would still be 10,000 in circulation.
You just don't see those p5 boards circulating anywhere because users are simply gone!
Remember boards were sold 6 years ago. Many many boards are collecting dust now or died due to old age. 10000 is quite reasonable (I recall it was much more than 10000) when comparing to OS3.5/3.9 sales (3.9 sold 15000 copies in 1999 still).
-
@itix
I'm quite sure I once saw the official figure to be about 11k, so certainly not "much more".
3.9 sold 15000 copies ??? Someone better tells H&P ....
Just as much out of line as the numbers you see for the Amigas sold by C= from time to time (I think 17 million was the "record").
-
Porting AmigaOS to x86 architecture is a waste of time, not only would it be competing with Windows, it will also compete with BeOS, Linux etc.
I've said it before, but what Amiga needs is it's own brand identity to set it apart with a genuine reason to chose Amiga over Mac or x86. Twenty years ago the A1000 had clear hardware superiority over the competition, hence the selling point. Even this wasn't enough though, it required several killer applications before it would become a viable alternative.
As we've seen with Windows, aggressive marketing beats superior architecture every time, but Amiga is fighting for survival with comparatively poor hardware and a marketing budget that wouldn't pay for a round of drinks.
Simply put, we need something unique that no other platform has got. It might be some absolutely stunning game that wins rave reviews, it might be an add on / application that appeals to artists / musicians but just selling a box that runs legacy Amiga software at a premium price is suicide. Likewise, recompling Amiga to run on an x86 box will offer no advantage either.
That isn't to say that Amiga shouldn't adopt Athlon or similar as it's core CPU, as development investment is clearly in excess of that set aside for PPC. Programming issues aside, a top end Athlon based Amiga would be cheaper to build than it's PPC counterpart.
-
Porting AmigaOS to x86 architecture is a waste of time, not only would it be competing with Windows, it will also compete with BeOS, Linux etc.
You are competing with Windows no matter what CPU you use. That is a fact.
Running AmigaOS on the x86 has the advantage of a higher spec and lower costs. The only real disadvantge of the x86 is legacy app integration.
You are right about the need for a killer app... but sadly now killer apps take millions to develop... I can download a freeware program for Windows, programmed by a CS student that will out-perform any program I can get for the Amiga platform for all the tasks I need, I cannot justify nearly £1000 for an A1 system... I could justify a cheap x86 board though to run AmigaOS... Infact I did just that when I bought a Mini-ITX for £70 to run AROS :-)
-
The idea behind AmigaDE is that some day the AmigaOS will run on PCs and handhelds and fruit computers. But in the meantime we need a "meeting place" to regroup and readjust our strategy. That "meeting place" just happened to contain a CHRP motherboard. I think it mainly had to do with having access to PowerPC technology through their existing code base and not having any x86 software available for Amiga at that time.
The AMD k6-II was the recommended platform for development of AmigaDE when it first came out. The x86 development platform got the shaft when the Amiga community rejected AmigaDE.
Hardware DOES have something to do with the selling of AmigaOS but until there is a common software base for AROS, AmigaOS, and MorphOS there is not going to be much left to pick up the pieces of the shattered Amiga community. AmigaDE could have done it but the only chance I see of the Amiga community pulling out of this is if AInc. and Tao Group come out with an Amiga version of AmigaDE and it catches on. Otherwise the best we can come up with for multiplatform compatability is SDL and the Amos Basic spinoffs that are going to use it. :-(
-
@ALL
http://amiga.emugaming.com/tripos.html
You may find this article interesting..
:-o
-
SamuraiCrow wrote:
(SNIP)
The AMD k6-II was the recommended platform for development of AmigaDE when it first came out. The x86 development platform got the shaft when the Amiga community rejected AmigaDE.
Selecting AMD K6-II would be a bad choice i.e. Super Socket 7 hardware reference is in EoL at that time.
-
@Erol
I'm more interested in CAOS, AOS is powerfull, but CAOS sounds even better. Doubt there is anything left of it source wise..
-
As we've seen with Windows, aggressive marketing beats superior architecture every time, but Amiga is fighting for survival with comparatively poor hardware and a marketing budget that wouldn't pay for a round of drinks.
Depends on "superior architecture" e.g. one would defined Windows's legacy/software investment protection as the desirable characteristics.
-
Nice comment, I agree with most of it but there is a little problem in part of it. The endian issues are not that simple. The PPC *can* work in both modes,
Except for PowerPC 970, hence Virtual PC issue.
the x86 can't.
Refer to i486 (and greater) instruction set Bswap instruction.
To everyone else: Don't start lecturing people about x86 hardware, most of us *ARE* PC users apart from Amiga users
Sticking with i386 POV doesn’t help things...
-
dmac721 wrote:
It's not a matter of which hardware is better. I agree PPC beats x86 in most things hands down.
Careful with generalisations. IF you want to start yet another X86 vs PPC then go right ahead...
-
-D- wrote:
That about sums it up... there are technical merits for using a byte order agnostic CPU like the PPC... but they are outweighed by the cost benefits of the x86.
Guess it took me a while to get it, but once I realized that an _emulated_ '040 on a dirt-cheap XP mobile chip crunches numbers pretty much identical to a Pegasos G4, the advantage of something like AROS became pretty apparent.
RC5 only highlights the lack of vector rotate functionally in the X86’s SIMDs(hence evolution from SSE->SSE2->SSE3->SSE-whatever comes next). General desktop applications doesn’t highlight this margin i.e. refer to MacOS vs Win32 comparisons. AMD and Intel’s SIMDs feature set is driven mainly by the general desktop market.
-
Waccoon wrote:
Argo: not to mention on X86 you have to compete to some degree with Windows.
You're competing with Windows if you do Windows-like things. The hardware is irrelevant. Microsoft makes many standards used on x86 boards, but they certainly don't own the architecture.
Extensive Microsoft’s backing for MacOS X (MS Office, MS Virtual PC, MS IE5, XBOX-whatever SDK**, MS Media Player 9, MS educational titles) didn’t change the fortunes of PPCs in the desktop market. MS Windows 4.0 PPC and Windows CE 2.x PPC (for handheld Windows devices) for didn’t rescue PPC desktop market.
**Rumoured for MacOS X PPC, now superseded by MS's XNA initiatives.
Microsoft has been patient with PPC for a while now.
-
Erol wrote:
@ALL
http://amiga.emugaming.com/tripos.html
You may find this article interesting..
:-o
CBM's HW model wouldn’t be competitive enough against a clone army (Wintel market).
-
Hopefully, Sony's new "Grid" processor will be a real wake-up call to the industry.
Google VILW, Sony Cell, Trimedia VILW**(MIPS based media VILW processor). One may find that the certain key persons in designing Sony's CELL have a background with VILW processors.
ATI and NVIDIA VPUs are already an extreme variant of media VILW style processor. In terms HW, XNA is VPU centric (Primary HW partners are ATI and NVIDIA).
**AMD and Philips are also interested with this particular VILW design.
Game machines have always been far more advanced than PC hardware in all respects except I/O and CPUs
Not always i.e. note following and dates;
1. DOOM1's arrival during early 1990s
2. 3DFX's Voodoo during late 1990s.
3. NVIDIA's Geforce 256 in the late 1999.
-
hammer wrote:
Selecting AMD K6-II would be a bad choice i.e. Super Socket 7 hardware reference is in EoL at that time.
It was least common denominator at the time but it supported 3dNow! technology. Stuff written for the K6II would really fly on an Athlon64 nowadays wouldn't it? :-D
-
Now, getting rid of the GUI cruft that clogs a typical Linux system... that's what Amiga should be doing! I really wanted a QNX powered Amiga, especially since that company put a lot of effort and money into a new Amiga before Gateway screwed them over.
Yes please... I've been using qnx for a while now, and it would (at least as far as I can tell) make an excellent foundation for a home/business OS. Neutrino 6.x.x. and photon out of the box is very good stuff but really not suited for the average user, but add some Amiga-style usability and ease and we'd be all set.
Oh how good it is to dream... Why did the whole qnx-amiga thing fall apart anyway? Did QSS get screwed over?
-
porting AmigaOS to x86 is agreed as "a waste of time"
lets face it, the hardware sucks. and powerpc IS the future.
intel has stopped production of thier {bleep}ty cpu.
the mhz myth is slowly going out the door.
so why port a superiour OS like amiga to a {bleep}ty platform
thats already got a full market.
PowerPC rocks.
-
Would be interesting to know the number of Amiga Forever cd's sold to run on these crappy pcs. I would hazzard to say there are a whole lot more pc's owned by the members here than geniuine Amigas (still working that is).It is strange that so much can be done by this pc intel crap now a days.
-
mikrucio wrote:
porting AmigaOS to x86 is agreed as "a waste of time"
lets face it, the hardware sucks. and powerpc IS the future.
intel has stopped production of thier {bleep}ty cpu.
the mhz myth is slowly going out the door.
so why port a superiour OS like amiga to a {bleep}ty platform
thats already got a full market.
PowerPC rocks.
How about inputing some informed comments into this forum rather than using it as a place to dump your verbal diarrhea.
-
Hi Hammer,
RC5 only highlights the lack of vector rotate functionally in the X86’s SIMDs(hence evolution from SSE->SSE2->SSE3->SSE-whatever comes next). General desktop applications doesn’t highlight this margin i.e. refer to MacOS vs Win32 comparisons. AMD and Intel’s SIMDs feature set is driven mainly by the general desktop market.
I agree with your statement, I realize RC5 isn't an accurate "across the board" comparison between PPC and x86...and that "real world" tests prove such...
To clarify a little, what I'm saying is that an Athlon Barton, emulating another CPU, benchmarks nearly as fast (or faster, depending on the benchmark) as the MPC7447 used in the Pegasos G4, while both running comparable OS's.
(i.e, run "AmigaMark" CPU benchmark on the above stated machine in WinUAE, and compare the results to the MorphOS executable of the same program (or the 68k executable to compare JIT performance)).
I realize it's not an "100% accurate" or even ideal method to compare systems ..and there are many other factors involved...my intent was simply to highlight the price/performance disparity between the two setups, both running amiga/amiga-like OS's, and then to note the potential of AROS, which runs *native* on x86 and isn't emulating '040s like WinUAE.
My apologies if you already determined that from my post, and were only making a comment.
-
-double post-