@hatschi

Even if this is way off, I cannot resist:

Quote:

Luckily, he did not have to care about RoHS, although he was the LEAD engineer.

Damn, I am the teamLEAD...(what happens if we have to be LEAD free in July...but then, I am in the US)

_________________
AmigaOne XE G3/933/VIA/FM801/R200 (fixed),
G3SE/600/Voodoo3/Sil680/RTL8139/SBLive! (noiseless!)
µA1-MK2/G3/933/R200/CMI8738
XE/G4 (broken 7450/800)

@syrtran

So we are agreed that the whole "Amithlon embroiled in controversy" could not possibly have been a reason for the PPC decision. OK.


Quote:

Again, you are right. I don't know for sure if they were using the actual 68k. I only figured they did 'cause I thought it would've taken them even longer to write an emulator that would switch in/out only when necessary. Since I've got that right here on OS4, maybe they did write the emulator first.

Actually, it's probably easier to write the emulator than to write processor-switching code. Once you really start thinking about the PPC calling the 68k, which then calls the PPC, which then calls the 68k, at which point a task-switch occurs, and the new task is currently in PPC-land (as a result of being called from 68k) and executes a return instruction.... And then start to worry about stack setups, and where to hide away the other processor's stack space and state, you have enough of a problem even with an emulator. Trying to actually implement whatever solution you come up with on real, physical processors, which are not cache-coherent between them, and which have a somewhat hard time even catching each other's attention, would be a complete nightmare, both from a complexity point of view, as well as where performance is concerned.

Much easier (not easy, just easier) to either (a) do what PowerUP and WarpOS did, and treat things as two separate systems with a few clunky ways to interact, the correct usage of which is left to the user, or (b) run one CPU in an emulator on the other, thus ensuring cache coherency, avoiding (some) concurrency issues, having a single CPU that will handle interrupts, and various other nice things.

That, BTW, is the other reason I am not buying your argument of "they went PPC because the PPC already coexisted with the 68k on one card". Yes, there were such cards, but the way the processors coexisted made them pretty much useless for seamlessly, transparently mixing 68k and PPC modules. And given their history, I am sure the guys at Hyperion knew that.

@syrtran

Quote:

In any case, they still needed access to the custom chips while they were doing the port. Other than Amithlon, there was no way to do this for x86, unless someone made a special bridge board (PCI or ISA slot to Amiga CPU slot).

I am not quite clear on why they should need the custom chips. Getting AmigaOS to boot without custom chip emulation is not all THAT hard, even if you don't have the sources. If you *do* have the sources, and can recompile things, it should be downright straightforward. You will eventually need to replace all those parts, anyway, so you might as well start work there.

The first couple of weeks work on Amithlon were just that --- remove custom chip emulation, run AmigaOS, see where it breaks/hangs, fix. Of course, for me (lacking the source, alas) the "fix" part often consisted of "putting in custom-chip-emulation light", i.e. just making sure that, for example, code that waits for a vertical retrace on the ECS chipset eventually sees the right bits. With the OS source, you'd just make sure that check is never there in the first place...

I think Jobs was wrong as when he took the decision of having MacOS X on both Intel and PowerPC since the beggining,,,

Stop!... I'm not making any sense!

_________________
Indigo 3D Lounge, my second home.
The Illusion of Choice | Am*ga

@nzv58l --- what he said.

The BeBOX had a "geek" port. I lusted for that machine
for about six weeks. As Gasse proudly proclaimed, the
successor to the Amiga. Multiple cpu's. An OS that was a database
at its core. But then again had I lusted after the Apple Lisa when
I actually had the ten grand handy to buy that 1M 68K machine
with the dual 8" floppies. A couple fo dreams like sand through
my fingers.

I did buy an Amiga500. I later banged an A590 with a couple
of megs of RAM and a huge 40M drive into the connector slot
as well as another two megs of RAM into the expansion slot.

I couldn't justify an A3000 for the price it went for. But I
knew I could upgrade my A500 with an 030 in the future.
But then there was a firesale at Commodore and the deal
was just too good. I went fully 32bit.

I eventually ended up with an A3000T with both the CSPPC
and CVPPC plus a nice SCSI UW hard drive.

I now run AOS4-pre on my micro A1-C and about 98% of my
legacy apps acquired over the last twenty years that form
my software base.

Having a multitasking OS from 87 onwards has had a profound
effect on, how I see, things, computing.

nzv58I sums it up quite nicely. I sold my A3000T bit by bit to
people who wanted various parts and I gave other stuff away.
I have my replacement and am looking forward to the Troika
board and all the uses those USB2.0 ports can be put to.

Dual core or multi cpu support would be very nice for 4.1.
A higher end machine would go over well with me.
A good supply of apps would also make me happy.

Everything else is possible somewhere in the future. Right now
we are re-establishing a foothold, it's a precarious position that
requires support instead of all the naysaying, hindsight and
second guessing that has been the order of the day.

Last edited by kindergip on 14-Jan-2006 at 04:14 AM.
Last edited by kindergip on 14-Jan-2006 at 04:13 AM.

First off, I want to express my support for the Hyperion team for making Amiga OS4 possible. I've got my uA1 running pre-3 (soon pre-4), and I think Amiga OS4 out-shines my wifes XP system which is running at a little over 3 times the MHz.

But anyway. . .this thread is about hardware.

What has only been slightly touched upon so far (in this particular thread) is something that has the potential to make OS4 stand out as a worthy bearer of the "Amiga" name even from the hardware side of things, namely, the CELL technology. Of course, this could only work well. . .
*If CELL based desktop systems become easy to acquire (at not too high a price)
*If some important applications would be developed to make full use of the CELL technology (even if the OS itself does not use it to its full potential right away)

I know the above does not seem likely to become reality (in the short term anyway). But as long as we are just throwing out ideas of what we wish would happen with Amiga OS4 and beyond on the hardware front. . .well, that's my wish.

@PEB

Finally, someone is making some sense. I was starting to get worried :)

Yogi

@PEB

Now I like the sound of that, It would give the Amiga something that stands out of the crowd. It`s hard for someone to consider an AmigaOS as a main OS, because, everything done on an Amiga can be done on a PC with Win,Linux, and now OSX.
The main thing that would put Amiga back in a mass market is killer apps, games, or just plain retro nostalgia. The last being more like the C64DTV. The problem being how would a new vender lets say Troika for example, be able to use a processor, that is meant for game consoles? The Cell microprocessor will be mass produced for sure, but said not to be in the desktop market. There would have to be a license with Sony, Toshiba, and IBM.

_________________
Pegasos-II board G3/600 MHz MorphOS3.x/ 1GBRam,
Power Mac G5, Efika MorphOS3.x
SAM440ep 600Mhz AmigaOS 4.1 512
AMIGA 1200-DBOX,BLIZZARDPPC175 MHz,64MB,Mediator1200SX
VOODOO3, OS 3.9 & 4.0, Amiga CD32

@PEB

I totally agree... i don't know if i was clear enough in my previous post, but
you hit the nail right on his head.

If (this is a doomed word somehow) there'd be widely available, fast and cheap
PPC/POWER/CELL/whatever-of-this-family hardware, i'd be all for it.. but
OTOH i wouldn't be that much concerned about x86 technology..

Best thing for us would be IMHO to be a bit different (would be too boring if
everything and everybody would be the same, wouldn't it?) *and* to be able to
keep up performance wise with the rest *and* to stand out in one or the other
field.. Cell could help for that goal.. somewhere in the future... but right
now there is nothing unfortunately to keep up..

But thats only in the Desktop field... our chances are probably in totally
different fields than the Desktop (as already said, PDA, MDA, Smartphones,
STB, etc.pp.), which doesn't mean that the Desktop should be forgotton.. it
may only take less priority for some time until there's again enough momentum.
After all, AmigaOS is damn scalable, so wheres the problem of having the very
same OS on my PDA and on my Desktop

Hope you got my point.. i'm still not fully awake..

_________________
Regards, Cyborg.
AmigaOS4 development team member

"In the beginning was CAOS.."
-- Andy Finkel, 1988 (ViewPort article, Oct. 1993)

First off, from myself, a BIG, BIG, vote of thanks to Dave Haynie!

I've been originally a Commodore User from the first C64 that I acquired in 1982. THAT was also the year that I got into VCRs, and shortly after, a VHS separates-system vid-camera. I even used that C64 for titling and generating scrolling credits.
Stepping on up through C128D (and SX64) to KS1.2 A500, and then A2000HD (when the price dropped) . . . I used that A2000 (with 68060 & PicassoII/A2320) up until this A1-XEG4 took over in 2003(?)

I have no opposition to ANY processor that AmigaOS4.0 runs on . . . I just want to run it. 4.0 (presently on AmigaOne hardware) gives more of that pleasurable operating system experience that no non-AmigaOS (apart from possibly BeOS) could provide.

But really, in THIS day and era, it should be possible to run applications in an operating system-agnostic (not sure if THAT's the word) way. Wasn't the concepts of JAVA and HALS steps along that path?

Presently I use no x86 machines but its more the issue of OS that decides that. We don't use MSWindows in this household because no other OS has such latency problems, so many openings for viral intrusions, so many non-coopperation issues, . . . much like I wouldn't drive a Lada Samara, or run a powerboat that needed excessive bilge-pumping!

In fact, the THIRD PowerPC machine in this house resides just to my left -- a 2000 Dual 500Mhz Gigabit Ethernet PowerMac running Tiger . . . . . . I wouldn't mind dual-booting THAT to AmigaOS4.0!! Next up it'll likely be a Laser Printer for the dual 500Mhz!

Last edited by vortexau on 14-Jan-2006 at 10:21 AM.

_________________
-vortexau, who's A1 XE-G4 remains at half-RAM !
A2000HD (from 1991) 060 64Mb PicassoII with OS3.5 . . . still working.

@Cyborg

"different for the sake of being different/interesting" has already been addressed several times in the thread.

If anyone's interested, I have a 5 1/4" Compaq Bigfoot harddrive I'd be willing to sell, it's different, definately not the boring old harddrive.

I'll even throw in an IBM M-Wave Modem/Soundcard combo.

Proprietary is a dirty word

_________________
"If you want to tell people the truth, make them laugh, otherwise they'll kill you." - Oscar Wilde

@hazydave

Quote:

My whole "obsession", as you might call it, with performance here is twofold. One is my simple claim that anything called "Amiga" should be to today what Amiga was back in the day. In particularly, it should be capable of serious multimedia performance. I'd probably buy into the AmigaOS on the PS3 idea, simply because that would actually achieve that -- it could probably render my HD MPEG-4 videos faster than any PC, anywhere. It's not general purpose computing, but it's useful and "Amigaesque".

I know you would see Amiga like 90 years bu now is not possible form many reasons. At the moment we can make only step by step operations. AmigaOS was old and now is a modern os. It need any modern features too but there is a cut with the past. Next step could be a port on x86 platform. Could be. I know many and many Linux users that lament lack of informeations about drivers. Yes is same thing on PPC but with a specific platform (like xbox or maybe an x86 platform) could be really simply. We should change first how Microsoft drive the market. in the 90 years the market was free. Not today.
For the multimedia performace i can only say to you that mplayer on windows stop movements when i move its windows. not on amigaos. these are no multimedia brutal performance but can understand how an os is important and you should know very well since with a 7mhz (my fantastic A2000) we can be a lot of things also today (not your HD MPEG-4 of course.. but i must try.. )

Quote:

Second thing is this: there are not enough Amiga fans around to drive serious hardware development. I think that's been demonstrated. There may be enough to drive sufficient software development, but we'll have to see even about that (and I'm just talkin' OS, not applications). To make the market viable, you need non-Amiga people to buy into the vision.

There are many hardware manufacturer that are not capable to develop Linux drivers simply because Microsoft force him to don't do this. Why should they enter in Amiga market? So we need some "stupid" (ehi i'm joking..) fans to develop drivers.

Quote:

So how do you convince the end-user who's not an Amiga fan to buy? You need to make it very cool, or very cheap, or something compelling. You can pretty much assume they already have used PCs and maybe even Macs; you're not likely to deal with people who haven't touched computers. This is basic grade school education these days. My claim is that, when you're selling something that's a small fraction of the compute power of the default choice (the PC IS the default choice) to an end user, they are not compelled. This has been Apple's problem, too. And that's all before you even bring up the software question...

I'm agree with you. I have bougth an A1 at 1300 Euros and is too much for a modern pc. but it is our market today. And how about a Mac at 1800 Euro? today i buy 3-4 PC at this price. but there are many users that bought it. I like to see OS4 on a x86 platform but today we cannot do this because we haven't a company like Apple that with IPOD market can do all they want. We have only a bunch of companies that try to keep our platform alive. nothing else.

Quote:

In reality, you don't need a 3GHz Athlon64 to do web browsing, email, or word processing. But it's the easiest growth path for regular computing, staying orthogonal with respect to all code, just one kind of CPU, even when multiples are present. That's not the most efficient way, but it's well proven.

Yes. but we have used our amigas or many years. Today the GHZ growth is a fashion nothing else. What do you have in your DVD to play an high definition HD-MP4? a dual athlon x86? guess not.

However i'm very pleased to speak with you. hope you can become a regular user of this forum first and maybe a POWER user in the future.. how about a new motherboard??

_________________
http://www.amigasoft.net

@T_Bone

Quote:

"different for the sake of being different/interesting" has already been addressed several times in the thread. If anyone's interested, I have a 5 1/4" Compaq Bigfoot harddrive I'd be willing to sell, it's different, definately not the boring old harddrive. I'll even throw in an IBM M-Wave Modem/Soundcard combo. Proprietary is a dirty word

You got me wrong.. i didn't mean "different fo the sake of being different",
i meant it the way like "sticking out of the masses in a good way" (tm).
Just like Apple still sticks out of the masses although hardware wise isn't
that much of a difference left with x86 architecture.

Be it by design or maybe because of incredible performance in some more
specialized fields, which *may* be accomplished due to a future Cell architecture
or new custom chips to help the CPU or whatever..

_________________
Regards, Cyborg.
AmigaOS4 development team member

"In the beginning was CAOS.."
-- Andy Finkel, 1988 (ViewPort article, Oct. 1993)

@kindergip

I agree with you

@ all the naysayers

Sorry, but I don't get it. Os4 is almost there and we are (mostly) all complaining about porting it to x86. I don't know if we can wait another year?!Can we?
I can't immagine how many times the "when it's done" smilie will clutter up AW forums again..

As soone as the x86 port will be almost done, we will have 24 threads asking and complaining to Hyperion why they didn't code it to support dual core and dual processors and PCI-express (looks like suche an HW will be standard in a year by now). And what should hyperion live from in the next year? Our love and air? The solution could be another T-Shirt thingy for 500$ each shirt, so these poore guys could at least buy some bread for the next year of porting os4x86. The motto on the T-Shirt could be:

We are
The Amiga Comunity
the best comunity of The World
and We Know It Better than anyone.
Our motto:
when it's done!
How to get it?
Complain to Everything
so on the last minute they Have to
change everything
to
make Us happy,
to let Us hear
When it's done!


and if we menage to trick the friedens brothers telling them at the mensa: " you have already had your lunch" - as soon as one twin had it- we'll have half of the budget free to choose our AmigaTWO cheap-x86 motherboard and writing a proprietary bios for it.

by the way who should write all the drivers as posted somewhere above?
there are..how many.. over 1000-2000 A1 out there? Is all the community really going to buy the new amigatwo next month, each with a little different hw to write drivers for it? And who will write apps in the meantime?

I can remember some years ago, at Petro Tyschtschenko's age, as a lot of ppl in the community thought of the idea of moving AmigaOS to x86 as the reason to drop the amiga..

Can't we just wait untill they free OS4? Untill we know what the mystery HW is?

@Amigo1

(UNRELATED)

Quote:

As soone as the x86 port will be almost done, we will have 24 threads asking and complaining to Hyperion why they didn't code it to support dual core and dual processors and PCI-express (looks like suche an HW will be standard in a year by now)

AS I understand it Olegil has PCI-express gfx working on a pci-reiser card with PCIe slots, I think he said that there was nothing magical needed to make it work.

Olegil might tell you more, or search the forum.

_________________
Yes I own an Amiga. A non-upgraded A500 that is unpacked once every 3 years.

If you are going to quote me, do so fully or not at all.
/Ins

@Cyborg

Quote:

Cyborg wrote: You got me wrong.. i didn't mean "different fo the sake of being different", i meant it the way like "sticking out of the masses in a good way" (tm).

That would be hard to do. Finding a hardware platform that is competitive (and sticks out in a good way) against x86 is a monumental task, especially when if you decide to not compete at all, and just use x86, rather than try to compete with it, it doesn't cost you a dime in hardware developement at all, for the same performance. Using x86, you're comparable to the competition, for free.

Plus, that requires a hardware department. The guys currently working on new hardware for AmigaOS4 arn't even making hardware that compares favoribly to the old Eyetech boards and CPU speeds, much less even bargain bin x86 hardware. We're going backwards in performance. We can "stick out" only when compared to classic Amiga's, that's about it.

Think about it, 16-bit Windows 3.1 runs on modern hardware we're nowhere near competitive with, in price or performance, and the hardware situation isn't getting any better, it's getting slower. x86 hardware gets faster every year without having to find an eyetech or a troika every year to make a new board.

About the only market for these things are the people reading this thread. (note that I'm included, but I'd like to see market expansion, at least a little )

Quote:

Just like Apple still sticks out of the masses although hardware wise isn't that much of a difference left with x86 architecture.

Apple decided to stop competing on hardware, they decided to become comparable instead. They did this because they simply couldn't stick out anymore, they were having trouble just keeping up. Now they don't have to try to keep up, it's free.

Quote:

Be it by design or maybe because of incredible performance in some more specialized fields, which *may* be accomplished due to a future Cell architecture or new custom chips to help the CPU or whatever..

From what I hear from people talking about the CELL, it's a completely different beast and porting to it would be more difficult than porting to x86, and that's just considering software. You'd also need someone to build them for you. x86 boards you can get at the corner, and are better suited for desktops anyway.

Besides, what good would CELL do us? Multimedia? There's no software for it. Sure it can be written, but hells bells man, we don't even have MOZILLA yet. What companies are going to port all the professional multimedia software from Windows to AmigaOS if we can't even expand our market base past how many people fit in my sons school cafeteria??

_________________
"If you want to tell people the truth, make them laugh, otherwise they'll kill you." - Oscar Wilde

@Cyborg

Quote:

i wouldn't be that much concerned about x86 technology..

Cell doesn’t increase PPC code parallelism nor it accelerates PPC based code.

By the time Cell arrives in the market place it will be in competition with “Fat” multi-X64 cores.

_________________
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)

@Amigo1

Quote:

Amigo1 wrote: As soone as the x86 port will be almost done, we will have 24 threads asking and complaining to Hyperion why they didn't code it to support dual core and dual processors and PCI-express

Right, because having a successful product that could be sold to thousands is less important than preventing 24 threads on a web forum somewhere.

I get it now.

Amiga Inc may be hiring a CEO, I nominate you

Last edited by T_Bone on 14-Jan-2006 at 10:43 AM.

_________________
"If you want to tell people the truth, make them laugh, otherwise they'll kill you." - Oscar Wilde

@afxgroup

Quote:

What do you have in your DVD to play an high definition HD-MP4? a dual athlon x86? guess not

Like Cell's SPU, refer to NVIDIA's PureVideo or ATI's AVIVO.

_________________
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)

@The_Editor

Quote:

Judging DnetC crunching stats I've often wondered how an Altivec optimised app on Os4 (A1) would compare to the rendering times of the Pc (its running AMD64-XP3500).

SSE2/SSE3 is missing certain Altivec vector instructions which are not overly important to overall 3D rendering performance e.g. G4’s Altivec didn’t rescue it from Cinebench 2003.

SSE3 specifically targets software vertex shaders.

_________________
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)