@Stephen_Robinson

Quote:

Stephen_Robinson wrote: @everyone I'll say one good thing about this thread (and the 13 others), it's stopped us all whining about Pre-Release 4 not being out!

(13? I thought 2/3, but heck, that's not many threads considering a Dave Haynie appearance, we see more threads started by people saying they're leaving than we do about Dave Haynie showing up, Heck i think that's backwards )

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

@redfox

Quote:

I often regret not purchasing a newer Classic Amiga, such as the A3000T or A4000T

Well, I bought A3000/030@25Mhz (in 1993) and my local PC vendor (for Fast RAM upgrade) at that time commented that the A3000's chassis/case was over engineered.

_________________
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)

@Leo

Quote:

Leo wrote: Quote: Only x86 and alphas are little endian Every PPC, except the G5 also does little endian. Leo.

Please read the entire text! i said that only x86 and alpha are little endian processors and the others (that include PPC) are big endian or have the two modes.but i had a mistake it´s not dec alpha that is little endian only, is the DEC VAX.Sorry

Last edited by Fransexy on 16-Jan-2006 at 01:19 PM.

_________________
No PowerPC, No Fun
Make Amiga Great Again

@Seehund

Quote:

Hazydave is Dave Haynie, just to make that clear.

Thankyou.

Quote:

What battle? I think it's been an interesting and pretty civilised discussion so far,

You're right. I quess I'm just kind of tired of these endless discussions.

I keep forgetting that many people are just looking for viable alternatives, and only a very few are actually just stirring the pot.

---
redfox

@Seehund

Quote:

So what's the great difference compared to "AmigaOnes"?

The main difference is that they will be much more affordable.

Quote:

Even worse specs than the poor "AmigaOne"

The PowerVixxen will be lower performance than the A1, the Amy'05 not, according what I know about its current specifications, which is not even final.

Quote:

Why do we so stubbornly refuse to learn from the death of the Amiga and the failure of the "AmigaOnes"?

The AmigaOne, despite its high price, was selling like hot cakes, Eyetech simply could not deliver the quantity that was needed and now they are unable to deliver at all for unknown reasons. Every batch they produced sold out really quickly. So what we learned from that is there is huge demand for PPC hardware running AmigaOS4, even with high pricing. So if Troika can manage a more reasonable price tag, it will be a blowing success.

Quote:

Trying to be "special" is bad when "special" isn't synonymous with being the best.

Would OS4 on a x86 box be the "best"? I doubt it. The PPC architecture has amazing multimedia potential, the x86 has nothing that can compete with Altivec today, and the possibilities the CELL offers in multimedia is simply amazing. There's nothing even close to it on Intel's roadmap. Giving up all that advantage because one particular company failed to deliver hardware would be foolish. According to Hyperion there are several companies working on hardware some of which have not been announced yet. Obviously they are no fools and if they saw that the PPC is a dead end, they would move to a different CPU architecture ASAP. I'm sure they won't let all their work go to waste.

@COBRA

Quote:

COBRA wrote: Would OS4 on a x86 box be the "best"? I doubt it. The PPC architecture has amazing multimedia potential, the x86 has nothing that can compete with Altivec today, and the possibilities the CELL offers in multimedia is simply amazing. There's nothing even close to it on Intel's roadmap.

Are you aware of Motorola's and IBM's roadmap WRT PPC? I could be wrong, but I've heard they wern't planning anything more on the high end, and embedded class CPU's were their focus, not Altivec enabled multimedia monsters that couldn't even compete with today's $150 x86 CPU's.

Quote:

Giving up all that advantage because one particular company failed to deliver hardware would be foolish.

It would help if we knew why they couldn't deliver the hardware. I wish Eytech would speak publicly about this, rather than just leave the situation at "Eyetech couldn't do it..." in everyone's eyes.

Quote:

According to Hyperion there are several companies working on hardware some of which have not been announced yet. Obviously they are no fools and if they saw that the PPC is a dead end, they would move to a different CPU architecture ASAP. I'm sure they won't let all their work go to waste.

I'm sure they won't either, that's why it's nice to see a thread every once in awhile that shows them that if they HAD to, there would still be people here who would still be interested in AmigaOS4 if they have to go x86.

It's not like MOS, where they are stuck with PPC, because that's genesi's whole business plan, to promote PPC weather MorphOS is successfull or not. They'll let MOS die and just keep pushing PPC in other OS areas if it comes to that. We have options, they don't. < -- (not a MOS flame, I'm a fan of the BABY, not the bathwater)

Last edited by T_Bone on 16-Jan-2006 at 04:17 PM.
Last edited by T_Bone on 16-Jan-2006 at 04:17 PM.

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

@T_Bone

Freescale are still on the road to relase teh 8641(D) - that is an embedded proc, true, but that is not a disadvantge necessarily. The e600 core is quite okay. It has been slowed down *a lot* in current 7447/7448 because of the slow FSB. The 8641 will not suffer from that. It will susatain for a while and be quite competitive in the high integration, low energy, high data throughput market.
Btw.: the e600 core includes Altivec.

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

One thing I look forward to with the PPC is system on a chip. Powevixxen or 4U2 with AmigaOS4/MOS would be cool. Imagine picking up a "matchstickbox" from your pocket, plug it to you 42" flatscreen and have a fully working, userfriendly desktop with some mediacenter app onscreen! Or even have the device inside the screen.
It can be made alot cheaper than the alternatives out there.
I guess thats what hyperion is trying to do. And that´s where the Amiga way of doing things can really make a difference.
Even the apps are low footprint! Everything just fits!

I do hope for a port to a powerful/cheap desktop computer in addition though.

@T_Bone

What with all the people screaming bloody murder here and elsewhere over the last "IDE DMA not working while using the network chip" I say Eyetech are wise to stay below the radar. There's little money in sitting around here being yelled at all day.

I note how Seehund seems to have a weird definition of failure, though. If something sells out, how can it not have been a success? I'm certain they would sell more batches still (would be nice to find some working riser cards and figure out the old PCI graphics card issue, though). We need a board with higher spec, but that's mainly just faster CPU and better graphics card. The rest of the board isn't too bad, really.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

Quote:

The rest of the board isn't too bad, really.

Agree, I'd like to add one more thing on the "improvement list" though:

An ability to handle any kind of PC-133 memory-dimms, in pairs.

_________________
Site admins are people too..pooff!

@olegil

Quote:

olegil wrote: @T_Bone What with all the people screaming bloody murder here and elsewhere over the last "IDE DMA not working while using the network chip" I say Eyetech are wise to stay below the radar. There's little money in sitting around here being yelled at all day.

True, but they could have made a statement. They did this to Amiga Inc and outed them (in 2001/2002 or so?), they could have at least been that proactive in their own affairs. At least we'd have some idea of how things might be for anyone who follows them.

Quote:

I note how Seehund seems to have a weird definition of failure, though. If something sells out, how can it not have been a success? I'm certain they would sell more batches still (would be nice to find some working riser cards and figure out the old PCI graphics card issue, though). We need a board with higher spec, but that's mainly just faster CPU and better graphics card. The rest of the board isn't too bad, really.

If you're going to sell water in the desert, supply is probably the biggest factor in determining success. Aftertaste could probably be tolerated.

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

@T_Bone

Quote:

Are you aware of Motorola's and IBM's roadmap WRT PPC? I could be wrong, but I've heard they wern't planning anything more on the high end, and embedded class CPU's were their focus, not Altivec enabled multimedia monsters that couldn't even compete with today's $150 x86 CPU's.

I am not aware of such. IBM are in fact broadening their targeted markets, they are moving into home entertainment by providing high performance CPUs for the next generation game consoles. For example the XBox 360 has a very powerful 3.2GHz PPC processor with 3 cores, with one VMX-128 vector unit per core (the IBM equivalent of Altivec). The Nintendo Revolution will also be using PPC chips from IBM, probably lower performance than the XBox. And I don't think I have to mention the revolutionary CELL architecture. It doesn't sound like they are going low-end only.

As for Freescale, their roadmap contains the e600 and e700 cores, which are the successors of the G4 family with clocks speeds of 2-3GHz. See here.

There is now a third company developing CPUs based on the PowerPC architecture called P.A. Semi. They have announced some high performance CPUs with incredibly low power consumption, nicknamed PWRficient. See here.

So the PPC architecture has a bright future. The only question is, will those "secretly working" companies manage to bring those goodies to AmigaOS4?

@olegil

Quote:

I note how Seehund seems to have a weird definition of failure, though. If something sells out, how can it not have been a success? I'm certain they would sell more batches still (would be nice to find some working riser cards and figure out the old PCI graphics card issue, though). We need a board with higher spec, but that's mainly just faster CPU and better graphics card. The rest of the board isn't too bad, really.

I think that Alan said he lost about 500 euros on each board sold...

Last edited by pixie on 16-Jan-2006 at 05:07 PM.

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

@tomazkid

*cough* parallel port *cough*

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@pixie

If he lost money on each board sold and is selling them out, then perhaps he did not order them in enough quantity. To be on the safe side he ordered them in very small batches. If he did his homework he should not have lost money on the boards. It also matters on what that additional cost to him was about. Is developement of the board considered in this number? If it was then perhaps he should have not come out with the Micro A1 and instead just fixed what was known not to work on the original boards.

Is there a possibility that the original spec of the A1 was too much? Perhaps a more flexible standard should be made. Lets get rid of all the stuff that is not needed for an A1 specification.

Quote:

So the PPC architecture has a bright future.

PPC has *no* future on the desktop market. Plain, simple. Period.

Leo.

_________________
http://www.warpdesign.fr/

@Leo
but does it matter? There is no way that the Amiga is *EVER* going to conquer the desktop market. Luckily, the desktop as we know it is changing.

PowerPC is going into all sorts of devices, and thats why its not a bad place for AmigaOS to be.

_________________
Playstation Network ID: xeron6

@T_Bone
Quote:

AROS doesn't have it by choice. It was not one of AROS's design goals, it was designed to be source compatible, not to emulate binaries.

Not entirely true, or at least not in hindsight. Current bounty is for seamless UAE integration:
http://www.thenostromo.com/teamaros/bounty_details_7.html

Not meant to start a debate, just to offer the facts.

Personally, IF OS4 were to ever go x86, I would _hope_ that it was via emulation. That seems to be ignored in these dicussions.

Why? Binary compatibility. AROS doesn't have it, (although see above bounty), and there simply aren't that many new apps, at least not enough to be able to ignore all the old.

The other way would be a fat binary similar to Apple's approach, but that's meant to be transitional, not 'forever.' Some older apps that are still useful we may never see source to, so I'd prefer to take a bit of a hit (again, on an AMD64 as as example, how much of one is yet to be determined, but I'd be perfectly content running a 3GHZ AMD64 and have it running for all intents and purposes as a 1-1.5GHz PPC). I'm unsure that performance is attainable as I seem to recall PearPC discussion and it seemed for the most part to be running at 1/10th or so of actual speed, so this may be entirely not easy, but would be the ideal. Then again, with 68k code, the numbers look much better, but that's only helpful for immediate (sometimes dated) software, not more moving forward.

I don't know the original design goals of AROS, or why they came to the conclusions they did. It's highly likely the debated about 68k emulation from the start, and may have decided UAE was enough, or it was more work than people on the project, which are both valid...and no clue if PPC emulation was ever an option there.

Native code is always going to be faster, but breaking compatibility (at the binary level) is a tough call IMO, moreso in a specialty/small market.

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

@Fransexy

Quote:

A curiosity: Only x86 and alphas are little endian the others CPUs in the industry are big endian (or have the two modes)

Yep. This used to cause some annoyances when porting traditional Unix software to Linux/x86. Data file conversions among other things...but it is a well known/defined problem as well.

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

@T_Bone

Quote:

I just had another thought. Since Petunia isn't portable, if AmigaOS4 were ported to x86, why not just leave 68k emulation out? I don't plan on emulating 20 year old software forever, and OS4 native software is becomming more and more common.

Yes to the second (not emu forever), but the key word is forever IMO. It's a call between immediate apps, dated as they are, but they DO exist (68k)- IBrowse currently being rather 'important' if not modern standards compliant, versus a clean break. AROS went for source compatibility and not binary, and has few apps as far as I'm aware. That may be a good long-term strategy (drop binary compat/clean break), but for a small hobby OS market it's not good for allowing immediate/short term growth.

This isn't a shot at AROS by any means, it's just a comparison of objectives and impact, which I'm sure they discussed when it was started and perhaps throughout it's lifetime so far.

As a total 'out of dark/no research comment,' as PCI -Express becomes the standard, it might be a possiiblity to add a PPC card on the PCI Express bus, as it does have significant bandwidth (what, > 1GB/sec each way or so iirc?), but of course there's that 'odd/custom hardware' rearing it's head again. That could work similarly to the Cell concept (not the single clip implementation of the ps3 but the overall design goal where multiple Cell devices could form their own nodes, or new 'cells,' in a Cell distributed network), but obviously it's not a sane way if the goal is 'OS4 to the masses' by any means, just fun to think about.

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