@pavlor

Quote:

Criticism yes, but I am against Hyperion bashing.

Is it bashing to question Hyperion when they post in a thread discussing an x86 port, and the post contains no real arguments regarding the topic of the thread, but only a promise of an "ambitious" project they have started, but haven't told us anything about?

If they had a strong trackrecord of delivering solid products in a timely maner, I would give them the benefit of the doubt, but again, why should be believe they can deliver anything ambitious within a reasonable timeframe? Why is it bashing to ask them that? It's a valid question IMO.

@VidarL

Quote:

I think it's fairly clear that Amiga Inc went PPC partly because of Hyperion wanted to. IMO that decision is THE major mistake that has been done in the aftermath of Commodore.

Amiga Technologies announced PowerPC transition in 1995, long before the formation of Hyperion. At that time, it was accepted by whole community. These plans never materialised, only Phase5 made powerPC boards, but programmers - including those from Hyperion - learned how to use these CPUs. in 2001 there were many skillful PowerPC developers, but only few that understand x86 - including OS4 core developers. Thus, how could they port AmigaOS to x86?

Quote:

Where would we be today if they had chosen to work along the Amithlon route?

OS4 on AmigaOne is faster than any Amithlon/UAE based "x86" computer avaiable. This emulation would kill our OS.

Last edited by pavlor on 22-Mar-2009 at 04:51 PM.

Ok i haven't read all this tread so maybe it been brought up already, but why cant the OS be ported to the same thing AmigaAnywhere is running off, intent/ Vp code as I understand it that would make it Processor/Platform Independence that way it would support a wide range of processors, plus any software developed for it wouldn't have to be redeveloped for a different processors. is there a reason why it cound'nt be done.

Last edited by amigang on 22-Mar-2009 at 04:55 PM.

_________________
AmigaNG, YouTube, LeaveReality Studio

@amigang

It would be really "cool", but how fast could be this VP?

Imagine a native X86 AOS version...

What should it be compatible with ?

A sandbox for 68k, a sandbox for OS4... Programmers still writing apps for both and for AROS and MorphOS AND for the new X86 OS.... That would be a mess.... (it already is)

I for one think that if all would drop 'older' OS's and focus on the most current OS, this would benefit the most.

Also there should be only one Amiga (compatible) OS. Otherwise it is killing us. No 'competition is healty' it is just dividing the left Amiga users.

Remember most Amiga users would not buy a CD drive or more memory or faster processors, and we where left with slow diskbased games ?

Well, now it is time to make up our minds and pick ONE route and drop older OS's. Sure we still need compatibility for older important apps, but STOP developing them !!

You know... You'd think I was going somewhere with all of this.... Ah well....

Martin

_________________
Theres a time to live and a time to die
When its time to meet the maker
Theres a time to live but isnt it strange
That as soon as you're born you're dying

@Metalheart

Quote:

Well, now it is time to make up our minds and pick ONE route and drop older OS's. Sure we still need compatibility for older important apps, but STOP developing them !!

Why? If somebody wants to write applications for OS1.x, why not? ANAIIS is good example of such effort.

@VidarL

Quote:

Is it bashing to question Hyperion when they post in a thread discussing an x86 port, and the post contains no real arguments regarding the topic of the thread, but only a promise of an "ambitious" project they have started, but haven't told us anything about?

The HyperionMP post did indicate that the X86 road is not the one that they are taking.
It also indicated that another solution may be in the works and did not rule out Amiga OS ports to another cpus.

Or could it indicate that they are doing something else and that we should not bother them with Amiga OS requests.

Last edited by Spectre660 on 22-Mar-2009 at 05:16 PM.

_________________
Sam460ex : Radeon Rx550 Single slot Video Card : SIL3112 SATA card

@pavlor

Quote:

OS4 on AmigaOne is faster than any Amithlon/UAE based "x86" computer avaiable. This emulation would kill our OS.

Now thats pure nonsense !!

An AmigaOne running PPC-code might beat a ca.2002 x86 running Amithlon and 68k-code, but it gets easily defeated when both have to run the same 68k-code. Things get even more clear if you use a more recent x86 and replace parts of the system with x86-code (the native GFX-drivers alone can have a huge impact).

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Quote:

Now thats pure nonsense !! An AmigaOne running PPC-code might beat a ca.2002 x86 running Amithlon and 68k-code, but it gets easily defeated when both have to run the same 68k-code. Things get even more clear if you use a more recent x86 and replace parts of the system with x86-code (the native GFX-drivers alone can have a huge impact).

I meant PPC code on the AmigaOne side and 68k code on the Amithlon/UAE side.
800 MHz AmigaOne is 3 times faster than WinUAE box of my brother (Core 2 Quad Q6600 2400 MHz) - if you compare speed of Doom, Quake (without 3D accelleration), emulators, OGR etc.

@pavlor

Quote:

OS4 on AmigaOne is faster than any Amithlon/UAE based "x86" computer avaiable. This emulation would kill our OS.

You should actually try out an A1 vs Amithlon on a fast x86 before making ridiculous statements like that. Amithlon was competitive with the A1 even back in the days of the Athlon 64 for purely CPU benchmarks. On a modern x86 like the i7 or Phenom II the performance gap would be enormous.

_________________
Who do you serve, and who do you trust? - Galen

@pavlor

Quote:

I meant PPC code on the AmigaOne side and 68k code on the Amithlon/UAE side. 800 MHz AmigaOne is 3 times faster than WinUAE box of my brother (Core 2 Quad Q6600 2400 MHz) - if you compare speed of Doom, Quake (without 3D accelleration), emulators, OGR etc

Doom and Quake? Seriously? I don't deny their importance to PC gaming history, but no one is playing Quake under WinUAE, and no one thinks that's a fair comparison. /me goes back to playing Fallout 3.

EDIT: And Amithlon can/could execute native x86 code, right? WinUAE can, too, in a way, but in its current state, the emulation pauses while the DLL-based x86 code is running.

Last edited by Trev on 22-Mar-2009 at 07:41 PM.

_________________
Sam440ep-flex 733 MHz/1 GB RAM/Radeon 9250/AmigaOS4.1 Update 2
borked A1200/Blizzard1260+SCSI-IV/Z4+MediatorZIV/Deneb/Voodoo3/CatweaselMk3
more borked A1200/MBX1200z/Indivision
A500/clockport/RRNet
A600/A603

@DrBombcrater

I admit that Amithlon is much faster than UAE. With AMD Athlon XP 2600+, it reaches (in some cases) even the speed of Pegasos II. But there are many oher areas, where Amithlon stays behind: eg. DVD playback (there aren´t good 68k applications and G4 has AltiVec) or OGR (Amithlon on AMD Athlon 1.3 GHz cca 3.470.000 keys/sec vs. cca 14.000.000 keys/sec on uA1) etc.

@VidarL

Quote:

Is it bashing to question Hyperion when they post in a thread discussing an x86 port, and the post contains no real arguments regarding the topic of the thread, but only a promise of an "ambitious" project they have started, but haven't told us anything about?

It's a futile argument. The technical justification for not embarking on this path is a shell game. x86 was avoided because it was considered to be high-risk in terms of attrition of users and piracy.

Any OS of relevance is now platform-agnostic. Somehow the rest found their way.

The religious argument is even more laughable. People were so desperate to avoid any connection with the PC that they were more than happy to chop off their nose to spite their face. "Users will discover Windows and the MacOS" they claimed. Duh... welcome to 1992. "The PPC has a better ALU!" Blech... that's only an issue when you have software to take advantage of it.

I have a great deal of respect for Hyperion and all of the others who have contributed to the evolution of the OS over the years. But it's not clear to me what the business plan has ever been. If there was one.

I genuinely hope that whatever they're working on is both ambitious and rewarding.

@Trev

Quote:

Doom and Quake? Seriously? I don't deny their importance to PC gaming history, but no one is playing Quake under WinUAE, and no one thinks that's a fair comparison.

I´m playing Quake under WinUAE... (but I am crazy )
Doom and Quake are wery good examples of application benchmarks - widely used and avaiable on many architectures and operating systems.

Last edited by pavlor on 22-Mar-2009 at 07:50 PM.
Last edited by pavlor on 22-Mar-2009 at 07:47 PM.

@pavlor

Quote:

. With AMD Athlon XP 2600+, it reaches (in some cases) even the speed

Erm its 2009 these CPUs are dead meat since ages, if you compare speeds do it with recent x86 Cpus and not 1999 models.
Compared to my old A1 my UAE system is much faster and i can run all the nice old classic games.

Last edited by Totally_Blind_Mule on 23-Mar-2009 at 11:14 AM.

@Trev

Quote:

Trev wrote: @Hans Quote: Creating a decent port to PS3 would require access to graphics chip documentation that is not available, or significant resources to get the SPE's to fake HW accelerated graphics I'm confused. I know you and others are working diligently on improving OS4's graphics subsystem, but someone really needs to explain to all the lay people why the ECS and AGA chipsets are good enough for OS4 while the PS3's virtual frame buffer is not. Honestly, no one's buying that. The Cell/BE PPE should be fast enough--certainly faster than the 603e and 604e used in the Blizzard and Cyberstorm accelerators. From the benchmarks I've seen, the 440EP is only 2.5x to 3x faster than the 603e, depending on the clock speed. I'm sure Cell/BE PPE performance has been adequately covered in other threads.

Whether it would work or not isn't really an issue. I'm sure that it would be quite good; definitely better than AGA. However, the development time and effort required is rather big for what you get in return. The costs of development vs. developing something else do not stack up in its favour, at least in my opinion.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

@Kronos

Quote:

Kronos wrote: @Hans Amd who is paying to keep the lights up TODAY ? Obviously not the income from OS4-sales (o.k. technicly it should be just enough to pay the electric-bill, but I doubt that was what you were aiming up). OS4 hasn*t paid it*s bills and it as it looks it will never do, call it cross-financed or just a hobby ....

Remember how development slowed down at one point because they needed to do other work that brought in money?

I honestly don't know where their money has been coming from. However, I don't think that doing something that would delay getting a return on all that OS4 work for another few years would make much business sense.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.

@Totally_Blind_Mule

Quote:

if you compare speeds do it with recent x86 Cpus and not 1999 models.

We (at least me) discussed here speed comparison of Amithlon and A1/Pegasos in the time of their arrival (cca 2001-2004). My point was that native PowerPC applications on A1/Peg were (then) faster than 68k applications on Amithlon.

AMD Athlon XP 2600+ is from year 2002.

Quote:

I have a great deal of respect for Hyperion and all of the others who have contributed to the evolution of the OS over the years. But it's not clear to me what the business plan has ever been. If there was one.

Well, if there is one, I don't know who decided/agreed on it... but choosing custom PPC hardware as the target market for a desktop OS isn't that clever if you ask me :)

_________________
http://www.warpdesign.fr/

@ferrels

Quote:

ferrels wrote: @Hans I'm not "expecting" Hyperion to do anything. I'm pointing out that porting an OS to a different architecture isn't as bank-breaking or time consuming as several people on here believe....or rather they're trying to make others believe.

But you're wrong on this point, and enough people have pointed out why.

Quote:

The same naysayers were complaining that AROS would never be ported to PPC (due to endianess issues and others) and Michal Schulz ported it in several months in his spare time at his own expense. That's a fact and cannot be argued unless you want to argue with him personally.

I never argued to the contrary. However, I pointed out that what he did and what you are proposing are not the same. I see that others have tried to explain to you why one man might be able to do that in several months, but porting OS4 wouldn't. I'm also sick of repeating myself.

Quote:

It doesn't matter if the port is from PPC to x86, or x86 to ARM, or x86 to CUDA. Porting is porting and it's done all the time. Let me be clear, I port "applications" not OSs between several OSs on three different architectures and if it was as difficult as several people on here purport, hell, I wouldn't even get up in the morning! And I never said it wasn't time consuming and at times difficult.

Yet you think that it's a matter of months in order to port the OS. It's not. An OS isn't the equivalent of one app; it's the equivalent of a whole room full of apps.

Quote:

I also understand that Hyperion wants to recoup its development/porting investment by squeezing every last penny out of SAMs, Pegs, and A1s. What developer wouldn't? I do think they are missing out on a lot of revenue now by not having OS4 running on the dominant CPU in the market. Maybe they feel that the Intel market will still be there when or if they ever decide to port to x86. I really don't know what they're thinking. But I do know that the longer they wait, the smaller the market will become for those who want OS4 on x86. They'll move on as many have because the current Amiga h/w platform is woefully underpowered and overpriced.

It's not about just recouping investment, it's about keeping the lights on while you're not making any money at all. Do you know any banker who would be ready to risk money on such a venture?

I personally think that Hyperion's bet bet is to get it working on more hardware, quickly. That last word (in bold) is the key; an x86 port would NOT be quickly.

Hans

_________________
http://hdrlab.org.nz/ - Amiga OS 4 projects, programming articles and more. Home of the RadeonHD driver for Amiga OS 4.x project.
https://keasigmadelta.com/ - More of my work.