@ppcamiga1

Wrong. Emu68 emulates the CPU. WinUAE emulates the entire hardware. Try to get it through the inches of bone you call a skull.

_________________
Doing stupid things for fun...

@ppcamiga1

Stop trolling.

@MagicSN

Unfortunately, I don't think he can. All his posts as fat back as I've bothered to look are the same. He calls everyone interested in 68K at any level scum, constantly berates AGA's lack of chunkiness, insists that PPC is the Amiga commodore were supposed to build (we all know it was PA Risc, but whatever), that everyone doing anything to do with 68K must cease and desist and port MUI to Unix instead. Oh, and that MUI is the only part of AmigaOS that is worth keeping. Which as a single statement sums up everything you need to know about his knowledge of anything,

_________________
Doing stupid things for fun...

@MagicSN

Quote:

MagicSN wrote: @Karlos Yes. Well, the current plan is to do the stuff on the library itselves, exactly as you suggest - in highlevel language, so in C.

This all sounds eminently achievable then. In terms of available low level documentation for the videocore, there could be some issues, but as it's supported by Linux DRM, there's still stuff to draw upon. Pun not intended.

Last edited by Karlos on 19-Jan-2025 at 02:49 PM.
Last edited by Karlos on 19-Jan-2025 at 02:48 PM.

_________________
Doing stupid things for fun...

@Karlos

Quote:

insists that PPC is the Amiga commodore were supposed to build (we all know it was PA Risc, but whatever),

Just a little bit of nitpick:

C= planned to replace the Amiga with something running Windows(NT?) on PA-RISC.
This wouldn't have been an "Amiga" even if C= might have called it that (which I doubt).

What the were "supposed to do" is a matter of opinion and with today's hindsight (and considering what was available at the time) I would say that would have been something like Amithlon with API extension based directly on the Linux kernel (read similar to OSX).

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

@Kronos

That's why I appended "whatever". There was no clear continuation through it, but the only plans - however nebulous - that Commodore had towards RISC was Hombre.

We got PPC not because Commodore thought it was a good idea but because phase5 made it happen. It was a reasonable choice at the time, but just as MUI is *not* and *never was* part of AmigaOS, nor was PPC ever something planned by Commodore. Except in ppcamiga1s tiny mind, where "PPC is nice fast and has chunky pixels like commodore should have made" and "MUI is only part of AmigaOS worth porting to Unix". I'm sure you've seen those regurgitated multiple times.

_________________
Doing stupid things for fun...

Back on topic, in case anyone missed it, I think an open source single library retrofit solution is a great idea that in time can benefit other users - even those who already have V4 support on their older cards should someone decide to implement an improved version.

The old modular approach had its charm and wasn't a bad idea in principle, but in practise, it's a lot of moving parts and indirection. You could probably build a more cycle efficient driver for just about every existing supported chip when going monolithic.

_________________
Doing stupid things for fun...

@Karlos

Wrong. Emu68 emulate whole os. Like winuae.
Winuae may use 2D graphics, 3D graphics, audio, network, hdd from host without emulation.

@MagicSN

Stop trolling

@ppcamiga1

UAE emulates the whole Amiga. Emu68K emulates only a 68040. Now go and haunt some other thread you Muppet.

Neither system emulates the OS, the OS just runs on the emulation. There is a difference but it does require more than just one braincell to grasp it, so I appreciate you may struggle. Emulating the OS would involve providing a complete third party implementation of Exec and all the key libraries. Hmmm. Providing a complete third party replacement for Exec. I've heard that somewhere before. And it wasn't UAE or Emu68...

Last edited by Karlos on 19-Jan-2025 at 04:48 PM.
Last edited by Karlos on 19-Jan-2025 at 04:47 PM.

_________________
Doing stupid things for fun...

@ppcamiga1

Quote:

ppcamiga1 wrote: @MagicSN Stop trolling

Go on, keep biting the hand that feeds you.

_________________
Doing stupid things for fun...

@Karlos

Quote:

He calls everyone interested in 68K at any level scum

everyone that start usuall bs about switch from ppc to 68k
where after so many years real 68k still not reach level of cheap pc from win95 era

Quote:

berates AGA's lack of chunkiness

yes this AGA should have chunky pixel in 1992

Quote:

insists that PPC is the Amiga commodore were supposed to build (we all know it was PA Risc, but whatever),

ppc was chosen as next Amiga cpu in 1995
Quote:

that everyone doing anything to do with 68K must cease

with emulated 68k
Quote:

MUI is the only part of AmigaOS that is worth keeping.

yes MUI is the only part of AmigaOS that is worth keeping after swithc to x86/arm
rest is too outdated

@Karlos

Wrong. Uae emulates things like timers, kb, mouse, joystick.
Everything important like 2D, 3D, audio, net, disc has drivers for years and is not emulated.

@ppcamiga1

Yawn.

_________________
Doing stupid things for fun...

@ppcamiga1

Quote:

ppcamiga1 wrote: @Karlos Wrong. Uae emulates things like timers, kb, mouse, joystick. Everything important like 2D, 3D, audio, net, disc has drivers for years and is not emulated.

Wow. I suppose the whole chip set settings, options to emulated specific RTG cards, other esoteric expansions,not to mention the phase5 PPC expansions (which require the specific ROMs to function) are all there just for show and no UAE ever booted until RTG was added.

https://github.com/tonioni/WinUAE/blob/master/custom.cpp that and all the supporting includes and adjacent files for Gayle etc are all just pretend, eh?


Honest, STFU, you don't have a single clue.

Last edited by Karlos on 19-Jan-2025 at 05:01 PM.

_________________
Doing stupid things for fun...

The Pistorm is an exciting and innovative product that AmigaKit is supportive of. The work that Claude and Michal have done over the last few years is truly remarkable. It is encouraging to see how this has helped keep our 68K machines alive. Having 3D added to the feature list is a game changer for Classic Amiga systems and opens many new opportunities.

A-EON Technology Ltd has invested substantial resources over the last 11 years in Warp3D. As the owners of the original source code and name, A-EON is continuing to develop Warp3D for the benefit of the Amiga community.

It is better if A-EON is consulted before a project such as this is announced in public using the Warp3D name and replacing the Warp3D Library. A-EON may have an official way of incorporating these new developments into the Warp3D package. This will be a better way forward than confusing users by producing an unofficial library baring the same name.

_________________
Amiga Kit Amiga Store
Links: www.amigakit.com | New Products | A600GS

@amigakit

I don't think it's sensible to even consider using the name. It's already confusing enough with Warp3D and Warp3D Nova.

_________________
Doing stupid things for fun...

@Karlos

stop trolling.

Karlos Quote:

Back on topic, in case anyone missed it, I think an open source single library retrofit solution is a great idea that in time can benefit other users - even those who already have V4 support on their older cards should someone decide to implement an improved version.

No doubt a single Warp3D.library is more efficient as it avoids jump table function call overheads. There should have been a RTG standard for the 68k AmigaOS like PPC AmigaOS to avoid the P96 and CGFX libraries but the 68k Amiga is a 2nd class citizen that is still not allowed to compete with Trevor's PPC. Warp3D should be a part of the AmigaOS too instead of charging for standard APIs and drivers that come with free OSs but that is what happens when there is no money made on outdated low production hardware anymore. Separating Warp3D into two layers of libraries, the API and hardware driver, is slower but has some advantages. It likely makes development, distribution and bug management easier. There may be cases where the Warp3D.library could be used for multiple drivers at the same time thus saving memory. The same criticism could be leveled against the AmigaOS in general which uses many small modular libraries, some of which could have easily been combined, especially when parent libraries open and require child libraries. Code in general can use function inlining which has a performance advantage up to a point instead of calling functions which has a code sharing memory saving advantage. The performance advantage of integrating the Warp3D.library into one library is likely small most of the time due to other bottlenecks. The assembly optimizations made to the unofficial Warp3D update only added about 1 fps to ~25 fps or 4% performance boost to Quake despite some code reduced in half. The Warp3D.library could likely be reduced to half the original size because the optimization quality is so bad (EGCS compiler?). Some 3D software received a large boost like when indirect rendering was used providing what feels like a doubling of performance and multitasks better than direct rendering but it was removed from Warp3D v5, perhaps due to bugs and poor performance that are more difficult to solve on PPC. PPC developers can not look at the PPC code and see what a mess it is and the 68k was a 2nd class citizen not worthy of a Warp3D v5 update or even fixing critical memory trashing bugs which has not changed with A-Eon acquiring Warp3D. Perhaps some developer will be hired to fix Warp3D bugs and it will end up on A600GS and available for a fee. The Hyperion, A-Eon, and AmigaKit (HAA) Alliance are practically extreme niche software only businesses so they will make customers bleed to obtain software to subsidize Trevor's dead PPC fantasy.

Karlos Quote:

The old modular approach had its charm and wasn't a bad idea in principle, but in practise, it's a lot of moving parts and indirection. You could probably build a more cycle efficient driver for just about every existing supported chip when going monolithic.

It is funny that you talk about a "cycle efficient driver" with very cycle inefficient emulation. This is amateur hour due to noncompetitive Amiga hardware. I know the remaining Amiga optimists look at this and say this will be the highest performance 68k Amiga ever but it is a Frankenstein of cheap parts that is an abomination to the elegance and efficiency of the 68k Amiga technology. Between cheap ARM abominations and expensive PPC abominations, I choose neither like the rest of the sane world. Trevor is doing a good job of selling computers, but they are ARM and x86-64 systems after looking at the extreme lack of value of any Amiga hardware.

Last edited by matthey on 19-Jan-2025 at 06:11 PM.

@Karlos

A-EON maybe can change the official warp3D.library so they can launch the Pistorm 3D library? The developers would have to give their opinion on this idea or maybe another method. It is great that the Pistorm 3D Library is free of charge to Pistorm users with the bounty. Lets hope the idea of confusingly calling it Warp3D.library is averted by them getting in touch with the Warp3D owners, A-EON Technology.

_________________
Amiga Kit Amiga Store
Links: www.amigakit.com | New Products | A600GS