Post WinUAE 3.5 feature suggestions/comments

[PeterK]

Ok, a more serious feature request now:

The last time I tried to compile WinUAE myself was already more than 11-12 years ago on my old Win98 system. At that time there were still some 68k instructions not emulated by the JIT directly, but instead there was a fallback into the non-JIT code. I did not read the WinUAE sources for long. Are all of these old JIT fallbacks into non-JIT code now replaced by direct JIT code? Every forced fallback might result in a massive slowdown during the code execution. One example were the very complex PC-relative addressing modes of the 68k code.

[Toni Wilen]

Originally Posted by gulliver

Ok, sounds fair.

And what about all those graphics card that are not implemented yet? (Domino, Merlin, RetinaZ2, RetinaZ3, ImpactVision24, etc)

AteoBUS emulation? (also OmniBus, Prometheus which has been open sourced, GoldenGate, Crosslink, etc)

Matze`s GBAPII++? (open source)

MNT VA2000? (open source)

VA2000 : Anything that is user programmable: never or at least until it is totally abandoned. I don't add emulation for something that can have "hardware" change suddenly.

GBAPII++ is "just" a "clone" of existing already emulated RTG board.

Main reason for single RTG board was to have RTG mode that works with all operating systems (including unix-based). Only reason there is so many choices: they all use same chip family.

Usual answer: SVGA chip emulations are very complex, I don't do those. It is someone else's problem. Obviously ET4000 does exists because it was very popular PC board but I am not that interested in adding it because it won't enable anything unique, it is "just" another RTG board with practically identical resolution and vram limitations. Only "better" one is Blizzard/CyberVision but Permedia2 chip isn't emulated. (or Voodoo3 but it isn't emulated either)

Prometheus is already supported. (PCI bridges are very simple from emulation point of view)

ISA bus extensions look useless (and probably also very easy to emulate). Always include use case! Minimally it must enable some unique software to run that does something interesting.

GoldenGate: not going to happen without specs (PC side is a blackbox that is invisible to Amiga side). Commodore bridgeboards were only possible because detailed specs was included in official manuals.

Originally Posted by PeterK

Ok, a more serious feature request now:

The last time I tried to compile WinUAE myself was already more than 11-12 years ago on my old Win98 system. At that time there were still some 68k instructions not emulated by the JIT directly, but instead there was a fallback into the non-JIT code. I did not read the WinUAE sources for long. Are all of these old JIT fallbacks into non-JIT code now replaced by direct JIT code? Every forced fallback might result in a massive slowdown during the code execution. One example were the very complex PC-relative addressing modes of the 68k code.

I don't know. I don't touch JIT.

Aranym and WinUAE/FS-UAE use now mostly same JIT so ask them (Frode did some merging and I think some merging is still missing)

[B14ck W01f]

x86 Bridgeboards not manufactured by Commodore.

[honx]

Originally Posted by B14ck W01f

x86 Bridgeboards not manufactured by Commodore.

toni already mentioned, missing specs and documentation would make this impossible.

[gulliver]

@Toni
Whilst I understand your reasons and accept your choices, I still think it is invaluable to emulate all that hardware you find boring/uninteresting, because it is a very good way of documenting hardware and preserving it for future use/reuse (much like MAME guys do).

It is also very useful for preparing and modding a system setup within the comforts of a modern PC and then taking the drive to the real Amiga without any adjustments.

Doing the opposite allows you to verify wether a missbehaviour in an Amiga system is either a software issue or hardware failure.

[idrougge]

Originally Posted by gulliver

@Toni
Whilst I understand your reasons and accept your choices, I still think it is invaluable to emulate all that hardware you find boring/uninteresting, because it is a very good way of documenting hardware and preserving it for future use/reuse (much like MAME guys do).

I don't agree, and for documentation purposes, you first of all need documentation to implement. The MAME guys implement hardware so that they can run software. But there's no software that requires a Merlin card unless you want to experience ProBench. I can expect that there are two people on this planet who would like to do that.

[Toni Wilen]

Amount of unique software vs complexity of hardware needs to be balanced.

SVGA chips are extremely complex and also need complex interfacing to main emulation. (vs sound card or hd controllers or practically everything else)

[gulliver]

Originally Posted by Toni Wilen

Amount of unique software vs complexity of hardware needs to be balanced.

SVGA chips are extremely complex and also need complex interfacing to main emulation. (vs sound card or hd controllers or practically everything else)

I will have to agree that balance is sure a key.
But then WinUAE has reached a point of maturity in that there arent many other features to implement, besides bug fixes. It is rather "complete", so this only leaves room for the more exotic features/functions to be implemented. Of course, this in itself is a good sign of the progress you have made with it.

Anyway, what is your personal/informal roadmap you would like WinUAE to take from here on? I read you mentioned implementing Video Toaster emulation, but beyond that?

[Toni Wilen]

I don't have any real roadmap, just some list of things that would be nice to have, someday, when I feel like doing it.

Priorities change too easily, especially if new information gets released (like hardware specs, rom dumps etc).

[Old_Bob]

If I could have one WinUAE wish granted...

It would be to have a complete programming environment built in to it, with an editor, assembler and debugger all integrated in one nice package. Much like ZX Spin for the good old Speccy.

I guess that's probably a whole ton of work, though. But hey, I can at least wish

B

[Toni Wilen]

Originally Posted by Old_Bob

If I could have one WinUAE wish granted...

It would be to have a complete programming environment built in to it, with an editor, assembler and debugger all integrated in one nice package. Much like ZX Spin for the good old Speccy.

I guess that's probably a whole ton of work, though. But hey, I can at least wish

B

I don't think this is possible with Amiga. Too many different existing development environments with totally different requirements (system friendly to non-system friendly), different debuggers and so on..

[sean_sk]

Would love to see netplay as well. Like the others have said, I can't be bothered with other implementations as I only use WinUAE. The ability to play Lotus 2 online with my mate would be gaming nirvana for me.

[mark_k]

Separate filter/scaling settings for PAL- and NTSC-like modes. Would allow automatic change of aspect ratio. E.g. 3× for both H & V for PAL modes, but 3×H & 3.5×V for NTSC modes.

"PAL-like" modes (meaning ones with almost-square pixels) would be PAL, DblPAL, Productivity, Super72. NTSC-like would be NTSC, DblNTSC, Euro36, Euro72.

[Toni Wilen]

Originally Posted by sean_skroht

Would love to see netplay as well. Like the others have said, I can't be bothered with other implementations as I only use WinUAE. The ability to play Lotus 2 online with my mate would be gaming nirvana for me.

I have some plans but I am still not sure how it fits and if it really can work properly..

Originally Posted by mark_k

Separate filter/scaling settings for PAL- and NTSC-like modes. Would allow automatic change of aspect ratio. E.g. 3× for both H & V for PAL modes, but 3×H & 3.5×V for NTSC modes.

"PAL-like" modes (meaning ones with almost-square pixels) would be PAL, DblPAL, Productivity, Super72. NTSC-like would be NTSC, DblNTSC, Euro36, Euro72.

Problem with these kinds of features is that this is not complete enough: it needs to support all modes by mapping modes to different filter settings. (displaydata config lines sort of do it already but it is not GUI friendly)

It must also optionally support RTG boards and video port adapters, also by mapping filter mode to specific board.

[Steve]

Originally Posted by Old_Bob

If I could have one WinUAE wish granted...

It would be to have a complete programming environment built in to it, with an editor, assembler and debugger all integrated in one nice package. Much like ZX Spin for the good old Speccy.

I guess that's probably a whole ton of work, though. But hey, I can at least wish

B

Wow, I love this idea! Like a simple assembler IDE built into the WinUAE interface with integrated debugging has a run/debug button which opens a new window with your code etc. A nice simple assembly language compiler built into WinUAE. It would definitely make tinkering with the Amiga a lot more 'immediate' than starting up an emulated version of Devpac. As long as it compiled the code with 100% accuracy then I think it would be a great feature.

[ross]

Originally Posted by Steve

Wow, I love this idea! Like a simple assembler IDE built into the WinUAE interface with integrated debugging has a run/debug button which opens a new window with your code etc. A nice simple assembly language compiler built into WinUAE. It would definitely make tinkering with the Amiga a lot more 'immediate' than starting up an emulated version of Devpac. As long as it compiled the code with 100% accuracy then I think it would be a great feature.

Hi, take a look at this:
http://eab.abime.net/showthread.php?t=87570
It's not exactly what you ask, but great and nice additions.

Cheers,
ross

[Steve]

Originally Posted by ross

Hi, take a look at this:
http://eab.abime.net/showthread.php?t=87570
It's not exactly what you ask, but great and nice additions.

Cheers,
ross

Cool, thanks for pointing this out.

[ztronzo]

Hello Toni,
If you proceed with NetPlay in the future, I would gladly share the servers I use for the amigalive project if required... (currently running Debian Linux)

Any plans for an OSX Compatible version of WinUAE?
This was be the first question half of the people I met asked for when interested in playing multiplayer through NetPlay.

[idrougge]

Yes, the biggest problem with WinUAE is that it isn't MacUAE.

[DamienD]

...then boot camp your Macs and switch between MacOS and Windows