@matthey

“The MiSTer" at this point isn’t one piece of hardware, it’s the DE-10 nano and a handful of clones, and in addition we have the MiST and all its clones, all of them essentially using the same Minimig-AGA core. So it’s no easy task to determine whether MiSTer outsells the THEA1200, as that is also matter of how you define “sell" and how you define "MiSTer".

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

outrun1978 Quote:

The A1200's appeal is that it looks like an A1200, but I think we will have to see how it performs.

Right. My perception of THEA1200 is that it fits in the category of closed system Minis despite the transition to a full sized Amiga 1200 case and keyboard. Mostly open RPI hardware with retro emulation on the low end and mostly open MiSTer hardware with retro simulation in the mid to high end leaves Minis moving to full size systems with cost reduced hardware using ASICs to improve competitiveness. The move back to full size cases and ASIC hardware comes at a higher development cost but the best relaunched retro systems may be sustainable as they can offer more value for retro use than RPI or MiSTer hardware. The NeoGeo may have been a poor choice to revive compared to the 68k Amiga but it is still likely successful enough that other retro systems will receive similar treatment. More universal retro system support as a kind of cost reduced MiSTer with FPGA support for retro chipsets but hard retro CPU cores could improve value and should be considered too.

outrun1978 Quote:

I'll be honest I am actually spending a lot more time and energy nowadays with my Heber Multisystem 2 which is a fabulous Mister FPGA product. Using both the Amiga Vision and CD32 cores on this machine is giving me enough of an Amiga fix. The fact it can chameleon into many other systems too is a bonus. I really would recommend that people pick one up as it's a fantastic piece of kit.

The problem with the 68k Amiga is that standards and specs are not moving forward without affordable mass produced hardware using ASICs. There are some enhanced 68k Amiga games but it is usually for AGA/RTG, 68060 or less and 32 MiB of memory or less. The AmigaOS was actually downgraded to compiled for the 68000 and MMU use is not being added as it is slow to emulate and uses limited FPGA resources. The 68k Amiga will continue to die without NeoGeo like ASIC treatment where it can provide exponentially better value. Until then, it is just another EOL retro system and not even the best system on the MiSTer. The 68000 NeoGeo was likely chosen for revival using ASICs because it was more impressive than the Amiga MiSTer core but a ASIC 68k SoC with modernized 68060+ CPU cores would destroy it. Even a CD32 with 68060@100MHz upgrade destroys a 68000@12MHz or overclocked 68000@24MHz NeoGeo in performance with good compatibility despite the AGA handicap due to Commodore not taking advantage of one of the most upgradable early retro systems and perhaps the earliest upgradeable console in the CD32. The next earliest upgradeable console was the 3DO console, designed by Dave Needle and RJ Mical like an upgraded Amiga and because Commodore was not upgrading the 68k Amiga fast enough. The CD32 was released September 17, 1993 while the 3DO was released October 4, 1993. The 3DO console has many of the chipset upgrades that many Amiga fans would have liked to see on the Amiga. The 3DO large chipset and higher specs made the recent MiSTer 3DO core challenging to develop compared to the minimalist chipset of the Amiga core. Modern systems with GPU unified shaders for 3D also have minimalist 2D chipsets. Add chunky graphics modes to the Amiga chipset as Commodore planned for AA+ and like some FPGA Amiga chipsets have added and it is ready for 3D. Like CPU performance in a FPGA, GPU performance in a FPGA will never be competitive either. I could live without an Amiga with 3D hardware support but the MiSTer lack of 68k FPU and MMU hardware support is the primary reason why have been reluctant to buy one. As good as MiSTer is in some ways, it does not even reach a 1990 level of 68k performance and features.

kolla Quote:

I wonder what "supports RTG modes� implies, licenced P96 from iComp seems unlikely, but… heh

Amiga Forever still uses the old P96 as I recall. I expect the same for THEA1200. Even if Mike can negotiate with Hyperion to free the 68k AmigaOS including their half of P96 ownership and full ownership of Reaction/ClassAct, iComp owns the other half. P96 and Reaction should be permanently included with the AmigaOS but like Warp3D, which should also be included, some ppl make basic AmigaOS support expensive with add-ons and enhancements. As it is, it would not do any good to have RPi priced 68k Amiga hardware as a complete AmigaOS would cost several times what the hardware costs.

kolla Quote:

“The MiSTer" at this point isn’t one piece of hardware, it’s the DE-10 nano and a handful of clones, and in addition we have the MiST and all its clones, all of them essentially using the same Minimig-AGA core. So it’s no easy task to determine whether MiSTer outsells the THEA1200, as that is also matter of how you define “sell" and how you define "MiSTer".

Comparing THEA1200 to MiSTer sales would be difficult. I would not be surprised if THEA1200 sells at a faster pace when first released but I expect MiSTer hardware to sell better over time. I am just talking about the MiSTer and clones but even that is difficult to define as there are several modular add-on boards and sometimes no case/keyboard sold with a system. MiSTer should not need many add-on boards for the Amiga core to match the unexpandable THEA1200 though? Is the main DE10-Nano board and digital I/O board enough to roughly match for hardware?

@cdimauro

Quote:

Which also means that no such ROMs can be used on RunInUAE or similar products.

RunInUAE doesn't use ROMs. It's a launcher for EUAE. You can use ROMs purchased from Cloanto with EUAE if you want.

@matthey

Half ownership of P96? AFAIK RTG in OS4 is more of a reimplementation than an actual port, and all rights today owned by iComp. Cloanto had an earlier agreement about P96 for their AmigaForever emulation package (not without disputes, of course), but it was hardly transferable to other companies’ products. I don’t see how RGL can provide RTG without either having a license from iComp or distributing the shareware archive from aminet with perhaps a custom installer and uaegfx driver.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

kolla Quote:

Half ownership of P96? AFAIK RTG in OS4 is more of a reimplementation than an actual port, and all rights today owned by iComp. Cloanto had an earlier agreement about P96 for their AmigaForever emulation package (not without disputes, of course), but it was hardly transferable to other companies’ products. I don’t see how RGL can provide RTG without either having a license from iComp or distributing the shareware archive from aminet with perhaps a custom installer and uaegfx driver.

P96 was originally licensed for AmigaOS 4 and not well integrated. According to Google AI, it should be possible to develop a clean room P96 API replacement but it likely was legally safer and cheaper to buy P96.

Is it possible to have a clean room API reimplementation after licensing the API?
Google AI Overview Quote:

Yes, it is possible, but it requires a strict separation of duties to protect against copyright infringement or trade secret misappropriation. A properly executed clean-room approach ensures the newly developed code is a legally distinct, independently created work, not a derivative work of the licensed software. The practice of Clean-room design relies on independent creation, which is a core defense against copyright infringement. How to Execute a Clean-Room Reimplementation To be legally defensible, your project must follow a specific, documented procedure: o The Specification Team ("Dirty Room"): Engineers with access to the licensed code, API documentation, or proprietary information analyze the original system. They study how the API behaves, but their only output is a functional specification document. o Legal Review: The specification document is reviewed by legal counsel to ensure it contains no copyrighted code, proprietary algorithms, or trade secrets. It should outline only what the API does, not how it does it. o The Implementation Team ("Clean Room"): An entirely separate group of developers—who have never seen the original source code or any proprietary internal documentation—writes the new code based solely on the approved specification. Crucial Risks & Caveats o AI Tooling Risks: While using AI agents for clean-room engineering is popular, it carries substantial legal and ethical risks. If the AI was trained on the original code, the resulting rewrite may not be legally considered a clean-room implementation. o Copyright vs. Patents: While clean-room design successfully defends against copyright infringement and trade secret claims, it is usually ineffective against patent infringement. o License Obligations: If the original software has restrictive licensing (such as the GPL or LGPL), relicensing or bypassing these rules purely via an automated "rewrite" remains a highly contested legal issue.

As far as shared ownership of P96, Google AI states the following which sounds reasonable.

Do Hyperion and iComp have joint ownership of P96 RTG system for the Amiga?
Google AI Overview Quote:

Yes, but in a divided capacity. In 2017, the original creators of the Picasso96 RTG system sold the intellectual property to a joint ownership between Hyperion Entertainment and Individual Computers (iComp). However, the ownership is split by platform: o Individual Computers (iComp) owns the rights and development for classic 68k Amiga systems and modern expansions. You can find current updates and licensing for classic systems on the Individual Computers Shop. o Hyperion Entertainment owns and maintains the PowerPC (PPC) port of P96, which is integrated into AmigaOS 4 and its successors.

Based on this, RGL would likely have to license P96 from iComp, as you say, unless Cloanto/Amiga Corp could sub license it to RGL. Sub licensing has been used before by Amiga related businesses but broad interpretation of licenses can and has led to "disputes" and lawsuits. There may be legal loopholes for using the freely distributed and free to end user old P96 version which is on Aminet if a clean room P96 driver was developed. I doubt Elbox licenses P96 for Mediator graphics boards or they would include the latest version. Jens at iComp is unhappy that Amiga related businesses avoid licensing P96 but there is no money in the 68k Amiga without new silicon. It ends up being a similar situation as AmigaOS 4 for a micro niche market where the classes are milked for the AmigaOS, 2D & 3D drivers, enhancer software, etc which still does not compete for features with free Linux. The 68k AmigaOS lacks even more without P96 RTG, potentially Reaction/ClassACT GUI because of Hyperion ownership or networking support but the small memory footprint savings offsets some of a semi-modernized AmigaOS cost when not using memory wasting emulation. The Amiga micro niche milkers and IP squatters are parasites feeding off a dead Amiga corpse that maybe could be revived without them.

@matthey

Cloanto should have one since workbench 3.* uses p96

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

pixie Quote:

Cloanto should have one since workbench 3.* uses p96

Saying "workbench 3.* uses p96" is odd wording. P96 is a patch on the 68k AmigaOS 3.x which has only partial RTG support. It uses exec.library/SetFunction() on many graphics related AmigaOS library functions to replace functions with its own functions that may then call the original functions if no RTG is used. It is hackish in its current form as it patches many different versions of AmigaOS libraries and SetFunction() itself can be problematic with multiple SetFunction() patches of the same function.

I expect "Cloanto should have one" means Cloanto should have a license for P96. P96 is an AmigaOS add-on which is allowed for an owned copy of the AmigaOS and there is no requirement for add-ons as far as license. The old version of P96 was shareware.

https://aminet.net/package/driver/video/Picasso96 Quote:

Picasso96 is ShareWare. Requested fee US$20 or DM30, free to use for PicassoIV and Pixel64 users (the manufacturers of those cards already paid for these licences) and the really few ones that already have sent us a donation. No distribution restrictions as long as the the archive is complete and none of the files within is changed. BBS notes may be added.

It looks like the old P96 archive could be included with the AmigaOS, likely without a license. As I recall, the old P96 archive was included on AmigaOS 3.5 and 3.9 CDs but not installed. P96 or CGFX with graphics card driver(s) usually came with a graphics card or bus card purchase and could be installed on top of AmigaOS installs. Real hardware dried up without new 68k silicon so that revenue stream dried up for the P96 developers. RTG for emulation then became more popular but that is EOL support which customers invest in minimally if at all. P96 devs and now iComp would like to get some revenue from this market but if a clean room developed P96 driver and the old P96 archive are used, it may be legal. Maybe it is possible to change the license but the old P96 is still available on Aminet unlike the old free version of PPaint which A-Eon had removed but Amiga Forever still includes. The Amiga milkers and IP squatters are trying to monetize the micro niche market they perpetuate.

Maybe Mike has had enough and put his foot down. Jens/iComp was licensing the 68k AmigaOS from Hyperion so perhaps thought he had leverage to get Cloanto/Amiga Corporation and their licensees like RGL to license P96. Hyperion also tried to intimidate their way into selling the Hyperion version of the 68k AmigaOS to RGL. RGL performed due diligence and knows who owns the Amiga IP though. The court appeal that allowed Cloanto to sue for Amiga IP copyright violations reinforced this. Now Mike has leverage as he controls 68k AmigaOS distribution and could refuse to sell Jens copies of the AmigaOS, Kickstarts and other Amiga IP licenses he needs to produce hardware. Jens claimed he bought P96 for the Indivision as protection to someone else buying it and locking iComp out of its hardware market but he has bigger concerns now. Mike should be in a much stronger position to negotiate with iComp to permanently include P96 in the AmigaOS with the integration advantages of AmigaOS 4. Likewise, Mike should have leverage to permanently include Hyperion owned Reaction in the 68k AmigaOS with the integration advantages of AmigaOS 4. As much of AmigaOS and add-ons as possible should be in the same repository and available for a low price considering all the free OSs and that the AmigaOS needs development to compete with even the free OSs. Mike has talked about a free and open source AmigaOS which would be an improvement to what he have now. Almost free and partially open source may be more practical, provide more control with fewer forks and provide some development funding. For sure, an AmigaOS with add-ons that costs more than the RPi competition's hardware is obviously not going to proliferate the Amiga.

@Rob

Quote:

Rob wrote: @cdimauro Quote: Which also means that no such ROMs can be used on RunInUAE or similar products. RunInUAE doesn't use ROMs. It's a launcher for EUAE. You can use ROMs purchased from Cloanto with EUAE if you want.

I was recalling that Hyperion had issues with Cloanto because they included old Kickstart ROMs like the 1.3 in their software.

RunInUAE is the perfect product where to include them.

So, it should be something different then.

@matthey

Quote:

matthey wrote: kolla Quote: Half ownership of P96? AFAIK RTG in OS4 is more of a reimplementation than an actual port, and all rights today owned by iComp. Cloanto had an earlier agreement about P96 for their AmigaForever emulation package (not without disputes, of course), but it was hardly transferable to other companies’ products. I don’t see how RGL can provide RTG without either having a license from iComp or distributing the shareware archive from aminet with perhaps a custom installer and uaegfx driver. P96 was originally licensed for AmigaOS 4 and not well integrated. According to Google AI, it should be possible to develop a clean room P96 API replacement but it likely was legally safer and cheaper to buy P96. Is it possible to have a clean room API reimplementation after licensing the API? Google AI Overview Quote: Yes, it is possible, but it requires a strict separation of duties to protect against copyright infringement or trade secret misappropriation. A properly executed clean-room approach ensures the newly developed code is a legally distinct, independently created work, not a derivative work of the licensed software. The practice of Clean-room design relies on independent creation, which is a core defense against copyright infringement. How to Execute a Clean-Room Reimplementation To be legally defensible, your project must follow a specific, documented procedure: o The Specification Team ("Dirty Room"): Engineers with access to the licensed code, API documentation, or proprietary information analyze the original system. They study how the API behaves, but their only output is a functional specification document. o Legal Review: The specification document is reviewed by legal counsel to ensure it contains no copyrighted code, proprietary algorithms, or trade secrets. It should outline only what the API does, not how it does it. o The Implementation Team ("Clean Room"): An entirely separate group of developers—who have never seen the original source code or any proprietary internal documentation—writes the new code based solely on the approved specification. Crucial Risks & Caveats o AI Tooling Risks: While using AI agents for clean-room engineering is popular, it carries substantial legal and ethical risks. If the AI was trained on the original code, the resulting rewrite may not be legally considered a clean-room implementation. o Copyright vs. Patents: While clean-room design successfully defends against copyright infringement and trade secret claims, it is usually ineffective against patent infringement. o License Obligations: If the original software has restrictive licensing (such as the GPL or LGPL), relicensing or bypassing these rules purely via an automated "rewrite" remains a highly contested legal issue.

As the Oracle/SUN vs Google case already proved, APIs aren't IPs neither they can be copyrighted blocking the development of clones (like Google did with Java's API).

So, the answer is definitely yes: it's always possible to make new implementation of P96. And of Warp3D as well.

As long as you don't use their copyrighted names (if any registered), it's absolutely fine.
Quote:

As far as shared ownership of P96, Google AI states the following which sounds reasonable. Do Hyperion and iComp have joint ownership of P96 RTG system for the Amiga? Google AI Overview Quote: Yes, but in a divided capacity. In 2017, the original creators of the Picasso96 RTG system sold the intellectual property to a joint ownership between Hyperion Entertainment and Individual Computers (iComp). However, the ownership is split by platform: o Individual Computers (iComp) owns the rights and development for classic 68k Amiga systems and modern expansions. You can find current updates and licensing for classic systems on the Individual Computers Shop. o Hyperion Entertainment owns and maintains the PowerPC (PPC) port of P96, which is integrated into AmigaOS 4 and its successors. Based on this, RGL would likely have to license P96 from iComp, as you say, unless Cloanto/Amiga Corp could sub license it to RGL. Sub licensing has been used before by Amiga related businesses but broad interpretation of licenses can and has led to "disputes" and lawsuits. There may be legal loopholes for using the freely distributed and free to end user old P96 version which is on Aminet if a clean room P96 driver was developed. I doubt Elbox licenses P96 for Mediator graphics boards or they would include the latest version. Jens at iComp is unhappy that Amiga related businesses avoid licensing P96 but there is no money in the 68k Amiga without new silicon. It ends up being a similar situation as AmigaOS 4 for a micro niche market where the classes are milked for the AmigaOS, 2D & 3D drivers, enhancer software, etc which still does not compete for features with free Linux. The 68k AmigaOS lacks even more without P96 RTG, potentially Reaction/ClassACT GUI because of Hyperion ownership or networking support but the small memory footprint savings offsets some of a semi-modernized AmigaOS cost when not using memory wasting emulation. The Amiga micro niche milkers and IP squatters are parasites feeding off a dead Amiga corpse that maybe could be revived without them.

AFAIR the problem which Jens has with Cloanto/Amiga Corp is that he can have problems with P96 because it includes part of the Amiga software to properly work, and this is clearly an IP from Amiga Corp, which is protected by the copyright law.

But I might recall badly, eh!