@matthey

Quote:

I hope the sources from AmigaOS 4 are not lost.

I hope considering Trevor almost certainly bailed Hyperion out that as a stipulation of that transaction the sources were put in a safe which can be opened and used by A-EON if Hyperion ever go under. That would have been sensible.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

Quote:

matthey wrote: The embedded market is very diverse too. [..]

What I tried to say was that the embedded market is much more diverse than the desktop side.

Quote:

The 68k has been recently used for industrial embedded markets with the Fido. http://www.innovasic.com/products/fido1100-communication-controller

Very interesting to me, thank you for that link.

Quote:

I believe it is important to have an upgrade path to 64 bit even for embedded use. A 32 bit CPU can often be better performance, smaller footprint, more energy efficient and cheaper though. A 32 bit CPU with a small footprint can go further without needing a 64 bit CPU. The Raspberry Pi certainly didn't need a 64 bit CPU as it lacks the memory and expansion options to take advantage of it.

You may be correct, though I think 64-bit for embedded basically means Linux. I suppose a future Amiga OS in such setting could need it, at some point.

I am very positive to your 68k ASIC idea, is there a way it can become reality?

@ matthey

Quote:

I believe the 68k AmigaOS has more competitive features for the embedded market than the PPC AmigaOS does for the desktop market. ... Lack of SMP support, full memory protection and security are devastating to the desktop market but there are embedded applications where they are not needed.

I see your point here. And it sounds good.

Quote:

The 68k has been recently used for industrial embedded markets with the Fido. http://www.innovasic.com/products/fido1100-communication-controller

Very neat. Good features, too. So regarding tapping into the embedded market: maybe a 68k Pi-type device (as you previously proposed) that could be utilized in similar applications as this Fido device? Sure. Works for me.

Quote:

The high "power consumption" of the G5 made the whole system more expensive and louder which some people do care about.

True. Many did care about it, but they bought the computer in spite of that. My thought was that despite the power problems of the G5, many bought it anyways because the computer enables the user to do needed tasks faster than previous G4s. And this leads to the thought that people typically want speed more than they want power efficiency.

Quote:

AmigaOS PPC hardware would need SMP support as well as more software to realize the full value of the hardware but then it would still be overpriced.

You're not their target market, obviously. The A4000 was overpriced when it was released but people still bought it.

Quote:

Classic 68k Amiga hardware does not offer good value either. The situation is not conducive to expanding the Amiga user base.

Then why are the Vampire products so popular? The Vampire, which I consider classic hardware, have value to their purchasers. I would agree that classic hardware would probably not help expand the Amiga user base. Although a standalone Vampire, as has been mentioned, may expand it to those who may have been interested in the Amiga but never bought a Classic.

Quote:

I expect the AmigaOS will leave the clenched fist of Hyperion at some point considering their business ineptness. I hope the sources from AmigaOS 4 are not lost. Maybe the next owner will realize hiding away the sources does not help proliferation.

They could be considered inept by some, but at least they delivered an Amiga product rather than vaporware. I still maintain that putting AmigaOS 4 in the open source world would not improve the Amiga situation and would fragment the AmigaOS even more (as demonstrated here by another open source OS).

@BigD 

Quote:

I hope considering Trevor almost certainly bailed Hyperion out that as a stipulation of that transaction the sources were put in a safe which can be opened and used by A-EON if Hyperion ever go under. That would have been sensible.

Definitely. Hopefully this is true.

@OlafS25

Quote:

to get updated software on AmigaOS running natively you would need someone risking really big money. We saw that with Trevor acquiring several software projects. If the software would have stayed 68k and bugfixes and new features would be added there would have been some potential but to get it running natively you must invest a lot of work by skilled developers (who are rare and of course want to get money) and then you have no new features and just running it natively. To get something that can compete with modern software you would then need to invest lots of time in it with high risk to to get the money back.

I was thinking of simply making the acquired 68k software available (as is) for AmigaOS 3 users. AmigaOS 3 users include those with classic systems as well as those who emulate it. I would guess some of their aquired software is written in 68k assembly, so porting it to PPC would diffinitely be costly.

@davidf215

Quote:

They could be considered inept by some, but at least they delivered an Amiga product rather than vaporware. I still maintain that putting AmigaOS 4 in the open source world would not improve the Amiga situation and would fragment the AmigaOS even more (as demonstrated here by another open source OS).

I agree with your points here. People also complain the AmigaOne is expensive and a PC does a better job of the Amiga experience. But at least they got the AmigaOne out there instead of leaving it all for dead emulators.

I've pointed out before what happens with open source, when people can have full reign with their own ideas, building an evolutionary OS tree of desent with modifcation, using Linux distros as an example.

But I was told aganst such diversity, that if it were open sourced, that it would somehow bring us together as one and be good for the whole community to work together. I'm paraphrasing from memory here. But the simplified version would be no it wouldn't fragment us.

Last edited by Hypex on 11-Feb-2019 at 01:57 PM.

@davidf215

Quote:

I still maintain that putting AmigaOS 4 in the open source world would not improve the Amiga situation and would fragment the AmigaOS even more (as demonstrated here by another open source OS).

There's only one way to find out for sure. Unfortunately, that decision is not ours to make.

The effect of fragmentation among Linux distros is overblown. Most people who use Linux-based operating systems use one of the major distributions, the remaining distros having few users. This is what happens when people's freedoms are not restricted and they can do what they want. Most people want to be lemmings, but some people just want to make operating systems.

_________________
"Unix is supposed to fix that." -- Jay Miner

Quote:

BigD wrote: I hope considering Trevor almost certainly bailed Hyperion out that as a stipulation of that transaction the sources were put in a safe which can be opened and used by A-EON if Hyperion ever go under. That would have been sensible.

The AmigaOS 4 IP could have been used as collateral for Trevor's loan. While Hyperion likely did not have many choices, they were likely worried about the conflict of interest Trevor has with A-Eon. If majority ownership by Trevor had been accepted by Hyperion, Trevor could have forced the sale of Hyperion property to A-Eon and Hyperion minority shareholders could not stop it. The next best option for Trevor would be to get the AmigaOS 4 IP as collateral, get some stock and take any property not nailed down like Warp3D. Then if Hyperion goes bankrupt, Trevor gets the collateral and in the unlikely event they are able to turn the business around, he gets AmigaOS 4 updates. The only problem is, Hyperion went into preservation mode reducing spending and development on AmigaOS 4. This theory would increase the chances that Trevor is the anonymous financial support behind Cloanto. It would not be necessary to win the lawsuit as draining Hyperion into bankruptcy would suffice. It looks like the success of AmigaOS 3.1.4 and the pro bono work of Hyperion lawyers have sustained Hyperion though.

Quote:

hth313 wrote: You may be correct, though I think 64-bit for embedded basically means Linux. I suppose a future Amiga OS in such setting could need it, at some point.

There are some embedded applications which need a 64 bit address space or benefit enough from 64 bit integer data processing (less than high performance general purpose CPUs lose overall performance with 64 bit). It is possible to only use 32 bit addressing on a 64 bit CPU thus avoiding the 64 bit pointer performance penalty and some of the code density loss (something similar to the x32 ABI for x86_64 hardware). Heavy full featured Linux operating systems probably are pushing resource requirements up in some cases. ARM hardware often has an advanced MMU supporting virtual addressing and monolithic kernels so more advanced and bloated OSs are making their way into embedded. I am not opposed to 64 bit or MMUs but I would like to see hardware which tries to improve the efficiency of the microkernel ala the Amiga.

Quote:

I am very positive to your 68k ASIC idea, is there a way it can become reality?

There are several pieces of a puzzle which would need to fall into place. I tried to make it happen several years ago as part of the Apollo Team but was rebuffed, ironically, by Gunnar. The plan would need backing by the right people at this point (they probably know who they are). I am available to try again with the contacts I made.

@Hypex and @bison

BSD, which I prefer over LInux (although LInux is okay), is open source but it too suffers from fragmentation (although not as bad as LInux).

The only version of open source I would favor (and it really isn't open source as most define it) that would help prevent AmigaOS fragmentation, would be very similar to the governance of the OpenGL technology:

Quote:

From the OpenGL About page: The OpenGL Architecture Review Board (ARB), was an independent consortium formed in 1992, that governed the future of OpenGL, proposing and approving changes to the specification, new releases, and conformance testing. In Sept 2006, the ARB became the OpenGL Working Group under the Khronos Group consortium for open standard APIs. The OpenGL Performance Characterization Committee, another independent organization, creates and maintains OpenGL benchmarks and publishes the results of those benchmarks on its Web site.

An "AmigaOS Architectual Review Board" (or whatever name is chosen) would be the centerpoint in maintaining AmigaOS code and standards, thus maintaining OS unity (via compliance testing rules and standards) and prevent it from fragmenting while simultaneously advancing the technology. The Consortium is basically a governing Trust for the technology.

In this manner, the source is available for interested parties who will produce a certified and compliant product/solution. To ensure IP related code changes for AmigaOS by third parties are not lost, the party who is granted access to the source code must agree that all its code changes are owned by the Consortium and must be shared with (given to) the Consorium in order to earn compliance certification. Also, in relation to this thread, if a company or group want the AmigaOS on PowerPC, 68k, POWER, or whatever, so long as the OS complies with the standard, then it can be produced and certified by the consortium (if compliant, of course).

@matthey
Quote:

There are several pieces of a puzzle which would need to fall into place. I tried to make it happen several years ago as part of the Apollo Team but was rebuffed, ironically, by Gunnar. The plan would need backing by the right people at this point (they probably know who they are). I am available to try again with the contacts I made.

Go for it.

Quote:

hth313: I am very positive to your 68k ASIC idea, is there a way it can become reality? matthey: There are several pieces of a puzzle which would need to fall into place. I tried to make it happen several years ago as part of the Apollo Team but was rebuffed, ironically, by Gunnar. The plan would need backing by the right people at this point (they probably know who they are). I am available to try again with the contacts I made.

You have to be realistic here. The Apollo core is nowhere near ready (maturity, feature-wise) for an ASIC. Even less so back when you had the discussions with Gunnar. There was good reason the proposal was rejected at the time... it was far, far too premature.

Quote:

You have to be realistic here. The Apollo core is nowhere near ready (maturity, feature-wise) for an ASIC. Even less so back when you had the discussions with Gunnar. There was good reason the proposal was rejected at the time... it was far, far too premature.

The Apollo Core never will be ready for an ASIC with the current goals and it is unlikely anybody would want to make an ASIC out of it. Instead of hyper optimizing for an affordable FPGA for individual customers, a larger expensive FPGA for developers should be used with the goal of logic as close to the desired ASIC as possible. While the Apollo Core has shown what is possible with compatibility, the project should be working on a 64 bit ISA which is not a dead end (including upgraded 64 bit addressing support and an SIMD unit with floating point support), multi-core and SMP support which retains compatibility, an MMU and perhaps protection ring for a microkernel ala Amiga (memory protection and security), etc. The project moved the 68k Amiga forward and adds a few thousand users but the current long term goal stops well short of modern hardware which requires an ASIC. The team of an advanced project like this would need to work with OS and compiler developers all of which would need support and cohesion.

@davidf215

Quote:

BSD, which I prefer over LInux (although LInux is okay), is open source but it too suffers from fragmentation (although not as bad as LInux).

BSD is also what Apple chose as well to base their MacOS replacement on.

Quote:

The only version of open source I would favor (and it really isn't open source as most define it) that would help prevent AmigaOS fragmentation, would be very similar to the governance of the OpenGL technology:

Sounds like a good concept. They have the position of maintaining one standard. Be good is AmigaOS/X actually had OpenGL as well.

@davidf215
Quote:

An NDA for AmigaOS 4 source code between two companies could work for this.

NDA?
Even Ben Hermans knows that NDAs aren't worth the paper they're written on. Which is why he would want at least $1M in an escrow account to cover against leaks/breaches.

I know because I asked to fork 3.1 source. The asking escrow amount was lower, but still very high.

Quote:

Otherwise, use Aros. Aros source is already available.

I've always admired Aros' mission, and it is one of the most remarkable reverse engineering projects ever. It has its uses, but I'm not sure how effective it would be as an embedded OS.

Also, it fell into the same trap as every other Amiga "NG" OS. They focused so much on OG that now they're struggling to make it truly NG without breaking compatibility.

Last edited by agami on 13-Feb-2019 at 04:03 AM.

_________________
All the way, with 68k

Yes, its quite obvious that the fastest game in town is X86, but thats unrealistic from the point of getting the Dev team to reprogram/convert from PPC to x86.
Ssolie did say that he gets enough of mainstream platform at work. When he gets home and want to have fun, he dabbles with PPC/AOS.
He finds it intresting to work with.

You can have your opinion regarding AOS devs communication or pace, but its hard to argue against his point regarding having fun with what basically boils down to his hobby (with some pay).
He said all of this during a Amigameet in Sacramento a few years ago.

So those wanting a different platform is living in a dreamworld.

@Overflow

Yes, given the resources, it is good to stay with PPC. And yes x86 is faster, but people who want that may choose Aros or emulation on a x86 machine. Or let the Amithlon-guys create something.

Arm is nice, but where are motherboards with support for RadeonHD cards ?

And as mentioned before, there is a nice Blackbird board with Power9 coming soon. But IF AmigaOS was running on such a board people would complain about a modern browser or other productivity software.
I switched from a AMCC 460ex SoC to a A1X1000 PWRficient PA6T, the cpu is no problem at all.

@agami

Quote:

Also, it fell into the same trap as every other Amiga "NG" OS. They focused so much on OG that now they're struggling to make it truly NG without breaking compatibility.

OG? Operating Generation?

@Overflow

Yep. It is better to let Amiga platform die. Slowly and painfully, like you seem to like.































What a relief right now there are at least 3 initiatives that will save Amiga (none of them are "official" by the way) and very far from the crazyness populating here.

Enjoy your soap operas

@Hypex

Quote:

Hypex wrote: @agami Quote: Also, it fell into the same trap as every other Amiga "NG" OS. They focused so much on OG that now they're struggling to make it truly NG without breaking compatibility. OG? Operating Generation?

I think it's a typo.. he surely meant CG (Current Generation). A much more obvious acronym would be CGA (Current Generation Amiga) which as a side note, has fantastic graphics! Much more so than EGA (Elderly Generation Amiga).
[an attempt of being funny here ]

I'm really looking forward for the Next Generation Amiga, should it ever become reality..

Last edited by Amigo1 on 13-Feb-2019 at 07:30 PM.

@Overflow

Quote:

Ssolie did say that he gets enough of mainstream platform at work. When he gets home and want to have fun...

So, because Ssolie get's enough mainstream at work we all have to suffer? Pray tell, what else does King Steven decree us dreamers should wake up to?

_________________
All the way, with 68k

@Amigo1

Quote:

I think it's a typo

Not a typo. OG as in Original Gangster.

https://www.urbandictionary.com/define.php?term=OG

_________________
All the way, with 68k