@agami Quote:

agami wrote: @cdimauro I agree with almost everything you've written here. Just a few comments: For Amiga OS (Classic) Quote: Future: uncertain due to legal battles between Hyperion and Cloanto. I would change that to Future: Uncertain, due to Hyperion. I know it takes two to tango, but we all know who the culprit is in this case.

I share the same idea. However in my previous post I wanted to make a neutral analysis of the current situation.
Quote:

Quote: Regarding the Amiga o.s., it's better to create a new one which is inspired by Amiga, but with modern features and removing the too tight hardware integration. I too have said this very thing, for many years now. Here on this forum, Amiga.org, and even a lengthy article on the old Amiga Round-table podcast web site. Then I see @OlafS25 make the same comment, only to turn around and state how that would not be an Amiga. NEWS FLASH! Nothing will be an Amiga. It was one of those moments in computer history that can't, nor should it survive unchanged in a 21C computing landscape. How many people look at an M1 Mac running macOS 11, with all its objects, support for modern coding languages, and even a Terminal app, and proclaim: "But it's really not a Macintosh LC II running System 7.x. Where's the single tasking? Where's the floppy drive? Where's the single-button mouse? Can it run Mac Paint?" It's like those parents who force their children to live out their long lost dreams of glory.

Olaf is clearly biased towards the Amiga (so, 68K) platform: hence his writings.

I only share his opinion regarding the facts: it's a fact the vast software base for an Amigan is represented by the Amiga (again: 68K) software.

However this platform is dead, as I've stated, so you can and should enjoy it as it is.

Or, as I've suggested, there are two ways to better enjoy both games and applications, and this might revive the platform, but only from the emulation side. And nothing really new, and no "game changer" on 21C, where people enjoy the technology in quite different ways, as you explained also in your example.
Quote:

For the record, again, I firmly believe that Amiga paradigms of the late '80s and early '90s can find a solid user base in contemporary 21C computing. I'm certain it would be a lot larger than the user base of all A-Cube/A-Eon AmigaOS 4.x systems combined. (not as easy a challenge as it may sound). The average new user would buy it because it is cost effective, supports a unique and empowering UX, has the critical contemporary standards-based internet apps (browser/email/synchronous chat/video-conferencing), and is easy to develop for because it supports modern coding languages and frameworks, and IDE.

Here we've different visions. IMO the track is lost and there will be nothing that can revive it.

Only a new Amiga-inspired o.s., which has also modern features, can attract geeks which are interested on new challenges with new cool stuff.
This might bring the platform to get some space in the embedded market, where there are always chances for very lightweight and reactive products. Then it might also gain desktop support.
Linux shows that finding a huge audience in the consumer side is very difficult. But... who knows what can happen.
Quote:

The old Amiga user would buy it because they would see the inspiration of Amiga in the OS and UX, and they can play their old games and run their old apps with the built-in 68k Amiga emulator.

As I've said before, I prefer to use "enhanced" games and applications "on steroids" (and perfectly and transparently integrated on the host o.s.), using modern technologies, but keeping the same "mechanics".

That's because I've already played the old games, and used the applications as they are. I can do it again, and I do it from time to time. But enhanced versions will give the opportunity to jump again on the wagon and spend time on it.

To give an example of what I say, I share the Fightin' Spirit review which also contains the interview that it was made to my graphic artist (Giacinto Platania): http://www.dizionariovideogiochi.it/doku.php?id=fightin_spirit#artwork
If you scroll a little bit down starting from the Artwork section, you'll find an Artwork HD subsection, which shows how he would have realized an "enhanced" (not only HD / higher resolution, because it's using also 32 bit graphics).
Now imagine a similar work done for the music and sound effects, but keeping the same mechanics: the game will give existing Amiga users a motivation to play it again, and it will also open the door to new users.

You can think the same with applications. Imagine Personal Paint, for example, which has its icon your regular desktop: you double-click it, and then it opens a window like any other native applications of the host, and allows you to access the files in your host o.s.. Or, launch Cygnus Editor, and you can do the same. Etc. etc..
But eerything using the native UI widgets that the host o.s. is providing as well as the maximum performances (thanks to a JIT compiler which doesn't need to emulate the original chipset, and it's entirely focused on making the 68K code running as fast as possible with those os-friendly applications).

I think that I've gave a clear idea of what it can be done.
Quote:

It's just like any religion: If you don't get new adherents, the religion dies.

Indeed.


@Rob Quote:

Rob wrote: @OlafS25 Wether or not Hyperion abused the terms of the 2009 settlement was not Cloanto's fight until they made it theirs. Regardless of Hyperion's actions I think Cloanto's purchase of Amiga Inc would have put them on a collision course with Hyperion because the 2009 settlement meant that Cloanto would inherit the same restrictions making the purchase next to pointless for Cloanto.

Michele Battilana tried to find a solution / agreement: going to the court was his last possible action.

Anyway, I assume that if someone kicks your backside you don't react, right? You stay there and "enjoy". The same if someone eats your food: maybe you even say him "Buon Appetito"...

P.S. As usual, no time to read again.

@Rob

finally it is about money... both sides are not really doing it to help the community, there is still money in what people here call classic and because of that both Cloanto and Hyperion are now fighting. If the "classic market" would not grow and money still to earn there would be no lawsuit. Simple as that.

@cdimauro

It depends of how you define "dead" in this context. In sense of todays terms yes indeed, but that is true for all amiga platforms. Even all "NG" camps together propably would only be a low 4 digit number of active users, a couple of thousands at best. You can hardly call it "living". The "classic" camp including people using emulation is much bigger but still it is very small in todays terms. Living means for me new development and new users. Development in 68k area is much more than it was in recent years, both hardware related and software (most games of course, not applications). OS are in development too, like 3.2 or Aros Vampire branch. New hardware is developed or in development like the Vampire range but also other products. And of course hardware is sold. In this context even something like A500 Mini is interesting because it potentially creates interest outside. If users (including emulation) are returning or not is impossible to say. But if you look at activity in many amiga related forums it seems so. To me if looking at it I would say now glass is half full. Of course if you compare it to one of the big markets it is not much even now. It always depends on the perspective. Of course that means platform is and stays retro and certainly not becomes a serious competition on a global scale again. I do not think that this would be possible. To make it modern and competitive you would need to change almost anything and you would need lots of developers to do it. There would not be much left of what we know now and what we like, a platform technical different with different GUI. I do not see the world (both users and developers) waiting for another platform.

Last edited by OlafS25 on 13-Oct-2021 at 11:57 AM.
Last edited by OlafS25 on 13-Oct-2021 at 11:41 AM.

@OlafS25

The mini is a gaming console, very doubtful that it will attract many new users. Or how do you define an Amiga user anyhow? I have lots of friends and colleagues who have huge Amiga games collections on their various emulation platforms, but that’s also all it is - for them, Amiga is just one of many sources for old games, and any update to the OS is not just pointless, it’s frawned upon.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@OlafS25

Quote:

To make it modern and competitive you would need to change almost anything and you would need lots of developers to do it. There would not be much left of what we know now and what we like, a platform technical different with different GUI. I do not see the world (both users and developers) waiting for another platform.

Nah, you don’t need that many developers - a small team of skillful and dedicated people is most often much more productive than “lots of developers”. The important part to bring over would be the Intuition user experience - screens (private, public, different resolutions, dragging…), the menu system on rmb and with multiselect, and a Workbench clone. A DOS and commandline shell resembling current would also be nice, but not a must. Ditto for stuff that is there, but not altered much from day to day, like datatypes, deficons etc.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@OlafS25 Quote:

OlafS25 wrote: @cdimauro It depends of how you define "dead" in this context. In sense of todays terms yes indeed, but that is true for all amiga platforms. Even all "NG" camps together propably would only be a low 4 digit number of active users, a couple of thousands at best. You can hardly call it "living". The "classic" camp including people using emulation is much bigger but still it is very small in todays terms.

That's the point: it's a niche marker, only for some thousand fans.
Quote:

Living means for me new development and new users. Development in 68k area is much more than it was in recent years, both hardware related and software (most games of course, not applications). OS are in development too, like 3.2 or Aros Vampire branch. New hardware is developed or in development like the Vampire range but also other products. And of course hardware is sold.

This is a small stuff. Even the current Commodore 64 market is more living than the Amiga one.
Quote:

In this context even something like A500 Mini is interesting because it potentially creates interest outside. If users (including emulation) are returning or not is impossible to say. But if you look at activity in many amiga related forums it seems so. To me if looking at it I would say now glass is half full. Of course if you compare it to one of the big markets it is not much even now. It always depends on the perspective. Of course that means platform is and stays retro and certainly not becomes a serious competition on a global scale again. I do not think that this would be possible.

Correct, and that's the point.
Quote:

To make it modern and competitive you would need to change almost anything and you would need lots of developers to do it. There would not be much left of what we know now and what we like, a platform technical different with different GUI. I do not see the world (both users and developers) waiting for another platform.

If you want to change something to the current platform then this leads to a failure, as we've seen.

The Amiga and post-Amiga platform cannot be changed, otherwise it simply doesn't work anymore.

You need a NEW platform, which is Amiga-inspired, and which solves all previous limits. This is the only thing which can start a new market and get interest from the external (the real) world.

As kolla reported, implement most of the Amiga concepts (and APIs), and you give the Amigans more or less the same environment which they are used about. But on a modern, solid, platform.

@cdimauro

Quote:

As kolla reported, implement most of the Amiga concepts (and APIs), and you give the Amigans more or less the same environment which they are used about. But on a modern, solid, platform.

If I only had more time.... ARIX

@michalsc Quote:

michalsc wrote: @cdimauro Quote: As kolla reported, implement most of the Amiga concepts (and APIs), and you give the Amigans more or less the same environment which they are used about. But on a modern, solid, platform. If I only had more time.... ARIX

I didn't checked all files, but from what I've seen the project took the right direction, IMO.

It's a pity that you had no time to continue it, but unfortunately time is something which cannot buy on the market.

Any reason why you used C++? I'm not used neither I like it: too much complex, and it might reduce the people that can contribuite.

BTW, I saw that you used Python for generating files, but IMO interleaving so much C/C++ code inside a Python source makes everything less readable. Maybe it's better to separate the C/C++ code to external "template" files, which are loaded by the Python script which then generate the final code.

Python would also be better to for writing tests, but it requires some wrappers.

Anyway, as a starter, ARIX looks good to me, at least for the user-land. My expectation is that the final project develop & use its own kernel.

However, no time, no developement.

@cdimauro

Quote:

It's a pity that you had no time to continue it, but unfortunately time is something which cannot buy on the market.

The project is not closed, it is just on hold since Emu68 occupies 100% of my time right now. But I will gladly come back to it any time.

Quote:

Any reason why you used C++? I'm not used neither I like it: too much complex, and it might reduce the people that can contribuite.

There is no real reason, it came from the early stage where not all decisions are still clear. It can be just as well pure C, or Rust, or whatever. The only thing that matters is the ABI exposed to other components.

Quote:

BTW, I saw that you used Python for generating files, but IMO interleaving so much C/C++ code inside a Python source makes everything less readable. Maybe it's better to separate the C/C++ code to external "template" files, which are loaded by the Python script which then generate the final code.

Yes, that would be definitely better. And it clearly shows that I had not committed too much time to this project.

Quote:

Anyway, as a starter, ARIX looks good to me, at least for the user-land.

That was the aim - provide user-land environment and libraries which are inspired on AmigaOS only. Because of that
- ARIX processes are linux processes with address space separation
- ARIX threads are linux threads
- Messages are passed by value (unix sockets with optional OOB stream), not by reference
- Components used across processes are identified not by pointers but rather by 64-bit time-based UUIDs
- Libraries have to expose the library base and have to expose at least one LVO table, they can also expose symbols
- Mixing linux and ARIX libraries is possible every time
- Automating linking of ARIX libraries as well as using OpenLibrary call are possible
- ARIX is using only few .so objects for its own needs, everything else is going through kernel library which exposes system calls to the user space.

but, the user experience, or developer experience, should remain as close as possible to AmigaOS. Small code snippet from one test:


struct timespec t0, t1;
struct ID empty = NULL_ID;
struct Message msg = { { NULL, NULL, 0, 0, NULL }, NULL, NULL, 0, empty, 0, 0, 0 };
struct ID arixPort = ARIX_PORT_ID;
struct MsgPort *reply = CreateMsgPort();

msg.mn_ReplyPort = reply->mp_ID;
int i;
AssignPath("FONTS", "SYS:Fonts");
AddPort(reply, "Some port name");
FindPort("Some port name");

printf("doing %d iterations\n", NUM_ITER);
clock_gettime(CLOCK_REALTIME, &t0);
for (i=0; i < NUM_ITER; i++) {
PutMsg(arixPort, &msg);
WaitPort(reply);
DiscardMsg(GetMsg(reply));
}


I hope some similarities to AmigaOS can be recognised there ;)

Quote:

My expectation is that the final project develop & use its own kernel.

No, better now. From my experience all Amiga-related development teams are way to small for that. You can spend your life trying to follow evolution of hardware and adding drivers to your own kernel, or you take something developed by thousands of people, full time, such as Linux kernel.

BTW. Linux is not a problem there. Once ARIX would mature, you could use linux user land software (could! not have to!) and mix it with ARIX stuff. Boot time is also not an issue. The linux kernel itself needs only a second or so, and from there ARIX takes over (there is no /sbin/init or anything similar)

almost on-topic :

back @ work after long covid "hiatus"...

If one desire to be open for a mobile or for a tablet product, there are no options, ARM is the way.
(did benchmarks with latest ARM and x86 mobile products)

+ almost on-topic #2 :
Funny to notice how PowerPC still dominate @ Mars.
But also for radiation-affected environments, it seems SW solutions are being put forward to make past radiation-hardened HW solutions nearly obsolete.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@KimmoK

Quote:

Funny to notice how PowerPC still dominate @ Mars.

I hardly doubt it is because of PowerPC in general. The first radiation hardened CPUs from that company were build very long time ago, if I remember correctly the very first were still made by IBM. In areospace technology changing to a completely new solution is not really a good idea. There, slow evolution is much more preferred than any revolutions and pivots.

With every new version of RAD PPC cpu they already have thousands of hours of tests, certifications and documentations, including precise description of every single step in fabrication process. Now imagine the amount of work if they out of sudden decided to move from well tested (in their POV) PPC architecture to completely new and unknown (in their POV, again) ARM. Think of this all certifications, tests, failure scenarios and so on.

No, I really think it was just a lucky (for PPC) pick at the very beginning. The rest is just history. Just like with IBM picking intel CPU for their machines instead of motorola CPU.

@KimmoK

FYI, SpaceX for the Falcon 9 uses Linux and dual-core x86 processors on three redundant computers.

The triple redundancy gives the system radiation tolerance without the need for expensive rad hardened components. SpaceX tests all flight software on what can be called a table rocket. They lay out all the computers and flight controllers on the Falcon 9 on a table and connect them like they would be on the actual rocket. They then run a complete simulated flight on the components, monitoring performance and potential failures.

X86 and Space X delivers value for the US taxpayer instead of "gold plating" BS from IBM, Blue-Origin, ULA (Boeing, Lockheed).

Last edited by Hammer on 15-Oct-2021 at 04:59 PM.
Last edited by Hammer on 15-Oct-2021 at 04:57 PM.
Last edited by Hammer on 15-Oct-2021 at 04:54 PM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@michalsc Quote:

michalsc wrote: @cdimauro Quote: BTW, I saw that you used Python for generating files, but IMO interleaving so much C/C++ code inside a Python source makes everything less readable. Maybe it's better to separate the C/C++ code to external "template" files, which are loaded by the Python script which then generate the final code. Yes, that would be definitely better. And it clearly shows that I had not committed too much time to this project.

Sorry, my intention wasn't to criticize your work.

I rarely develop code by quite some time, but I spend some part of my daily work reviewing code from colleagues or external partners. So, I automatically do the same when I look at some code, and start reporting observations.
Quote:

Quote: Anyway, as a starter, ARIX looks good to me, at least for the user-land. That was the aim - provide user-land environment and libraries which are inspired on AmigaOS only. Because of that - ARIX processes are linux processes with address space separation - ARIX threads are linux threads - Messages are passed by value (unix sockets with optional OOB stream), not by reference - Components used across processes are identified not by pointers but rather by 64-bit time-based UUIDs - Libraries have to expose the library base and have to expose at least one LVO table, they can also expose symbols - Mixing linux and ARIX libraries is possible every time - Automating linking of ARIX libraries as well as using OpenLibrary call are possible - ARIX is using only few .so objects for its own needs, everything else is going through kernel library which exposes system calls to the user space.

Understood and it makes sense.
Quote:

but, the user experience, or developer experience, should remain as close as possible to AmigaOS. Small code snippet from one test: struct timespec t0, t1; struct ID empty = NULL_ID; struct Message msg = { { NULL, NULL, 0, 0, NULL }, NULL, NULL, 0, empty, 0, 0, 0 }; struct ID arixPort = ARIX_PORT_ID; struct MsgPort *reply = CreateMsgPort(); msg.mn_ReplyPort = reply->mp_ID; int i; AssignPath("FONTS", "SYS:Fonts"); AddPort(reply, "Some port name"); FindPort("Some port name"); printf("doing %d iterations\n", NUM_ITER); clock_gettime(CLOCK_REALTIME, &t0); for (i=0; i < NUM_ITER; i++) { PutMsg(arixPort, &msg); WaitPort(reply); DiscardMsg(GetMsg(reply)); } I hope some similarities to AmigaOS can be recognised there ;)

Absolutely!
Quote:

Quote: My expectation is that the final project develop & use its own kernel. No, better now. From my experience all Amiga-related development teams are way to small for that. You can spend your life trying to follow evolution of hardware and adding drivers to your own kernel, or you take something developed by thousands of people, full time, such as Linux kernel.

OK, so and if I've got it right, you don't plan to develop a new kernel, and you'll stick to the Linux one, right?
Quote:

BTW. Linux is not a problem there. Once ARIX would mature, you could use linux user land software (could! not have to!) and mix it with ARIX stuff. Boot time is also not an issue. The linux kernel itself needs only a second or so, and from there ARIX takes over (there is no /sbin/init or anything similar)

OK, but... any plan to move all Linux dependencies to a proper software layer / files? So that the host o.s. interface can be abstracted, and this might allow to have ARIX run on different o.ses?


@KimmoK

Quote:

KimmoK wrote: almost on-topic : back @ work after long covid "hiatus"...

Sorry for that: I hope that you already recovered now.
Quote:

If one desire to be open for a mobile or for a tablet product, there are no options, ARM is the way. (did benchmarks with latest ARM and x86 mobile products)

Try next week with Intel's Alderlake.
Quote:

+ almost on-topic #2 : Funny to notice how PowerPC still dominate @ Mars. But also for radiation-affected environments, it seems SW solutions are being put forward to make past radiation-hardened HW solutions nearly obsolete.

Then why post-Amigans don't move to Mars? It's the best place for them.

You can also ask Elon Musk to give you a ride there: he's already searching people for that.

The biggest problem with Amiga community is that some people do not accept that switch to x86/arm breaks compatibility.
It is simple oldest part of Amiga OS api are big endian only.
It is something that was designed when all civil 32 bit cpu was big endian.
It will not work on little endian.

Here are many dumb people try to payback Amiga users for rejecting their dumb idea of dumb port to x86/arm.
They want to "switch" to x86/arm and expect ppc users to resign from ppc.

This is stupid.
After breaking compatibility new Amiga NG should be hardware agnostic.
It should work on any hardware. ppc, 68k with mmu, x86, arm etc.
Everybody should be free to use new Amiga NG on any hardware that they want.
We do not want another split in Amiga community.

New Amiga Ng should be just Amiga gui and graphics on top of unix.

@ppcamiga1

Assuming OS would work on "any hardware" (what would be a huge task to do in my view...) but what about software? If you only have a "amiga-like" GUI on top of linux and not a port of all libraries there would be a amiga gui with only linux software. What would be the difference/advantage compared to existing linux desktops? I do not really understand your idea currently.

Michal wants to merge Linux and Aros that makes sense and a excellent developer like him can do that. That is much more than just a new desktop.

Last edited by OlafS25 on 24-Oct-2021 at 10:31 PM.

@OlafS25

Quote:

What would be the difference/advantage compared to existing linux desktops?

The difference would be a desktop paradigm that doesn’t constantly change at the whims of its developers. The difference would be a desktop with screens, not just virtual desktops, that can have different resolutions and depths - like Amiga - no linux desktop has this (Haiku does). The difference would be that you can configure these screens to be dedicated to one program, or to be shared (private vs public). The difference would be menu that pops up on pressing right mouse button, and that DOES NOT CLOSE until you have made all your choices and let go of the right mouse button - no linux desktop, or other OS for that matter, has this (afaik). The difference would be a spatial desktop file system browser (aka Workbench) where icons from multiple windows can be lassoed in one go, The difference would be that you can cancel drag’n drop by using right mouse button. Enough already, if you really are an Amiga user, you would know these things and have no need to ask…

Last edited by kolla on 25-Oct-2021 at 05:09 AM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC