Open Source Amiga OS

[Madshib]

Hello everyone. I have been silently watching this web space for quite awhile and only recently decided to jump back into the conversation with the hopes of an A1200 purchase from Petro T. in July.

I stopped using my Amiga when I moved out in 2000 and my dad bought me an Athelon64 based PC. I think it was his way of sending me into the cruel dark world
I used Windoze for the next few years(still do at work, but that's a different life) and discovered Linux and it's community back in 2008. I now use it exclusively. I like the ideas behind the community and the development process that goes into the big hitters like Red Hat and Debian.

Since I've resumed my boyish interest in the Amiga community I've noticed the nature of the development is from those that love the platform with little help. I'm not talking about the PPC route here, I like the idea behind OS 3.9 being continuously developed by fans and gifted (and patient) programmers.

This leads me to my discussion point/question.

Would the Amiga community enjoy seeing the OS Open Sourced for the community to continue development? This would be the original 68k code released into the wild, so to speak.

Also, what type of governance would need to be in place to control and manage it's growth in the right directions? (committee, OS Dev Board, etc.)

Finally, do you think that it would attract more programmers (both beginner and advanced levels) to help revitalize and/or expand the platform as with Linux?

[persia]

The owners of the OS, Amiga Inc, aren't interested, and neither is Hyperion.  Your best bet is to get involved with AROS if you are looking for open source.

[NovaCoder]

Hiya,

The only kind of 'open source' you get with the classic side is when people reverse engineer something (which obviously takes ages).  

I guess in an ideal world, the owners (whoever the hell they are) of the original OS could have released the source code to the community after they'd finished developing it (around the year 2000).

The other problem is that a lot of the original OS was written in 68k assembler which doesn't lend itself to being developed further (due to a lack of skilled asm coders).  
It sounds like Hyperion have spent the past decade converting all of that cryptic assembler code to straight C code.   I guess that makes them the only entity that's really in a position to open source Amiga OS.

Then of course you have AROS which has taken a different approach (built from the ground up).

[Argo]

See http://aros.sourceforge.net/ and http://aros-exec.org/

An open source reimplementation of Amiga OS 3.1.  Currently has versions that run on x86, PPC, and 68k.

[Iggy]

See http://aros.sourceforge.net/ and http://aros-exec.org/

An open source reimplementation of Amiga OS 3.1.  Currently has versions that run on x86, PPC, and 68k.

Well, a reimplementation of the API anyway.
And after all that trouble, what would you have?
An open source variant of an OS over two decades old?
Doesn't sound all that usefull.

What wrong with using the closed source version that makes you interested in this idea.

[commodorejohn]

What wrong with using the closed source version that makes you interested in this idea.

Hahaha, you don't know freetards do you.

[Manu]

Well, a reimplementation of the API anyway.
And after all that trouble, what would you have?
An open source variant of an OS over two decades old?
Doesn't sound all that usefull.

What wrong with using the closed source version that makes you interested in this idea.

What's wrong with having it open sourced so people interested can join easily? There's no
BIG money to make in the "Amiga" OS business anyways.

Also AROS isn't just two decades old, there's lots of new stuff there too. So it's as useful as any of the three AmigaOS 3.1 clones.

[Manu]

Hahaha, you don't know freetards do you.

What do you mean exactly?

[wawrzon]

Hello everyone. I have been silently watching this web space for quite awhile and only recently decided to jump back into the conversation with the hopes of an A1200 purchase from Petro T. in July.

I stopped using my Amiga when I moved out in 2000 and my dad bought me an Athelon64 based PC. I think it was his way of sending me into the cruel dark world
I used Windoze for the next few years(still do at work, but that's a different life) and discovered Linux and it's community back in 2008. I now use it exclusively. I like the ideas behind the community and the development process that goes into the big hitters like Red Hat and Debian.

Since I've resumed my boyish interest in the Amiga community I've noticed the nature of the development is from those that love the platform with little help. I'm not talking about the PPC route here, I like the idea behind OS 3.9 being continuously developed by fans and gifted (and patient) programmers.

This leads me to my discussion point/question.

Would the Amiga community enjoy seeing the OS Open Sourced for the community to continue development? This would be the original 68k code released into the wild, so to speak.

Also, what type of governance would need to be in place to control and manage it's growth in the right directions? (committee, OS Dev Board, etc.)

Finally, do you think that it would attract more programmers (both beginner and advanced levels) to help revitalize and/or expand the platform as with Linux?

you are talking about aros here, as others mentioned, in particular aros68k. it isnt exactly consumer ready yet, but if you wish to try it i can help you to setup it on your machine. note though a plain a1200 would not be eonough. aros needs some 4-6mb ram to fully boot and is better usable with faster cpus.

what concerns open sourcing the genuine os it has been proposed multiple times and against all reason met deaf ears. no chance there. the aros kickstart is the answer to that,

[A6000]

...... note though a plain a1200 would not be eonough. aros needs some 4-6mb ram to fully boot and is better usable with faster cpus.

That is disappointing, AOS would run on a 68000 and 512K ram, what more does AROS do to justify the faster cpu and 6M ram?

[OlafS3]

That is disappointing, AOS would run on a 68000 and 512K ram, what more does AROS do to justify the faster cpu and 6M ram?

Aros was developed for and on X86 and never intended to be run on 20 years old hardware :-)

So it is not optimized for classic hardware whereas AmigaOS (classic) was and is tightly connected to the old hardware and parts of the OS were in the ROM to save memory. Try to run anything modern on A500, if you find anything you can show it to me because I would like to see it.

There is certainly still room for improvement in AROS 68k and the Kickstart Replacement but I do not believe that it will ever work on a 1 MB A500.

[nicholas]

That is disappointing, AOS would run on a 68000 and 512K ram, what more does AROS do to justify the faster cpu and 6M ram?

The same reason that Amiga OS 3.9 won't run on a 68000 with 512K RAM.

In short, C vs Assembly.

[Acill]

Im going to have to also agree that AROS is your best bet here. Its basically what your looking for. Its open and has an active user base as developer base working to improve it.

One of the nice things about it is the relative ease to port over MorphOS and Amiga OS4 code over and from it.

[wawrzon]

That is disappointing, AOS would run on a 68000 and 512K ram, what more does AROS do to justify the faster cpu and 6M ram?

i think the previous answers to that are not exactly spot on. sorry pals.
its true that aros has been developed on comparatively fast x86, but is meant as cross platform and attempts to be also usable on lower end devices such as pi, or the 68k hardware. it isnt optimized enough yet, but the advance towards those lower end systems, in particular genuine amigas forces it into optimizations that could be avoided before. i can observe it very well, running it regularly on amiga hardware. it has become at least two time faster in the last time. its progressing constantly.

also 68k maintainers like winuae autor, toni willen, attempt to allow it run on lowest end hardware possible. aros is compiled without 020 and higher optimizations for now so far i know, so theoretically should run on 68000 cpu with enough ram. on the other hand the question is, why would anyone need or want to substitute operating system on unexpanded a500. no applications need it or could take advantage of additional features on such a limited system, that is actually best used with its genuine 1.3 kickstart. so why bother?

what concerns the system speed and demands, its wirth to mention that on cpu bound tasks, aros is exactly as fast as the genuine aos (i have benchmarked this). the memory allocation is still some 4 times slower i guess, but it is worked on and also tlsf gets implemented. remember also that systems like 3.9 have their demands as well, dont run within 512k ram and need more time to boot. my a4000/060 cold boots from slow internal ide with old slow 1.3 gig drive almost in a half minute, while almost half of that time the drive is spinning up and aros gets softkicked on genuine kickstart after reboot, aros loading up to full desktop/wanderer/workbench would be about 20 seconds then. but booting it to the shell without startup-sequence will take almost an instant.

[ElPolloDiabl]

You can download individual upgrades from Aminet. Or if you want the whole thing free, wait for AROS.

Open source is great, it means no one can ever shelve it and abandon it. Like what happened to Amithlon.

If your asking... Could you buy OS 4.1 and Amithlon and sell them both at $20 with no hardware lock in?