New fork of AROS 68K: ApolloOS

[Samurai_Crow]

Originally Posted by NovaCoder

Cool



So it will basically work like Linux where you just have to pull the application you want and then make it locally?

The operating system is in C itself but the 68k version supports Amiga binaries.

[AMIGASYSTEM]

I don't see much difference between AROS 68k Nightly (ABI-v1) and ApolloOS (Ex AROS Caffeine) the only difference is the Magellan Dopus5 based Workbench and some commands in the startup-sequence to activate some native Vampire Drives.

[Seiya]

exactly, it looks like the usual heated soup. But the future will tell us if this path is good or not.

[Samurai_Crow]

In the current AROS builds there is unavoidably fallback code to a non-existing VGA mode. This doubles the memory consumption of the hardware abstraction layer while dragging OOP.library in as a requirement to the AROS kernel. By trimming out the abstraction the Kickstart replacement image is expected to shrink.

[AMIGASYSTEM]

Yes on WinUAE you notice less, although to use many OS3 applications you need to replace some native AROS libraries with the old OS3 libraries.

This is my Aros One 68k:

[ Show youtube player ]

[Korodny]

Originally Posted by Thomas Richter

What is coming? Exactly what I was expecting to come, another incompatible fork. This is exactly what happens with open source software - branches, and by that more incompatibilities.

ApolloOS - if it ever evolves into something useful - would be a legal, open source replacement for the illegal, proprietary CoffinOS. That's a good thing.

[IanS]

Originally Posted by Thomas Richter

What is coming? Exactly what I was expecting to come, another incompatible fork. This is exactly what happens with open source software - branches, and by that more incompatibilities.



If you think that this layer is solving problems, I bet you are in error - it will create problems. Not engineering problems, of course, and not right now.



In case you don't understand: You don't solve the problem of a small market and too many incompatible standards by another incompatible standard.


Oh well...

Well there is an answer. Instead of lining the pockets of Hyperion with their legally dubious rights to OS 3.1, work on getting Aros to a state where it has parity or surpasses OS 3.1/3.9

Whether you like it or not, the developers of OS 3.2 are only adding to the confusion and splintering of the Amiga OS. You have created another fork. And even worse, if Hyperion don't get their way in court, all your work could be for nothing.

Remind me again how that helps unify the Amiga market?

[daxb]

But OS 3.2 combines all OS 3 versions. The twenty years old forks are AROS, MOS and OS4. All are what they are. OS3 is 68k based. MOS and OS4 are PPC based. AROS was for PC only and since some years some people try to get it running on 68k but the result is not accepted by many users. So, we have two PPC and two 68k based systems. All have different targets and I don't see how unifying Amiga market would be possible with that. IMO its all 20 years to late at least.

[Minuous]

Originally Posted by IanS

work on getting Aros to a state where it has parity or surpasses OS 3.1/3.9

They couldn't catch up even when AmigaOS 3 development was halted for 15 years. It is falling further and further behind the official OS.

You have created another fork.

Only if you consider every upgrade of a piece of software as a new fork of that software. It is the mainline trunk of development so therefore not a fork, by definition.

[IanS]

Originally Posted by Minuous

They couldn't catch up even when AmigaOS 3 development was halted for 15 years. It is falling further and further behind the official OS.



Only if you consider every upgrade of a piece of software as a new fork of that software. It is the mainline trunk of development so therefore not a fork, by definition.

The only fly in the ointment is the small matter of ownership and subsequent licensing. Because, so far, nobody can prove they own the rights to Amiga OS 3.x. Until that is settled once and for all, you simply cannot class 3.14 and 3.2 as anything other than a fork of the original.

It really is that simple. Whether you like it or not.

As for 68K AROS falling behind, it already surpasses OS3.x in many areas, and that will only improve. Especially if developers like Thomas R and co were to dive in and help sort things out. You can argue all you like, but Aros is open, portable and free, whilst AOS 3.14 and 3.2 are closed, legally dubious and Hyperion are charging for each 'upgrade'.

[AMIGASYSTEM]

IANS, I've been following AROS 68k for many years, although it has improved a lot, the system is still unstable, just change a video resolution to get crash.
I often report problems HERE an example, which almost always remain unsolved and never anyone intervenes to discuss it, so you don't grow up.

If we then add that AROS is divided between a standard OS3 and Vampire, the road becomes more difficult. given the diversity and compatibility between Amiga and Vampire.

[Minuous]

AROS is nowhere near 3.1 let alone any modern version. Even its own page says it is only 82% of 3.1. If it was any good everyone would be using it. To say it has surpassed anything is ludicrous when there are lots of pieces missing and the bits that are present are slow and buggy.

[PeterK]

Originally Posted by AMIGASYSTEM

I've been following AROS 68k for many years, although it has improved a lot, the system is still unstable, just change a video resolution to get crash.
I often report problems HERE an example, which almost always remain unsolved ....

Yes, confirmed, Aros 68k does not install/update anymore onto an existing drive like DH0: since the beginning of this year and the format of executables and ROMs has changed somehow. I had to copy all required new directories and files directly from the ISO over to my existing Aros system and added W and D protection bits afterwards, because the Installer seems to be broken.

Furthermore, I noticed that there are incompatibilities between AOS and Aros concerning datatypes and datatype.library.

[BSzili]

Originally Posted by Minuous

AROS is nowhere near 3.1 let alone any modern version. Even its own page says it is only 82% of 3.1. If it was any good everyone would be using it. To say it has surpassed anything is ludicrous when there are lots of pieces missing and the bits that are present are slow and buggy.

Speaking of which, who needs 3.2, which another incompatible fork, when we already have 3.9?

[Tigerskunk]

Relax, dudes.

It's all just hobby stuff, and everybody can decide themselves where to put time into.

[Minuous]

@BSzili:

3.9 is great and I have used it since its release. Unfortunately it was not possible to continue from the 3.9 codebase for all components, but that work has been redone and further improved. 3.2 is now far ahead of 3.9.

[clebin]

Originally Posted by Minuous

AROS is nowhere near 3.1 let alone any modern version. Even its own page says it is only 82% of 3.1. If it was any good everyone would be using it. To say it has surpassed anything is ludicrous when there are lots of pieces missing and the bits that are present are slow and buggy.

Some elements have surpassed OS 3.9, others lag behind. Some of that missing 18% is no longer relevant and will never be implemented. Bugs and missing features are a problem, yes.

All this has been explained to you many times. Being rude about AROS and its developers, which is pretty much your modus operandi for as long as I can remember, will not fix a single line of code on its own.

[Minuous]

@clebin:

I don't see how I am being rude or to whom. Some of the things stated in this thread are not correct and I thought I should point that out, that is all. I don't know why you perceive that as some kind of personal attack on anyone.

[torturedutopian]

> The twenty years old forks are AROS, MOS and OS4.

Heh, don't want to get into controversies -- I'm way passed that, been upset for way too long for no good reasons ;-) - but MOS is a complete modern rewrite, I wouldn't call it a useless fork

As for finding a proper legal way to provide the V4 with an OS, that's great. Although that's confusing to see there are so many options for a niche market such as ours :-) But let's not get upset and do what people seem right !

[Overmann]

This is great news! I'm in the process of learning to code and it will be a long time until I'm able to contribute to something like this, I'm sure, but I hope they manage to whip AROS into shape as a usable 68k OS.