[Vampire] OS3.x VS AROS

[OlafS3]

As a developer, I agree, we should go all in for AROS.

For one, it is open-source. This is a really important for a developer. It helps in understanding the OS and it gives security.
 
Also, being able to have a powerful development machine (i.e. a fast PC) can really improve software quality.

All the effort put into AROS and applications for AROS help AOS4 and MorphOS anyway (as long as these efforts are open-sourced as well).

However, as I understand, Vampire V2 gives you 128MB of memory. This is very little for todays needs.

I would want to see a JDK 8 port for AROS, but that seems just not feasible on 68k (There is JAmiga 2 for AOS4). If we had 4GB, we could bring most modern applications to the Amiga, but I do not know if this is possible on 68k. And this is holding me back... big time :nervous:

Gunnar has told me that they work on a standalone device that will both a lot faster and has more RAM, if I remember right 2 GB. That is enough for even modern needs. Do not ask me when it will be available and how much it costs, I cannot answer :-)

[OlafS3]

I said nothing about anything "for free", I said leave it to the user to decide.
If specific patches are required for the Vampire, release the patches so the user can apply them to his legally bought KS himself, then load it to the Vampire.
I see no problem with this.

As for me, I legally own KS3.1 and OS3.9 and would like to keep using the OS3.9 KS updates.

I think this is the plan. But adapted will be Aros because (as already mentioned) no license fee, open sources and guaranteed availability. And it offers RTG, in case of Aros a implementation of CybergraphX 3.

[OlafS3]

I said nothing about anything "for free", I said leave it to the user to decide.
If specific patches are required for the Vampire, release the patches so the user can apply them to his legally bought KS himself, then load it to the Vampire.
I see no problem with this.

As for me, I legally own KS3.1 and OS3.9 and would like to keep using the OS3.9 KS updates.

double post

[OlafS3]

@Thomas Richter

You are totally wrong there, you think everyone is only motivated by money. I am not against money, expecially when I invest more than spare time in a project. But that is not all... when I started with my 68k distribution I realized the potential of Aros for the 68k community. For some time I got mostly the same things told you are also saying... why Aros there is 3.X

There is 3.X but the sources are more or less useless today even for our small community sticking legally in nowhere.

For me it has the advantage of:
sources free available
no license fee
guarantee not to be destroyed by anyone out there

That are big advantages. Regarding being bad that open source being used for something commercial. Was that serious? If yes I cannot speak for every aros developer of course just for myself... I would be more than happy if it would be used. That was my vision years ago... one OS for real and future hardware and emulation, free distributable and evolving to a kind of standard.

[OlafS3]

@Thomas Richter

From your posts it is obvious that you have not done much with Aros 68k. I have no problem with that, I have the same interest left in 3.X... zero

But do not make judgements just based on opinions because you prefer 3.X

[OlafS3]

As Hyperion is the only entity entitled to develop OS' based on 3.1 source code, an enhanced version of 3.1 from Cloanto would still have to carry an OS3.1 designation.
OS4 for the 68K does not seem that unrealistic if the processor is powerful enough.

AROS68K still is not quite 'done' yet (it has more than a few issues).

And frankly, I don't mind paying for another person's work, so NO open source is NOT a necessity.

Yes... please enlighten me about Aros 68k

And there is no option to get 4.X for 68k, it is about licensing a small modified 3.1 version

[OlafS3]

double post

[OlafS3]

If I remember correctly, also a few gadgets were also backported from Aros to AmigaOS 3.9.

and tons of components can be added or parts replaced. Aros includes  AHI, CybergraphX 3, Themeing, enhanced menu, Poseidon, PCI/Support,  MESA/Gallium (still slow), Windows can moved out of the screen, impoved  memory management and many patches you needed to add on 3.1. You can use  both Magellan and Scalos now (my personal favorite is Magellan).

You can add MUI 3.8 easily by basically replace 2 files. WHDLoad works,  you can add Wazp3D/Stormmesa and so on. When we talk about aros here we  talk about the typical base libraries like dos, graphics, intuition,  gadtools and so on. I could bore everyone by continuing...

I am always fascinated how people that certainly never use aros in  general and aros 68k in special have such firm fact based views. I could  not and would never do that about MorphOS f.e. because I do not own and  use it.

As Terminills wrote you can fork aros if you think it is going in the  wrong direction and 3.1. compatiblity is main goal of aros (already used  against aros in discussions) so Aros devs certainly will not break  compatibility just for fun.

Where Aros is slower certainly when  doing operations in ECS/AGA, that is visible in benchmarks. Aros was  developed for and on X86 with graphic cards so the original hardware  never was a goal. Additionally CybergraphX in Aros 68k is a wrapper on  P96 so I could very much guess that Aros 68k directly addressing the  framebuffer on Vampire is at least not slower than a similar  configuration based on 3.X and P96. As I wrote on ECS/AGA that might be  different but I think most new software will be written for RTG and to  run old software certainly Aros is fast enough (most old games do not  use the OS anyway).

So how I see it to go Thomas Richters route:
Pay Hyperion lots of money for the license
Pay the owner of P96 for a license
Pay  one or more developers for adapting both (in case of 3.1 you certainly  need to sign NDAs what reduces the number of developers being able to do  that)

And then you are very much dependent on the owner of 3.1., you cannot fork and develop in a different direction.

Then what is what you pay for? Just a binary?

I do not see any sense in that...

[OlafS3]

as a example I compared graphics library routine by routine between Aros, 3.5 and 4.X

http://www.aros-platform.de/html/graphics.html

I did not compare in detail if everything is implemented everywhere, just if the routine is available

if you look at it you see that aros has in some cases new routines but also some missing but most are there

I had planned to do more of such comparations but dropped it because of lack of time. Perhaps I will do just to counterproof people that claim aros is not compatible to 3.X

[OlafS3]

Yes, guess what. Do you have the sources of Windows? Or Microsoft word?

You miss the point... you just compare Amiga OS 3.X or 4.X with Windows and Hyperion with Microsoft. I buy Windows as binary but I can trust Microsoft that they invest in their product. In case of 3.X it is obvious that Hyperion has no interest to invest in it, if they can make some cheap money they do but nothing else. So the platform becomes dependent of a very very small company that controls the sourcecode and few developers with NDAs. To me that sounds not very reliable. What happens if company changes direction, looses interest or goes bankrupt or is sold? Your whole business and planning is based on the software, finally software decides what people can do with a platform. So you say API stays stable on closed source platforms... as far as I know this is not the case neither for 4.X nor MorphOS and both are closed. Closed only means one or a small group of people controls the software. You say 3.X would be stable at API in opposite to Aros. That might be because only few people have both access and interest in 3.X. To me that sounds more like stagnation. You say people are using amiga solely for retro to play old games. That is certainly true to a certain degree but I think few people will buy Vampire just to run the old games with 68060 (what makes some even unplayable). My view is people of course still want use old software but also expect new software using new hardware features implemented in the FPGA. People should have the choice, a old patched 3.1. for pure retro but also something new. And I do not think that in the current legal situation 3.X could offer that.

[OlafS3]

@Iggy

A crazy idea and from my side a simple no

But you know that already...

[OlafS3]

Olaf, no to which part?

Personally I'm just not ready to queue up and get bonked over the head again by what is left of the Amiga community.
So no to Cloanto, AInc., Hyperion.

Why not an open solution?

no to the idea to finance a closed OS as foundation for the future, besides would Hyperion never agree to such a idea

I do not know what Cloanto would do or think but you would get in trouble when preinstalling without approval of Hyperion. I see only a future in a open platform not dominated by one entity, and as long people use the OS and not directly hack the hardware or compile with apollo specific commands they are on the save side. Using a closed OS as base is a risky bet and if Thomas Richter mentions Windows here again, Windows exists for decades and is steady in development, something you cannot say about Amiga. The API changes there too, that to "closed is stable", of course mostly you can still use old software. If the Vampire project would change to a closed software like 4.X as main platform I would be out personal, perhaps even out of the community at all.

[OlafS3]

Forgive my ignorance, but to what actual closed-source development are you comparing open-source development of AROS?

Are you fantasising about legitimate, legal and commercial AOS 3.1 development as an alternative? If so, please explain why you think that would be an option. Without this option your pros and cons of open-source and closed-source development are totally irrelevant.

And, given the "competition" of 3.1 as we know it for more than 20 years, I have no doubts that any AROS68k developer interested in replacing 3.1 will be interested in maintaining not only source code compatibility but also binary code compatibility even though new developers will most certainly be discouraged from using some stoneage stuff. After all, without this binary compatibility AROS68k would be as useless for the vampire as some PPC-OSs derived from AOS principles.

3.1. API compatiblity is the main goal of Aros, in fact exactly this was used against Aros in the past claiming that Aros is just 3.1 on X86. If anybody finds incompatilities and how to solve them no Aros developer will be against the commit. And even if direction changes then Aros 68k could be forked and development would be ongoing on its own. From what I have read in forums 4.X is not very compatible to 3.X. It is not meant as bashing, just stating that they made different decisions obviously. Compatible would be unpatched 3.1 but you already can get that if you proof that you own it already. With access to the sources you could of course better integrate patches but you would always be dependent of the owner and approvals and few developers and would need to spend lots of money and at the end only owning binaries. I do not see sense in that. If the community really wants to go that route then be it but it would be not my way. And then they would immediately loose any rights to moan about the situation, they would deserve it.

[OlafS3]

In principle I understand Thomas's reservations, but;

Given the size of the community and the holders of the OS
Given the speed og development, or the lack thereof
Given rather little communication (or slow)
.. etc

Open source starts to look quite enticing.

Again, I agree with Thomas in principle, but OS 3.x doesnt have 10 fulltime developers working on it.

there are only two developers (if I count it right) who have both access and interest in 3.X (including signed NDA with Hyperion) and Thomas Richter is one of them. And then he is praising it as the only way to go and opensource if no alternative. To me it sounds like a mixture of personal and economic interests. Perhaps I am wrong there but Thomas is very negative regarding Aros.

[OlafS3]

Are there any known details about this deal? It explains the statement about open-source and closed-source development which basically comes down to the old argument that a wise dictator is better than a democracy.

Which deal do you mean? There is none.

As I understand it Thomas Richter proposes that 3.X is licensed from Hyperion and P96 from the owner. Then developer have to adapt both to Vampire, of course getting a "meal" or similar for it. You only get access to 3.X sourcecodes after signing a agreement with Hyperion that you will not contribute to any competing OS, ruling out any MorphOS or Aros devs here and narrowing it to few developers, in fact besides him I only remember one other that seemed to be interested in 3.X and has access. That all looks like a receipt to high costs and high risks because in fact you do not really control the situation.