What's the deal with NG Amiga OS's?

[hth313]

Originally Posted by kolla

Well, ironically, the "official" OS3 developers are facing the same problems as AROS, and are starting on OS3.1 again - even though Hyperion claim ownership to both OS4.1 sources and OS3.1 sources, the OS3.1.4 developers say that Reaction sources - which indeed is part of OS4.1FE - are off limits. Why is that, do you think?

Well, I do not know about the details here, but this discussion made me kind of go ballistic.

(With "you" below I am not talking to you who are discussing here, but rather people doing shareware, owning copyrights and participating in closed source OS related matters without a viable business plan. If you realize you such person, then yes, I am talking about you).

This all comes down to the usual problem with the Amiga community and is also the reason why it will never go anywhere. It is also the reason that we need to go back to 3.1 and live somewhat "happily" from there.

What I am talking about is all the license mess and all the small popes trying to own their tiny share of something big. This shareware and closed source mentality has taken you nowhere and will just continue to do just that. On top of that mountain of mess we have all the law suites.

I can tell you that by doing this you are just doing harm and not helping mankind or even yourself in any way whatsoever.

Having recently come back to the Amiga I am baffled by all this. It is just so small minded and downright retarded looking at the big picture. In the Haskell community where I do my work, almost everything is open source with liberal non-GPL licenses. Things can be used, inspected, improved and innovated from. Everything just kicks ass and things keeps happening. In the Amiga world... not at all.

You will never get anywhere with the current attitude. Nowhere, I can assure you that.

This is also the reason why we have to go back to 1994 and essentially throw everything away. It is just so infested with stupid small minded closed source mentality it cannot be used. The final 3.1 Commodore source is where it is (or at least should be) possible to establish a proper ownership.

Just open source AmigaOS 3.1, it is the only sensible thing to do if you want to go anywhere. Forget about everything made ever since, unless people step forward and donate the source to the open community.

For you owners of the Amiga OS. If you cannot pay real salaries for at least two full time developers working on the software at this point, just give up! Rest assured, you will never earn money from this. I have tried to do the numbers in various ways using educated guesses and I cannot see how you can cover more than about 10% of the cost of employing two developers with license sales and updates. It is just a black hole for commercial development and will be that forever with the current modus operandi.

P.S.
And by the way, the clock is ticking, the people that might find interest in all this are not getting any younger..

[plasmab]

The copyright owners aren’t interested in making money. It’s a trophy. “I own Amiga” etc. It will never be open. As much as I’d like it to be

[hth313]

Yes, I suppose that is the only sensible explanation.

[plasmab]

Oh and the rough formula for making money as a company is a million USD per developer per annum. You might half that if your costs are particularly low but it’s that order of magnitude to be successful. That figure covers all the other bodies needed to support said Developer...

That’s the ballgame. Amiga is 2 orders of magnitude off that

[kolla]

@hth313

Hear hear Which is why we'd all be better off with Cloanto winning lawsuits, since they on several occasions have been on record stating that they wish to see 68k Amiga OS legally open sourced. Hyperion on the other end, are claiming that all Amiga OS sources are their exclusive intellectual property. And then there are a handful of developers who see themselves as keepers of the holy grail...

[kolla]

Originally Posted by plasmab

It’s a trophy. “I own Amiga”

Exactly.

[hth313]

You could be right on your 2 order of magnitude.

I was "nice" enough to count on developers alone, not really having much overhead and got 1 order of magnitude being on the optimistic (but hopefully realistic) side of revenue.

[plasmab]

Originally Posted by hth313

You could be right on your 2 order of magnitude.

I was "nice" enough to count on developers alone, not really having much overhead and got 1 order of magnitude being on the optimistic (but hopefully realistic) side of revenue.

Developers aren’t the major overhead. Testers are. And with Amiga there are so many permutations of hardware and configuration that testing is v expensive. Add on the fact that good chunks of the OS are in assembler (making unit testing a nightmare) and the bits that are in C have no unit tests adds up to the fact that testing a change would cost a fortune.

Further add on that nobody makes money from operating systems these days.. they are a loss leader to sell other software and services.

It’s not just Amiga. It doesn’t make sense to develop operating systems anymore. Even Sun/Oracle stopped doing it because it didn’t make sense and they have a much bigger user base with customers that are prepared to pay.

EDIT: of course you can take a risk and ship without comprehensive testing but if something goes wrong it could cost you more to put right than you made. You would be better investing your money in the stock market. And this whole venture would need startup capital... that has to come from somewhere

[OlafSch]

Originally Posted by swinkamor12

Amiga OS 3.1 API was outdated even before Commodore bankruptcy.
Very little from Amiga OS 3.1 API is still worth use today.
Except BOOPSI and datatypes almost nothing.
If You have OS that is binary and source compatible and You want to use 68k code in Your new software it may still make sense to use 3.1 API but it not work for x86 AROS.
Forget about AROS on x86 it is not worth of use crap, it is not binary compatible and it has not memory protection, it has not support for multi cores, drivers are many times slover than win/osx/linux.
Use Amiga Os in uae instead, it will be only little slower but binary compatible.
On Amiga Os 4.1, MOS, Amiga OS 3.x in uae You can use API form 3.5,3.9 and other like MUI, Reaction.

if all the nonsense is spread I have to react...

Classact and MUI works on Aros (I must know because I developed a distribution on Aros 68k).Did not test Reaction because closed source but assume it would too. There it is also binary compatible. It has no support for multi-core on 68k but Amiga OS is neither binary compatible nor has memory protection or multicore support either. It just runs faster but then if running Aros 68k in UAE on modern hardware I also outperform most of the PPC hardware.

[gulliver]

Originally Posted by plasmab

Developers aren’t the major overhead. Testers are. And with Amiga there are so many permutations of hardware and configuration that testing is v expensive. Add on the fact that good chunks of the OS are in assembler (making unit testing a nightmare) and the bits that are in C have no unit tests adds up to the fact that testing a change would cost a fortune.

Further add on that nobody makes money from operating systems these days.. they are a loss leader to sell other software and services.

It’s not just Amiga. It doesn’t make sense to develop operating systems anymore. Even Sun/Oracle stopped doing it because it didn’t make sense and they have a much bigger user base with customers that are prepared to pay.

EDIT: of course you can take a risk and ship without comprehensive testing but if something goes wrong it could cost you more to put right than you made. You would be better investing your money in the stock market. And this whole venture would need startup capital... that has to come from somewhere

I am witnessing more or less this.

Testing seems the most tedious aspect as mentioned due to hardware permutations, but it is an unavoidable path that development has to go through, to have something that works on most common hardware.

And this unfortunately means that despite all precautions, there will always be some piece of hardware that will behave oddly and make developers life miserable even after release.

And for the money side of things, my guess is that developers are doing it for just pure love and hobby. They only seem to get compensated with beer money when the OS starts selling and bringing back dividens. So it is not a profitable endeavour per se.

And for the parent company, it seems it is a very low risk gamble in their case. They get the money from selling the OS with very little effort and nearly zero investment behind. But they are also aware they wont make themselves rich by selling niche OSes for dead platforms.

It is what is.

[Gorf]

Originally Posted by hth313

What I am talking about is all the license mess and all the small popes trying to own their tiny share of something big. This shareware and closed source mentality has taken you nowhere and will just continue to do just that. On top of that mountain of mess we have all the law suites.

I can tell you that by doing this you are just doing harm and not helping mankind or even yourself in any way whatsoever.

Having recently come back to the Amiga I am baffled by all this. It is just so small minded and downright retarded looking at the big picture. In the Haskell community where I do my work, almost everything is open source with liberal non-GPL licenses. Things can be used, inspected, improved and innovated from. Everything just kicks ass and things keeps happening. In the Amiga world... not at all.

You will never get anywhere with the current attitude. Nowhere, I can assure you that.

Sadly this is so true.
Tons of stuff on Aminet - but most is closed source even if it is freeware ... why?
Even new things are still closed down :-/

That is even true for some AROS specific software ... just compiled for an od version many of just one architecture, and the author no longer active of course.

There are so many small/tiny tools and programs, that so some nice trick, and you would like to see how this is done on an Amiga-like OS ... but no, you have to reinvent the wheel again and again.

Same for the MUI/Zune dilemma: all the closed source custom mcc for MUI ... WHY??? And WHY not opening up MUI?
You could even dual-licence it and allow only other free software to use the open variant...

[Gorf]

Originally Posted by plasmab

It’s not just Amiga. It doesn’t make sense to develop operating systems anymore.

Tell that Google - they should stop developing Fuchsia at once!
Same goes for MS and Win 10 IoT core..
Or Amazon and FreeRTOS...

[gulliver]

Originally Posted by Gorf

Tell that Google - they should stop developing Fuchsia at once!
Same goes for MS and Win 10 IoT core..
Or Amazon and FreeRTOS...

They earn money by selling the services they build around the OS, like online stores, user data mining , pluggings, support, etc.

But if it were for the OS alone, in nearly all ocassions they might end up loosing money.

[plasmab]

Originally Posted by Gorf

Tell that Google - they should stop developing Fuchsia at once!

Same goes for MS and Win 10 IoT core..

Or Amazon and FreeRTOS...

Most of those are Linux based. And none of them make money directly due to sales of the OS.

[Gorf]

Originally Posted by plasmab

Most of those are Linux based. And none of them make money directly due to sales of the OS.

None of the mentioned systems have anything to do with Linux!

(and if or how they make money was never the question.)

[Gorf]

Originally Posted by gulliver

They earn money by selling the services they build around the OS, like online stores, user data mining , pluggings, support, etc.

But if it were for the OS alone, in nearly all ocassions they might end up loosing money.

Probably... but no-one said otherwise. It just makes still sense for these companies to develop operating systems - despite of the global Windows(desktop) and Linux(mobile+server) duopol.

eComStation/ArcaOS and QNX still earn money the traditional way....

[plasmab]

Originally Posted by Gorf

None of the mentioned systems have anything to do with Linux!

(and if or how they make money was never the question.)

That was the whole question!! AmigaOS cannot make money. Did you even read the thread?

The operating systems you've hand picked are all IoT systems. Thats a bubble right now but there is still no money to be made from the OS sales. Once the bubble dies down it will become the same as server/desktop operating systems.

EDIT: And the reason Amazon et al will put money into developing those type of Operating Systems without having the ownership is so they dont have to support them long term. Its the support that kills you.

[gulliver]

Originally Posted by Gorf

Probably... but no-one said otherwise. It just makes still sense for these companies to develop operating systems - despite of the global Windows(desktop) and Linux(mobile+server) duopol.

eComStation/ArcaOS and QNX still earn money the traditional way....

These cases you mention are kind of far fetched:

eComStation has failed in driving a healthy business selling the OS alone, and its user base is continually shrinking despite running on commodity hardware (aka PCs). Revenue was so poor that now it became ArcaOS, which is still a 32 bit OS/2 Warp 4.52 + some drivers and open source components. And the future still looks dim, as it has certainly become a niche dying OS (I cant help but love it anyway!). So it is not a good example of a thriving commercial OS.

And QNX lived well thanks to Blackberry in the early 2000´s and some hardware manufacturers that still support it. But long gone are the days of having a nearly captive embedded market and as a cool desktop solution (remeber the qnx4demo floppy image?), as there are quite a lot of cheaper and better supported alternatives nowadays. So, unless the scenario changes, it is still another dying OS, this one is just dying a more slower death.

So the writing is on the wall for both of them. They are not healthy business examples, they are leftovers of a great past, now inevitably dying despite their coolness factor.

[Gorf]

Originally Posted by plasmab

EDIT: And the reason Amazon et al will put money into developing those type of Operating Systems without having the ownership is so they dont have to support them long term. Its the support that kills you.

In deed it makes almost in any case more sense to keep it open.
And that is why all AmigaOS-like operating systems should be open source

[Gorf]

Originally Posted by gulliver

And QNX lived well thanks to Blackberry in the early 2000´s and some hardware manufacturers that still support it.

Well right now it looks like QNX it the product that keeps Blackberry alive.