@bison

Yepp. You are probably right. What I would consider a must is native gfx for both 2D and 3D.
If that could be done we would have a real alternative without inventing the wheel that seams impossible for now. It is no reason why we could at the same time still have a real native platform for research.

@bison

Last edited by nikosidis on 03-Apr-2018 at 09:57 PM.

@OlafS25

Thanks for your kind words. I also like your AROS 68k work.

I think it might be a matter of impressing people and bring inn some stuff others don't have.
If AROS show a stable version of 64-bit with SMP and lets say show of a game or raytracer in lightning speed many will go mad about it :)
I realy hope AROS can show that to the Amigaworld some day and if so, many might join the platform.

I think at least we can agree about 64-bit is the future for amiga NG systems. If not NG will stop. We also need SMP supported and maybe there is a underlying layer in the form of Linux or whatever that could take care of the memory protection, driver problem.

@OlafS25

Cannot speak for the other NG OSes, but on MorphOS I cannot see development has stalled:
Major OS update, new IDE, new IMAP email client.

It's not all rosy, but it's quite okay.

_________________
My programs: via.bckrs.de
MorphOS user since V0.4 (2001)

It's a little puzzling that one still can call operating systems that already are around 20 years old "NG" :)

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

maybe "2nd Gen. Amigas", would be a better, or more accurate name for Amiga "NG" systems now?

Edit: actually, AGA A1200's and A4000's could be 2nd Gen, and everything "NG" after that could be "3rd Gen".

Last edited by BoingBear on 04-Apr-2018 at 02:55 AM.

@nikosidis

My mere $.02.

First, I think the only way to go is open source. I'm still in agony that programs I uploaded to Aminet are too difficult to improve or even bug fix these days because (a) I no longer own an Amiga, and (b) even if I did, I lost the code irretrievably when my A500 went up in literal smoke. Seems to me that the only way to avoid that is through open source. So I think AROS is awesome at least in principle.

I've tried out AROS using the Icaros distribution, and now I think it's awesome in practice as well. Paolone even includes a program I wrote for AROS 6 years ago in the Icaros distribution. So I think Paolone is pretty awesome, too.

I've been hoping to scrape together money for a standalone Vampire whenever it's available. I think Vampire + AROS has a great future.

I know a lot of the work that went into this is thankless, I know it's really tough and a true labor of love. So I thank the developers for what they're doing. (Sorry I can't do more than thank in words, but...) I love having that Amiga feel back.

All that said:

(At this point I wrote something fairly extended and then discovered that the main suggestion I had was actually there on the website, and findable without too much work. So I'll stop here.)

Last edited by simplex on 04-Apr-2018 at 08:11 AM.

_________________
I've decided to follow an awful lot of people I respect and leave AmigaWorld. If for some reason you want to talk to me, it shouldn't take much effort to find me.

@fishy_fis

I didn't mean to imply that I think AROS is pointless. Quite the opposite, it's actually my main Amiga experience right now while I'm trying to bring my A1200 to a happy and fully functioning place.

I'm looking forward to trying ICAROS 2.2.2 when it is released and running it on real hardware too.

The crux of my thoughts around AROS are really is the dev community large enough to keep above water on supported hardware or is the effort ultimately doomed and should be shelved in favor of hosting AROS on linux (like AEROS).

What a strange thing this is. This urge to keep a dead platform alive.

@nikosidis

This is going to be a bit of a rant, but alternative OSes like Haiku and AROS have missed a massive opportunity by not focusing on the Raspberry Pi.

To me, the Pi has been the kind of disruptive force that should have driven small operating systems to up-sticks and leave x86 as quickly as their feet will carry them. The Pi costs next to nothing, is available everywhere, is tiny, has a large community and crucially only has a few configurations to support.

Haiku could've sold a shit ton of preconfigured Pis at open-source events and got a load more devs on board. AROS could've attracted plenty of users and devs the same way. Even in the Amiga community there's enough people who've never tried AROS and remain completely ignorant about it, so make it an 'impulse purchase' and maybe they'll give it a go.

Instead we say "here's an OS that you can try in VirtualBox, although it's a bit of a 2nd rate experience that you'll probably tinker with for 10 minutes and give up on. If you have a large house, you can build your own desktop machine with a list of supported components that aren't really supported and might just be broken in the latest build. And by the way, that list of hardare doesn't actually exist anywhere on the web, so you'll have to ask around and do a bit of trial & error."

And we wonder why these OSes haven't taken off! I don't usually criticise open-source devs because they work on what interests them and that's how it's got to be, but over the past few years so many opportunities have been squandered by not supporting the Pi and that's a crying shame.

Even now, I hear MorphOS is potentially going x86. This as Apple and others are looking at ditching x86 for ARM-based machines - jumping on a ship just as the mainstream seems to be jumping off it. If this is the case, I beg them to reconsider. The problems with x86 are never going to go away and the solution is staring you in the face.

Last edited by clebin on 04-Apr-2018 at 11:22 AM.
Last edited by clebin on 04-Apr-2018 at 11:20 AM.
Last edited by clebin on 04-Apr-2018 at 11:19 AM.

@clebin

there was some activity porting AROS native to Raspberry. I remember two devs working on it. I was very enthusiastic and would even have donated money to it because it really made sense. Cheap, lots of users, defined more or less static hardware (unlike PC world). But then it suddenly and quietly stopped, perhaps bigger problems, perhaps loosing interest, real life, whatever. It is the basic problem of Aros, everything depending on few developers with slow progress and halffinished projects. I agree to you... ARM in general and Raspberry in special was a big chance and Aros would have been in a good position...

I would immediately donate some money to a raspberry port even today

Last edited by OlafS25 on 04-Apr-2018 at 11:58 AM.

@nikosidis

Yes killer applications could create more interest. As others stated... linux kernel for drivers with a special hosted version of Aros on it would be a good thing because it would improve hardware support a lot. The problem is that you need selected hardware for it that is supported and you have to invest time to get it configured, most people today are lazy, if it not works more or less out of the box they drop it. Also you need full memory protection for security reasons. But that would propably break most existing amiga software so it is not a simple recompile.

I think most important would be in this order:
extended driver support by using linux kernel
memory protection (for security reasons)
SMP
64bit

64bit and SMP exists to a certain degree. I do not know how useful this version is. Critical are drivers or that you have at least a number of full supported devices. Memory protection is critical if you really want to use it as main system and not rely on obscurity.

Last edited by OlafS25 on 04-Apr-2018 at 11:56 AM.
Last edited by OlafS25 on 04-Apr-2018 at 11:43 AM.

@clebin

It everything was that easy. AROS actualy tried to go for Pi but for now failed. It is to few developers and a very hard task to write drivers etc. etc.
AROS for sure has a hosted version for Pi called AEROS

https://www.aeros-os.org/

Your are damed right about people being lazy. People want something works and not use countless of hours fideling with this and that. I understand it very well.
This is why I would like to see some new AROS laptop hardware for sale preinstalled with everything.
Ready for action.
MorphOS is also in the same situation when people need to find some secound hand old Apple PPC computer and install the OS yourself etc. etc.
This is not sexy.
Find some statement platform, something that works and if people like they can build their own stuff but that is not for regular users.

Other good idea is what we talked about. Use Linux kernel being compatible with almost all hardware and make some USB stick people can buy ready for action actualy being a product with log and manual or let people download and install themself.

Lets say you had the Amiga name and come up with something like that. Amiga on a stick 20£ will run on any computer.

Last edited by nikosidis on 04-Apr-2018 at 12:45 PM.

@OlafS25

RPis are wonderful little beasts. I do think they suffer a bit of an ergonomic problem for anything other than a desktop setup because of all the dangling peripherals like keyboards, mice, external HD etc.

I'd love to see more computer-in-a-keyboard cases, perfect for occasional home use, like a C64 or A500.

@nikosidis

I never said it was easy.

I know how AROS works and I'm not trying apportion blame. I'm just reporting the current situation as I see it. IMO Raspberry Pi was and is the biggest gift to alternative OSes that we've ever had and not one of the OSes that I've been watching over the years has treated it as anything more than a side-show. It's a shame. I know why, but it's still a shame.

(Criticising open-source projects is a tricky and sensitive thing which is why I almost never do it. It's hard to criticise a project without seeming to blame individual devs. It's hard talking about anything strategic without being reminded of the futility of talking about strategy - especially with AROS which is the least strategic project on the planet!)

Last edited by clebin on 04-Apr-2018 at 03:59 PM.
Last edited by clebin on 04-Apr-2018 at 03:54 PM.

@clebin

it is the problem of pure open source... people do what they want as long they want because nothing to earn. In a big community there would be enough people involved so if one or two loose interest there is not a big difference. In case of Aros one or two loosing interest means the project fails. That seems to be the case regarding Raspberry. Missed opportunity as you said...

@clebin

least strategic?

No plan to take over the world?



;)

@All

Ten years ago Intel released the Atom processor, which brought to us the age of Netbooks. These machines are long dead now and almost forgotten, however if only Intel had re-released today the boards and processors netbooks and Atom mini-pc were based on for 30 or 40 dollars, they would have been a good contenders for Raspberry Pi systems, and a perfect mini-platform to run AROS on.

I'd say there are currently 3.5 classes of Amiga, Classic 68k is one, OS4 and MorphOS on PPC is another and AROS mainly on x86 is a third. The 0.5 is Vampire with it's Apollo Core and either OS3.x or Aros68k. I differentiate it from Classic 68k because of the 64bit and AMMX extensions of the core but there isn't much software yet that takes advantage of them. If more software uses the Apollo core features or the core moves to an ASIC it would then be a new class of Amiga for me.

@clebin

Quote:

The Pi costs next to nothing, is available everywhere, is tiny, has a large community and crucially only has a few configurations to support.

Yes, it does pretty much check off all the boxes. I especially like the new Pi 3B+. It's rare for a company to spend an entire development cycle improving an existing product, but that's what they did.

I am disappointed that AROS on Pi seems to have fizzled out. I think I'd like that even better than AROS on Linux, but it's probably a lot more work to get going.

_________________
"Unix is supposed to fix that." -- Jay Miner

I think it is one of AROS' strong points that it can run hosted, so you can have a rock solid foundation in a maintained and developed operating system, and pretty much use AROS as a desktop environment. And this without emulation of hardware etc slowing it down.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC