@cdimauro

Quote:

cdimauro wrote: PowerPCs have no future, so it's "funny" to see that the PowerPC-based entries dominate on the votes.

Yeah, you know, vote like crazy about a bright future for PPC (and talk "positive" about it in online forums) and it might just happen, like a self-fullfilling prophecy...

Quote:

@TRIPOS: I see no future on keeping the same Amiga o.s. limits. There might be a future only for solutions that remove such limits, even if this means dropping the compatibility with the legacy software.

Hence the MorphOS team set out a new future path in 2011 (or earlier) for achieving exactly this, on a new ISA. IMHO there is no "more future" (to connect with the topic of this thread) than this!

But again, if your perspective is more of a retro point of view, then your personal Amiga future is probably more in line with real Amiga H/W and the real Workbench OS, right?

Between these two options, the current "NG" operating systems seems kind of "stuck in the middle"; not being entirely retro, but not being very future either...

@TRIPOS: my retro point of view is that WinUAE is THE solution for such stuff. It allows to run/play any legacy 68K stuff with the highest level of compatibility.

That's also if you are interested on running applications & Amiga o.s. at impressive speed, thanks to the JIT plus the ability to run on very high-end hardware (to give a quick data: SysInfo just gave me 485 MIPS and 90 MFLOPS with WinUAE 3.2 x64, on my i7 4790K. I used SysInfo because it's the most used benchmark on Amiga, but I prefer some real application), and there's room for several improvements, especially on the FPU side.

This picture might change with some FPGA-based project, but only if it gives a concrete added value for its money. Otherwise it doesn't make sense to invest here.

WinUAE introduced support for PowerPCs, and OS4 already runs on it, so it can also be used for legacy PowerPC apps for this o.s. (albeit some limits should be removed, as we know). If the latest MorphOS will be ported again for Amigas with BlizzardPPC or CyberPPC boards, the same applies to it.

So, legacy preservation is essentially done.

Regarding the future, AROS already runs on the latest modern hardware: there's a very bright (and cheap) future here. It also removed the 2GB memory limit, so it's concretely one step behind all other post-Amiga o.ses.

Of course, it still lacks memory protection, resource tracking, SMP, and... a decent filesystem. But what's really missing is... developers. With only 4 or 5 developers which do something only on some spare time, it's not possible to complete the 1.0 goals in short terms, and then seriously thinking about those innovations.

From what I've seen on Neuss, the MorphOS team is the biggest and most active, so I think that MorphOS/x64 will arrive well before that AROS reaches the 1.0 goals...

Regarding OS4, it's bound to the dead PowerPCs with no plan (or possibility) for an ISA change, developers continue to patch the old Amiga sources to introduce small things to strictly keep backward compatibility with 68K and already existing PowerPC apps, and, after years of work, trying to introduce SMP became a nightmare for them and there's no light seen at the end of the tunnel. Here a deep mindset change should be made by the company, but I don't see any signal in this direction, so I don't think that OS4 will have a future.

@cdimauro and @TRIPOS

Quote:

PowerPCs have no future, so it's "funny" to see that the PowerPC-based entries dominate on the votes.

If wishes were horses.

For those voters the future only goes as far as 2016. But I can't really blame them; No one has provided a long term roadmap and they have been fed a steady diet of "next year".

_________________
All the way, with 68k

@KimmoK

Quote:

The name and genuine source code evolution from AOS1 are the strong points for AOS4.

I always wonder how much original 68K Commodore source code is left in OS 4.1 now? Surely it's all been rewritten by now?

@wawa

Quote:

some really motivated souls have been voting their fingers wound, yet none in his right mind will believe there are more than hunderd active posters left in here, right?

Not surprising, but a shame to be honest.

I'm mainly a MorphOS user, but I voted for AROS (despite the fact that I don't really use it), just because logically, it has the most chance for future survival in the current climate.

AROS runs on cheap commodity hardware, x86, ARM etc. and it's free. The main problem with it, is lack of developers. Maybe when MorphOS ports to AMD64 there'll be closer work with the AROS team. In an idea world, when OS4/MorphOS projects do end one day, it'd be nice if the source was released and merged into AROS. That seems to have the most chance for survival in the long run.

Last edited by danwood on 19-Nov-2015 at 08:55 AM.

@danwood

>I always wonder how much original 68K Commodore source code is left in OS 4.1 now? Surely it's all been rewritten by now?

Yes, it must have been rewritten by now. Technically clean rewrite is better than just maintaining old code forever.

But still there is difference between rewriting old code vs reverse engineering a compatible SW module.

MorphOS has reverse engineered compatibility to AOS. And it does not make it technically any worse than AOS4 (it might be better, actually). Some just prefer continuation of original OS, as long as it evolves.



(I use both/all when I can. +Just wish there could be agreement of unifying R&D resources to triple the speed of evolution of the true amigan flavor of computing. Nice if we had just retro + NG branch, when we have very little resources.)

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@cdimauro

even if MorphOS/X64 reaches goals before Aros 1.0 (whatever that is, SMP is already in testing/development like on the Raspberry port or Arix). Of course developers would help to speed up development but I do not see the situation as dire as you. The strategy you propose and MorphOS seems to go to is very risky, there is not much 3rd party software development on and for MorphOS right now and if that changes has to be seen. For me Aros being not just opensource but also cross-platform is much more appealing.

I think MorphOS Team and the AROS developers see the matter very differently. MorphOS Team are from what I've heard no fans of open source. Open source tends to be all over the place, unfocused and it forks all the time, wasting development time. Which is the reason why most of the code they originally used from AROS has been rewritten into closed source so they don't violate any GPL but can keep control.

I'm not saying MorphOS Teams approach is better (I think so personally but it's really a matter of perspective), but it makes a permanent cooperation very difficult for the above reason. Someone has to give, and AROS can't for obvious reasons.

_________________
Amiga Forum - Sweden's best Amiga Magazine

My MorphOS Blog

"Free speech includes other peoples right to offend you."

@Yasu

they have decided how they have decided... I think all NG platforms are heading in different directions anyway and MorphOS team wants to drop legacy features so at the end all will be very different.

Last edited by OlafS25 on 19-Nov-2015 at 10:45 AM.

@Yasu

Quote:

unfocused and it forks all the time, wasting development time.

It's never a waste if it fixes bugs or enhances the platform.


Quote:

Which is the reason why most of the code they originally used from AROS has been rewritten into closed source so they don't violate any GPL but can keep control.

First off you are wrong on both points. AROS is APL not GPL and MorphOS has plenty of open source and AROS components. as a matter of fact if you look at the AROS svn you will see that polluks is backporting quite a bit of AROS to morphos now.

http://repo.or.cz/AROS.git/search/HEAD?s=polluks;st=author

http://www.morphos.de/downloads

There was an issue many years ago about them not releasing the changes however that has since been resolved.

_________________
Support AROS sponsor a developer.

"AROS is prolly illegal ~ Evert Carton" intentionally quoted out of context for dramatic effect

@danwood

Quote:

it'd be nice if the source was released and merged into AROS. That seems to have the most chance for survival in the long run.

i cant speak for others, but i doubt aros developers would want anything to merge in at this point. they might welcome opening the source of genuine amiga system as reference, but aros is done from the scratch with different priorities to start with, like endian independence and portability. replacing the respective parts of source by original amiga code would certainly be a step back in this respect. what concerns opening up os4, if this was ever an option, from what i gather aros developers do not agree with numerous design decissions there, so the gain in this is questionable.

@terminills

Quote:

It's never a waste if it fixes bugs or enhances the platform.

Of course not. But there is no denial that a lot of work in Linux is doing almost the same things but in different branches. Because someone thought "hey, I don't like this particular detail so I'm gonna fork everything and take this to a ever so slightly new direction".

Quote:

AROS is APL not GPL

Yeah, I meant GPL in the broadest sense. There are some Linux stuff there too.

Quote:

if you look at the AROS svn you will see that polluks is backporting quite a bit of AROS to morphos now.

One of the developers at Amiga30 in Neuss told me that a lot of those releases are old and available mostly to not violate any GPL/APL. That they have since written their own code and stopped using the old AROS ones. However, I have no idea of how much AROS stuff they are still using or plan to use.

I feel I have to point out that I couldn't care less if there is AROS code in MorphOS. I do like AROS for what it is and I praise the developers for their achievents. And if the fruits of their labor benefits MorphOS in a better way than the Team think they could do themselves, then all the better as it saves time. As long as the APL is followed of course and credit is given where credits is due.

The point is rather that MorphOS Team and AROS developers operates with very different philosophies. Both have their strong and weak points, but they are difficult to combine.

_________________
Amiga Forum - Sweden's best Amiga Magazine

My MorphOS Blog

"Free speech includes other peoples right to offend you."

@danwood

Quote:

I always wonder how much original 68K Commodore source code is left in OS 4.1 now? Surely it's all been rewritten by now?

I bet it hasn't been. Exec was rewritten, but the rest of it is probably just creaking along.

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

@cdimauro

Quote:

Of course, it [AROS] still lacks memory protection, resource tracking, SMP, and...

Thus making it "stuck in the middle" like the rest of the current "NG" operating systems, neither being what retro enthusiasts are looking for, nor having the essential features of a desktop OS for the future (Amiga or not).

And IMHO the only clean way to proceed, is to assume a "binary view" on backwards compatibility vs. modern (Amiga incompatible) features, it's either this OR that, if you end up in between you will have a messy result with neither full backwards compatibility nor all the required modern features. Again, I think the MorphOS team has the right approach!


Quote:

But what's really missing is... developers. With only 4 or 5 developers which do something only on some spare time, it's not possible to complete the 1.0 goals in short terms, and then seriously thinking about those innovations.

This has been the case from the start. Scattered, slow, unfocused development without a plan or management. This year we celebrate Amiga's 30 year anniversary. Worth remembering is that AROS is almost 20 years old. Some people keeps shouting that only open source like AROS is the way to the future. Me, I look at when AROS started and where it is now and easily draw the conclusion that Open Source has not been the magic formula some people claim.


Quote:

From what I've seen on Neuss, the MorphOS team is the biggest and most active, so I think that MorphOS/x64 will arrive well before that AROS reaches the 1.0 goals... Regarding OS4, it's bound to the dead PowerPCs with no plan (or possibility) for an ISA change, developers continue to patch the old Amiga sources to introduce small things to strictly keep backward compatibility with 68K and already existing PowerPC apps, and, after years of work, trying to introduce SMP became a nightmare for them and there's no light seen at the end of the tunnel. Here a deep mindset change should be made by the company, but I don't see any signal in this direction, so I don't think that OS4 will have a future.

I think you are right, Sir!

@danwood

Quote:

danwood wrote: @KimmoK Quote: The name and genuine source code evolution from AOS1 are the strong points for AOS4. I always wonder how much original 68K Commodore source code is left in OS 4.1 now? Surely it's all been rewritten by now?

Yes, the original OS sources were written in some ancient C language (and parts of it also in some other pre-historic language IIRC), and was rewritten for GCC in a process that also included some cleaning and preparations for migration and less H/W dependencies. Should the 3.1 version that was used as a foundation for OS4 instead be compiled for 68k, it would NOT compile into the same binaries as Commodores 3.1, because the sources simply aren't the same (and neither is the toolchain). It would achieve the same functionality probably, but it wouldn't be the same. And this puts it at the same level as MorphOS and AROS in my books, whose comparable "3.1" components could also be compiled into 68k but wouldn't be the same as Commodores binaries either (they would achieve a similar result, but they wouldn't be exactly the same). But who cares, really? And the 3.1 parts were only a fraction of what became OS4 anyway, even in the beginning. Most things are new features, including the exec, either written from scratch or ported from other platforms.

@TRIPOS
Oh come on guys do I need to announce every bit of news to prove that aros has run circles around your beloved os?

...dream your dreams but next year AROS for ARM will have definitely more users (registered) than mos and os4 together. "users" means supporting users... donation/buying whatever... = active imho

The competition is not OS4 or MOS, it is linux. OS4 and MOS are like sisters.. let us act that way ; )

Also while AROS is open source I and others can build their own distribution adding proprietary stuff. So all the talk of open source vs closed source is bs.

AROS has both sites. With ARIX there comes another try to capitalize and push it.

So on topic:
AROS has a bright commercially supported future ahead going to beat most niche linuxes soon.

But as stated above.. AROS + AmigaOS 3/4 + MorphOS should be united as much as possible...

Last edited by phoenixkonsole on 19-Nov-2015 at 07:12 PM.
Last edited by phoenixkonsole on 19-Nov-2015 at 06:49 PM.

_________________
AROS Broadway - AEROS - Aminux - AmiCloud - indieGO! Appstore - AmiWallet - VAN lossless video codec - AMC Amiga media Center -KrypUnite - LibertyNet - MinX - amigaNX

@phoenixkonsole

Quote:

AROS has a bright commercially supported future ahead going to beat most niche linuxes soon.

I believe it when I see it.

_________________
Amiga Forum - Sweden's best Amiga Magazine

My MorphOS Blog

"Free speech includes other peoples right to offend you."

@Yasu
lean back and see ; )

_________________
AROS Broadway - AEROS - Aminux - AmiCloud - indieGO! Appstore - AmiWallet - VAN lossless video codec - AMC Amiga media Center -KrypUnite - LibertyNet - MinX - amigaNX

@TRIPOS

You talk of unlaid eggs so to say :)

A possible new MorphOS on a different ISA is years away (that is official information from the developers who should know ). Aros is already there so we can compare again when MorphOS is running somewhere else, at the moment it is running on PPC and nowhere else and I do not need to remind you comparing PPC versus X86/X64 based platforms. Everything PPC related is toy compared to current standards .

And even if then there is no software for it. 68k integration will be similar to that of Aros based on UAE and feel different compared to current PPC (something that often was used against Aros by MorphOS and AmigaOS fans). So we will see when it becomes real in years perhaps how many of current users really will use it, how much software is ported and which platform is more useful then.

@OlafS25

Quote:

Everything PPC related is toy compared to current standards

As I am still using an Atom based system for some applications, I obviously don't agree with that statement.

Anything that has broken the GHz barrier has plenty of CPU power (even ARM).
What really hampers us in using current hardware to even its fullest potential is our legacy of 31bit memory addressing and our lack of SMP support.

For example, if MorphOS were ported to a 1.8 GHz T2080 based system, in its current state it would be limited to a single thread using about 1/8th of that CPU's potential computing power.
Couple that with a 1.5GB memory limitation, and it does cripple potential performance.

Remove these legacy constraints and you have SEVERAL times the potential performance.

BTW - This applies to AROS as well.
Even with high speed X64 CPUs, with only single thread support and a limited memory address range, potential performance is crippled.