@Kronos

You haven't been reading.

Quote:

Sure and while we are on it lets just fix the other issues like 31Bit addressing, open system structures, lack of SMP support

I said if you can't fix those issues, you may as well embrace them. It was only a few posts ago. Do try to keep up.

These limitations are the 68K shadow that PPC/NG failed to emerge from which is why the NG moniker is a joke anyway*. Since there's more chance of satan ice skating to work every morning than these being fixed in a backwards compatible way before we all die of old age, then the only sensible thing to do is to find a cheaper, more available architecture to enjoy those limitations on. And mate, second hand G5s or gold plated multicore 64 bit PPC with quasi modern GPU are not it. You need readily available, readily supported hardware that doesn't cost an arm and a leg.

*Christ, even my now 11 year old GPU has more RAM than can be safely addressed by 3.x, 4.x or MOS/ABox.

The raspberry pi has been around since 2012 and keeps on going. I knew a lot of people that bought them over the years and I can tell you they were all younger than me and they all wanted them for fun/hobby/esoteric use, save a minority that had some very specific application in mind. You should have tried buying a Pi4 when the chip shortages were at their worst. But even then they were more accessible than PPC/NG HW.

Last edited by Karlos on 02-Jan-2024 at 09:38 AM.

_________________
Doing stupid things for fun...

@Kronos

Way back in the early days of MOS, I remember you, specifically, telling me that the answer to my 3.x dev question on amiga.org was to install MorphOS and carry on from there. You probably don't remember this but I could probably find the exact post if I dug a bit.

You didn't say this because you thought MOS was a deadend platform stuck on an obsolete architecture with no prospects beyond a fully backwards breaking migration to x64.

_________________
Doing stupid things for fun...

@Karlos

what you wrote is crap the same as it was in 2001.
ppc is bad because expensive switch to pc it will be cheaper.
you get the same answer as other in 2001 for pc I have windows.
using ppc is fun because it is not pc.

only what changes is in 2001 was pc today commodity hardware.

like other you expect that people will fool themselves
ad sat not it not pc i cannot switch to windows and do everething one hundred times faster.
it is extremely dumb.

want people to use amiga like solutions on commodity hardware
provide something that will be no more than ten years behind from windows/linux/os x.
stop attacking ppc and start working on something like amiga os x
amiga gui and graphics on unix base.

@Karlos

Quote:

Karlos wrote: @Kronos You haven't been reading. Quote: Sure and while we are on it lets just fix the other issues like 31Bit addressing, open system structures, lack of SMP support  I said if you can't fix those issues, you may as well embrace them. It was only a few posts ago. Do try to keep up.

Embrace them == lets add some for the fun of it
At least that would be switching to 68k mean for me today. If things have changed by 2030 (or sooner) I will revaluate.

As for me making a post a long time ago that you think hasn't aged so well....

Lets say you won the internet for this year.

Last edited by Kronos on 02-Jan-2024 at 09:52 AM.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@ppcamiga1

2001? What bollocks is that? Even 2013 I was still actively working on stuff for OS4.1. Remind us what you were doing?

I didn't give up on NG, NG gave up on itself.

_________________
Doing stupid things for fun...

@Kronos

Quote:

As for me making a post a long time ago that you think hasn't aged so well.... Lets say you won the internet for this year.

No, sir. It's you that hasn't aged well. You used to be optimistic and enthusiastic. I guess being stuck on PPC knocked that shit clean out of you. It figures.

_________________
Doing stupid things for fun...

@Karlos

There is a clear line between "optimistic and enthusiastic" and "fanboy".

I at some time crossed that line and may have over compensated a bit while others managed to change colors while firmly staying on the same side.

Which is better, which is worse? Reality is reality and history can't be rewritten (it surely can be made harder to see) so I choose to live in it making the best what is while not whining about what could have been or w##king of on what could (but won't) be.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Being optimistic doesn't make anyone a fanboi and that's not a term I throw around either. I don't expect you to do anything, either.

I really think ARM offers an excellent path out of the rut NG is in, but by the time anyone in a position to do anything about it realises, it probably won't be and if it happens at all, it won't be until well after it has ceased to be a good option.

In one of the "Culture" novels by Iain Banks, there's this oddball cult that only eats the waste scraps and entrails while deliberately discarding anything good and wholesome. Consequently, its members are perpetually ill and malnourished as a result. You couldn't ask for a better parallel, really.

Last edited by Karlos on 02-Jan-2024 at 11:34 AM.

_________________
Doing stupid things for fun...

@Karlos

Quote:

Karlos wrote: I really think ARM offers an excellent path out of the rut NG is in,

That "path out" requires some serious development effort on an 68k (or 68/ARM hybrid OS) that just isn't happening right now.

- MorphOS team made it pretty clear that they are not interested
- more or less the same for OS4 (their tangled&disputed ownership situation does not help)
- AmigaOS3.14 started 5 years ago and sofar they fixed some ancient bugs and backported some minor OS4 features. Which is all good for a team of unpaid volunteers. Who made it pretty clear that UAE/Vampire/Emu68 are NOT on their target list and are also by association entangled with the legal mess that everything Amiga(TM)
- which leaves AROS, available to the masses in the form of AfAOS, UAE ROM replacements and special distros for decades, but has somehow managed to gain 0 traction

Except for AROS there is absolutely nothing I could do about that (and very little concerning AROS) and I see no reason to jump from something seen as dead end but working to a "the future is bright" but broken system.

I kinda did that 20 years ago going from Amithlon to pre-April Peg1 and MorphOS1.0 and just don't have the patience to wait for Zombie68k to reach it's "MorphOS_2.0" moment.

If it ever does, or at least gets a clear&plausible roadmap to that point we can talk again.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Karlos

You know, people would ask me what OS4 has above OS3 apart UI, and I struggle to think of anything relevant. But the reality is, I'm just so used to it now there are things I take for granted, but there are changes under the hood. I just don't remember them.

Apart from the changes they made. Some being API changes that were nice for organising but just complicated it for programmers. It's almost a copy of OS3. The internal structure works the same. Well, you've used it, it's almost exactly alike. In the beginning being alike was a feature as the likeness also brought compatibility. But this also revealed a flaw. It was just too close and the problems were soon obvious. All along the excuse for keeping it the same internally was compatibility. But I think this was a fallacy used as an excuse. All 68K code has to be emulated regardless so it's bit of a moot point to claim the NG OS needs to be exactly alike to retain compatibility.

It's somewhat 20 years late, but I think they really needed to design a new API for OS4 native and above. While including a 68K compatibility OS layer. They already needed and still do stubs for 68K code. So that OS4 native would be a clean slate or in the least would have some cleaning up, so problems were not dragged over but solved. Right now, to move forward, it really needs to be split three ways. 68K, PPC, and ExecTG. I made that last one up, meaning Exec Third generation. What that is, regardless of CPU, is a new 64 bit API with SMP support. It would make sense to use existing PPC as it could for once make full use of the hardware. 68K and PPC would likely be sandboxed together. But in such a way as they would be Forbid to interfere with the operation of the system.

@OneTimer1

That's funny because I recall MorphOS on the Pegasos coming out first.

@agami

Quote:

I don't blame the turn-of-the-millennium Amiga Inc. et al for setting the initial PPC direction, but the key concept here is that it was the "initial" direction, and not the be-all and end-all direction.

I don't hold them responsible at all, they had no interest in PPC or even 68K, from what I could tell. I recall PPC announcements but they were just to tease us. OS4 was purely on an x86 developer box, where they were to develop OS5, that supposedly would have come out on both x86 and PPC. But, if you target x86 for a first release, why would you later target PPC where no one was making any commodity PPC hardware?

Quote:

I blame Hyperion and A-EON for sticking with PPC at a time when the perfect opportunity was presented to make a shift:

I forgot to mention they kept it going. There's obvious politics at play here. But, OS4 was just released before Apple killed PPC. They just got it working when suddenly the plug was pulled. By comparison the Pegaos was doing well until the ODW where it disappeared and later the well diminished Efika with unusually lower specs was put in place.

However, there is also a lack of finances, so even porting OS4 PPC to another PPC board would cost in the order of thousands. I wouldn't put it under 5 figures. And that's not even for a new architecture.

x64 is such a huge ecosystem that expecting any random board to be supported, like people would expect from AROS, is rather unrealistic. As well as, the old excuses about not having a license for x86, don't even make a dent when porting AmigaOS to x86 would mean dealing with endian issues.

ARM makes sense and unlike PPC, where people spread rumours of there being reference PPC boards in reach, ARM does actually have reasonable boards you don't need to be a developer for. However, it again has endian issues to deal with, since in opposite to PPC it's natively little endian. There is big support, but it looks like it's being phased out, and in some cases may need hardware support on board. A lot of talk has been made about changing endianess in AmgaOS or making it portable. But, this comes from AmigaOS not using a popular endian, so the argument is really about changing the current popular endian and not being tied to an unpopular one. Not that being tied to an endian is always a problem. Windows is tied to little endian and no one batted an eyelid for that. Of it's also the most popular so doesn't present any problems.

Last edited by Hypex on 02-Jan-2024 at 01:40 PM.

@Hypex

You forget that by that time ownership of OS4 had already fractured due to badly written license agreements or to different badly written license agreements making the developers go home and taking the ball with them.

So Hyperion did not (and still don't) have access to all the source codes needed to make a switch and they for sure don't have the resources to reimplement those parts yet again.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

You make it sounds like a swiss cheese admixture of 68K an PPC but the truth is even ten years ago when I was last active in there's not much in your typical OS4 installation that wasn't totally PPC native today. ARexx is an obvious exception (assuming that's not changed since).

_________________
Doing stupid things for fun...

@Karlos

I've noticed a modern idea is treating 68K as bytecode. Or, for 68K, it would be wordcode.

But this presents problems of it's own. If it's used as a generic base code for portability. Since, by comparison 68K code is not always clean. For example, 68K can modify itself, and code is mixed with data and variables. PPC, based on ELF, is cleaner since it separates all that by design.

Another is that 68K lacks SIMD. Now only specific applications use it and more so since later hardware lacks support. But the OS does use it internally just like the utility library would make use of 68020 code where possible.

For a native custom OS with all apps being loaded and binary compiled at load time on a portable CPU there's endian to think about. Be the OS fully native or the OS running QBox style as another app would determine how much work this is. But, despite all the complaints on the internet, modern CPUs and especially x86 have specific big endian support so it's not a big deal. At least not technically, only compilers have fallen behind. But I found GCC can be told endian for scalars, so AmigaOS x64 here we go, natively compiled to run on x86/64 in big endian. Sure, not as efficient as natural endian code, but it would still be wicked fast and movebe make Amithlon style bswaps practicality archaic.

@Kronos

Quote:

You forget that by that time ownership of OS4 had already fractured due to badly written license agreements or to different badly written license agreements making the developers go home and taking the ball with them. So Hyperion did not (and still don't) have access to all the source codes needed to make a switch and they for sure don't have the resources to reimplement those parts yet again

They had a viable path to more affordable hardware(and more licences sold), the same that MOS team took on when PPC died on desktop. They chose to support obscure and expensive HW.
FFS, even when they could have chosen more viable PPC CPUs for their HW projects, they either stuck to unobtanium(PA-6T) or shit like e500v2. If I didn't know better, I'd argue that Hermans, Dickinsons & Co probably used Amiga stuff for tax deduction or something like that. Completely devoid of any sense.

_________________

@Karlos

What has that to do with who owns the code or even has access to it.
It all being PPC sure won't help porting it to ARM or back to 68k.


Right now Hyperion/AEON/??? seem to have the right to distribute those PPC binaries as part of OS4.1 or at least none of the owners feel the need to make a legal fuss about it.

Port it elsewhere and you cut out AEON out of the equation who might not be so happy bout that and do seem to have legal ground to fight.
You also need to reacquire or rewrite everything they don't have the sources for.

-> messy situation which noone wants to trigger

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@WolfToTheMoon

The one that made the incompetent SoC choices is not only the one that would stand to loose the most on Mac port (or anything else) but also the one who has (or seems to have) the most reasons to bring Hyperion down.

But what he does not have is a legal way to make an Amiga or Amigalike OS without Hyperions consent.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada

@Kronos

Quote:

It all being PPC sure won't help porting it to ARM or back to 68k.

No, no. The goal is over there, track 45 degrees left. Question: How do you think all these old components became PPC native? Answer: by being re-implemented in a more up to date C standard. That is the actual hard work and in order to get to a PPC native distribution it was done already.

The issues around who owns what are an obstacle that has no engineering solution but "how do we get from PPC to X" is mostly solved already.

_________________
Doing stupid things for fun...

@Karlos

So?

If you don't have those reworked sources or if using them in that way would trigger another lawsuit they might as well not exist.

Sure you can fantasize on how such issues would evaporate but in reality everybody who holds any strings in Amiga(TM)land is either a megalomanic, scam artist or moron (or a combination of those) AND they all have axes to grind with each other.

Hence, neither AmigaOS4..1 nor 3.2 being a suitable starting point for Zombie68k-OS to have it's "MorphOS2.0 moment" which was less about the features added but more about a capable team taking full control with a somewhat coherent plan.

_________________
- We don't need good ideas, we haven't run out on bad ones yet
- blame Canada