The user acepeg who was present at the event, contributed with more information in the "MorphOS @ Alchimie 2013" thread at MorphZone.
Some clarifications/added info from the thread at MorphZone:
Fab: "Erm, about shaders, i didn't exactly say that... I said the driver infrastructure had been reworked to make it easier to deal with shaders (in fact i think the currently released r300 driver even makes use of this already to implement some features, but i may be wrong on that) . But i didn't say that full shaders support would come anytime soon. It's really a lot of work.
As for the future architecture, i mostly insisted on the fact it was still in thinking phase. One shouldn't hold his breath for this, this is a looong and hard way to go.
As for SAM460, it was a proof of concept. I don't know if it's being worked on or not. I don't have information about this."pega-1: "To make a long story short, we obviously evaluate future options for MorphOS alongside our current developments as well. This doesn't mean that the current path is about to die anytime soon."@TheMagicM
Ok, so that means that....
There will be a new 64 bit version, but not one that runs on PowerMacs, right? It will be for a different architecture?
They will continue developing MorphOS for currently supported PPC machines, and not only that, they will also increase the number of supported G5 systems. The latest roadmap has outlined at least 3 (probably 4 with a "MorphOS 3.7 bugfix release" not mentioned but highly probable). If they come up with new stuff, I'm sure they will add more releases after that. So don't worry!
But you have to think about
the reason to why we are using PPC at all. I can see these reasons,
- We (just like Apple) outgrew the 680x0 family, we needed more "juice"
- At that point (in the 90's), Apple formed an alliance with IBM and Motorola (later Freescale) to develop a new, suitable CPU family for future Mac computers, the PowerPC. These three companies were giants in the industry, and this alliance surely looked bright for the future, the PPC was surely to be a safe bet
- The PowerPC was Big Endian, just like the 68k. This opened up for the interesting and very appealing concept of making an Amiga OS who could run all applications and data truly *native*, just like before, without the need to box in the environment (like UAE for example). Just add a translator that "recompiles" the 68k binary streams into native PPC binaries, and you are all set, everything runs just like before, native down to the silicon!
This is what MorphOS set out to do; migrate the Amiga environment to a more powerful HW architecture with a bright future, that would handle the Amiga legacy Big Endian SW/data in the same way as before, in a Big Endian environment, while opening up for further improvements and development. This became MorphOS.
But then things changed. In essence, betting on the PPC horse turned out to be a bad bet. Eight years ago, even one of the funders of the alliance (and it's biggest/only important customer) decided to jump ship. This effectively killed the PPC as a desktop/laptop alternative.
Today we aren't using PPC and these old Mac's because they are the best. They aren't. Compared to other 2013 HW, these PPC machines aren't even good at all. Perhaps "good enough for us" (and it is, at least I am very happy with my Mac Mini), but that's a statement that slides a bit for every year that passes.
The fact is that yesteryear's Cortex-A9 ARM CPU's used in *smartphones*, was well into "G4" territory, raw CPU performance wise. Current Cortex-A15 ARM CPU's are in "G5" territory performance wise as we speak. And these are *cheap SoC chips*, made for battery powered smartphones!
In 2014 the new 64-bit ARMv8 devices will roll out, based more or less on ARM's Cortex-A53/Cortex-A57 chips. These will run our PPC machines into the ground, and they will *still* be highly integrated SoC's with low power consumption, low heat, etc, suitable for battery powered handhelds. But they will be used in other areas as well, like server environments, probably desktops, etc.
AMD is for example entering the ARM market with their "Hierofalcon" CPU:
- Up to eight 64-bit ARM Cortex-A57 cores
- Up to 2GHz.
- Dual 64-bit memory channels DDR3/DDR4
- 10-gigabit Ethernet
- PCI Express 3
This is definitely more of a desktop/server CPU than a smartphone one. And AMD will hardly be the only one, more will come.
And when it comes to the 4th generation Intel Core i3 / Core i5 / Core i7, then we are talking performance
several dimensions away from "our" PPC machines.
So we are not using PPC because it's the best, or even good, because it isn't, not by today's standards. We are using it because we are tied into it by previous decisions, and the ambition to run the legacy Amiga environment fully native. Which is, and has been during all these years, a very nice thing. After all, MorphOS is (unlike OS4) about *using*, not *waiting* for whatever might be around the corner!
But as PPC is floating away down the streams of computer history, we have a choice to make. Shall we let MorphOS float away together with the PPC and walk away and forget about it? Or maybe we jump into the stream ourselves and floats into history with it and become
retro users (IMHO the idea behind MorphOS was never about "retro", but taking the Amiga to a modern level). Or shall we start thinking about a future for MorphOS on a different architecture?
If you choose the last one, it will break the one of the core original ideas with MorphOS, that the Amiga legacy Big Endian SW/data should be able to exist native, because both ARM and x86 (whatever they may choose) is Little Endian. This is a big thing, a game changer. The new MorphOS becomes Little Endian. You can run Amiga legacy SW/data on Little Endian machines, but you must use something like UAE (which can be implemented nicely enough, look at AROS for example).
Of course, when you have made this decision you can as well look at other features that -by design and concept- is *incompatible* to Amiga, like true SMP, Memory Protection, 64-bit, etc. Suddenly it becomes theoretically possible to create a modern, multithreaded OS that can be compiled for either (or both) of the big two CPU architectures, meaning ARM and/or x86.
The result will be a modern Amiga^H^H^H^H^H MorphOS (slightly modified/extended API) Little Endian OS, capable of running Little Endian machines, with features previously not possible on Amiga. To an end-user, it could very well look and feel exactly as it does today, recent applications could be adapted/recompiled for it, legacy Amiga applications handled by some kind of UAE layer, more or less transparent. And the Big Endian PPC machines floats away in the history stream, maybe together with a few newly become "MorphOS Retro fans".
This isn't something new, it was presented at "
Alchimie 111111" already. If this scares you, think of the alternatives, the choices you can make. Future, or no future? And under any circumstances, this won't happen overnight. Read again the quotes I made above from Fab and Pega-1. Current MorphOS development on PPC continues as before, for a long time still. The "NG" stuff is made in parallel. And as something as the old C64 users, and later the "Amiga Classic" users used to say: "MorphOS PPC is not dead until the last Mac Mini Power LED goes black".
Personally, I'm relieved that they are actually thinking about a "MorphOS NG", a future for this great OS. I need and want something that is not Windows/Mac/Linux, on modern HW!
