@matthey

I choose PowerPC.Because it was the excitement of playing my first 3D FPS games on the Amiga 1200,Amiga 4000 or Amiga One.Like Heretic2,Hexen,Wipeout 2097 or Quake 2 (on my X5000) or Payback or Quake 3 and playing ScummVM at reliable fast speed.

I can choose better but I love PowerPC.

_________________
AMIGA 500 1.5 MB ACA500/ACA1232 accelerators
AMIGA 500 German
CD32 unexpanded
Amiga 1200 Tower on AmigaKit since years
AMIGA 1300 030/50 Mhz/32 MB WB 3.9 with lots of games&demos
AMIGA ONE XE G3 PPC 800 Mhz/1 GB RAM/RADEON 9250 128 MB/SATA CONTROLLER

@pgf_666

Quote:

Secondly (and counter to the same ageist argument), no one develops or markets a product based on the uniqueness of the CPU or its specs.

Yes they do, they do all the time.

Raspberry Pi and similar boards put their ARM SoC front and centre. Same goes for Arduino and Teensy boards.

Beyond the every day hacker and hobbyist, bespoke industrial systems and mainframes use CPU architecture uniqueness as a selling proposition. Console producers (Sony, Microsoft, Nintendo) care very much about the uniqueness of the CPU, they even work with suppliers such as AMD to create custom parts, and then they communicate that value to the game developers. A factory buying robotic assembly machinery care very much about the uniqueness of the CPU. Vendors selling to military and aerospace also focus on the uniqueness of their Mil-Spec and radiation hardened CPUs. Sending a probe to Mars? JPL will specify the CPU that supports their remote programability requirements.

Each of these market segments are extremely lucrative for CPU makers. It's very ignorant to base a world view of CPUs and microarchitectures on mainstream Microsoft Windows, Apple macOS, and Linux computing. There is no DirectX for factory automation, there is no Unreal Engine for for self-driving cars, and there is no OpenCL for independent targeting fire-control systems.

_________________
All the way, with 68k

https://www.pcgamer.com/analyst-predicts-20m-people-will-abandon-gaming-pcs-for-checks-notes-tv-gaming-platforms

@agami

Quote:

Raspberry Pi and similar boards put their ARM SoC front and centre. Same goes for Arduino and Teensy boards.

No, they don't. Go to the raspberry org's home page, and it's all about the things you can do with a Pi, and the information resources which are available. In fact, there is not a single mention of processors, or ram, or flash, or ethernet, or wifi.

On the products page, there is one mention of a "quad core processor", and zero mentions of the actual instruction set. It requires clicking on "More Info" for the Pi3b, and then selecting the "Specifications" tab to finally find mention of "ARMv8".

All of which confirms pgf_666's point --- people don't care about what's inside, people care what they can do with it.

And the thing is --- these days, that means "what can I actually do with a piece of computing gear", not "what can this gear theoretically do, based on its specs". Which is why there are many, MANY boards with better specs-per-dollar than the Pi3, and yet the Pi3 dominates its space --- it has infrastructure that goes with it; It has easy availability, it has well-standardised hardware extensions, it has millions of people using it (and thus when something goes wrong, the answer is often just a quick Google search away), and it has decent linux support.

Of course, there is a lesson to be learned here for "Amiga". Or for anyone trying to establish something "alternative" on the desktop... It's not about what's under the hood, it's about allowing people to do something that they can't do with the "established" or "incumbent" stuff. 35 years ago, that might have been the GUI. 30 years ago, having many colours on screen qualified, as well as actually running multiple things at the same time. 25 years ago, being able to network machines, and to have uptimes measured in months under heavy use. 15 years ago, being able to use more than two to four gigabytes of memory started becoming important. About ten years ago, "normal" laptops became powerful enough to be used for everything for most people; Also, meaningful multi-monitor support became important. Five years ago VR was what a GUI was 30 years earlier.

I am about as big a tech nerd as you will fine anywhere, and was running a BBS on linux back in 1992; Yet today at home, when it comes to actual computers, I have a bunch of Mac laptops and a Franken-Windowsbox (purely for VR).

@umisef
True.

The Pi is so cheap, so cheap that lots of people even buy one just in case. It's also powerful enough, has been on the market for a long time, is fast enough (a magnitude faster than any classic Amiga, even with PPC). And yes, people don't care for what's inside.

So I wonder why so many resources are being spent on building expensive hardware instead of porting (rewriting) OS4 to work on Pi. Also, it's ARM today, it may be MIPS in 15 years. The OS should be ready for that.

Last edited by Leo on 28-Apr-2019 at 06:22 PM.

_________________
http://www.warpdesign.fr/

@Leo

Quote:

So I wonder why so many resources are being spent on building expensive hardware instead of porting (rewriting) OS4 to work on Pi.

People give the Friedens / A-EON a hard time for diverting resources to support the SPE FPU implementation in the Tabor. I would think that changing CPU ISA is much more of a massive waste and diversion of precious resources at this stage in the AmigaOne project! No massive increase in users has occurred yet from the X1000 / X5000 machines but if the Tabor causes a massive boost it does make the possibilities for other future ISA roadmaps become feasible. At present only PPC makes sense (especially given the poor financial health of Hyperion) and a relatively low cost motherboard like the Tabor is a good idea.

I don't own a Rasberry Pi because I'd prefer to learn BlitzBasic on a Classic Amiga than mess around with RiscOS on a Pi to be honest. I'm guessing that many ex-Amiga owners would feel the same way and that the Tabor WILL appeal. There IS a market for the Tabor and RiscOS beat AmigaOS to the RaspberryPi and we can't put that particular genie back in the bottle.

Last edited by BigD on 28-Apr-2019 at 09:38 PM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

Quote:

Leo wrote: So I wonder why so many resources are being spent on building expensive hardware instead of porting (rewriting) OS4 to work on Pi. Also, it's ARM today, it may be MIPS in 15 years. The OS should be ready for that.

I believe it is because they want the system to be unique, not just another PC. It is intentional.

We may not agree, but for me it does not matter as there is AROS which does what I want. AROS is open source and portable. Among other things, work is being done to make it run on RPi.

@BigD

I have been hearing since maybe 15 years that changing ISA would be too long, too difficult etc... So the sooner it starts the sooner it's done. Since the market is now mostly focused on ARM and x64, PowerPC will get less and less affordable and less and less adapted to traditional desktop usage of the Amiga.

If money was spent on software instead of producing a power pc board this could also help. Companies already spend millions designing and producing ARM boards like the pi: why reinventing the wheel? Replacing arm or x64 processor with a power pc for the Amiga only make things slower, more expensive, surely not better.

_________________
http://www.warpdesign.fr/

@Leo

The hardware gives a profit margin possibility to A-EON that can be used for software development.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

Long time lurker here. I would like to see a new asic 68k cpu with 3ghz that I can stick into my Amiga and OS4 ported to that.

@umisef

Oh wow! Out of all the points I made you took the mature 3rd generation of Raspberry Pi and their current de-emphasising the ARM architecture, and because of that one point therefore @pgf_666's point is correct?

He said:
Quote:

no one develops or markets a product based on the uniqueness of the CPU or its specs

No one. Really?

If he had said that most mainstream computing products aimed at consumers do not use the uniqueness of the CPU as a selling point, then I wouldn't;t have said anything.
But he said "no one".

What's more. Doesn't matter how much you may want to think of A-EON hardware as aimed at hobbyists, it is not in the same category as other hobby boards.

The point I was making, is that in areas considered to be special applications of computing, the CPU architecture does matter.

_________________
All the way, with 68k

@BigD
Quote:

The hardware gives a profit margin possibility to A-EON that can be used for software development.

What margin? Trevor said several times in the past that he didn't see any return in investment.

"A return in my investment? Er, perhaps one day but in the meantime I do have the pleasure of traveling around the world meeting like-minded Amiga enthusiasts to share and spread my joint Amiga passion. Long may it continue."

see: http://www.generationamiga.com/2017/07/10/the-big-interview-with-trevor-dickinson/

_________________
http://www.warpdesign.fr/

"Special applications of computing"

Hmmmm I wonder what are those and how you define them.

@Leo

Quote:

What margin? Trevor said several times in the past that he didn't see any return in investment.

I clearly said "possibility" of a profit margin. How Trevor sees fit to manage his business / children's college fund is none of my business

The hardware business certainly does Apple no harm. A-EON with their clever marketing and swish cases COULD if they controlled their cost make a profit on the hardware that they produce however increasing the user numbers through cheap hardware is obviously being prioritised over the bottom line regarding the Tabor! Oh that every business could subsidise hardware costs / sell at cost to drive the software market (almost a console tactic except that console costs are recouped on hefty software licensing on EVERY SINGLE GAME)!

Last edited by BigD on 29-Apr-2019 at 02:45 PM.

_________________
"Art challenges technology. Technology inspires the art."
John Lasseter, Co-Founder of Pixar Animation Studios

@agami
Quote:

agami wrote: @umisef Oh wow! Out of all the points I made you took the mature 3rd generation of Raspberry Pi and their current de-emphasising the ARM architecture, and because of that one point therefore @pgf_666's point is correct?

Let's look at how the Raspberry Pi was marketed from the beginning shall we?
https://web.archive.org/web/20110506002027/http://www.raspberrypi.org/
https://web.archive.org/web/20120201113810/http://www.raspberrypi.org/
https://web.archive.org/web/20130115080030/http://www.raspberrypi.org/
https://web.archive.org/web/20160202004331/https://www.raspberrypi.org/

So the first link (pre-release) tells about a small cheap computer and lists it as using an ARM processor in the specifications. The second (post release) say "An ARM GNU/Linux box for $25" which is still IMO a specification rather than promoting an ARM platform. The third doesn't even mention ARM on the front page nor does the fourth capture.

@megol

But isn't this also misses the point? RPI WAS manufactured with that function in mind. The old Amiga wasn't and it was still able to operate under similar conditions. So you can say that ARM was consumed by the RPI in that specific combo and you can also say that it helped set it apart and differentiate it from both the public perspective and the actual hardware-software combo and it then became synonymous with the RPI (at least in my mind - not a RPI enthusiast yet still gathering enough info to get the gist of what it's about). By comparison, you can say that the 68k was unique in its ununiqueness and simply being able to fit in that system of new, custom chips. Isn't there more "closeness" by design of the ARM and the RPI? I mean, from the ARM not being an endgame type of processor. For all intents and purposes the 68k line was an endgame scenario. The migration happened because of consequences which opens up this discussion even further - how do you prepare for something like that? what is the future flexibility of that ARM-RPI connection? because the past has shown that the 68k-Amiga connection could be severed or at least sort-of contained. Maybe it's a movement in both directions at the same time...

@Srtest

No, not missing the point of correcting something which was the only reason I posted.
Remember that what was written originally:

Quote:

agami wrote: @pgf_666 Quote: Secondly (and counter to the same ageist argument), no one develops or markets a product based on the uniqueness of the CPU or its specs. Yes they do, they do all the time. Raspberry Pi and similar boards put their ARM SoC front and centre. Same goes for Arduino and Teensy boards.

This is unqualified bullshit. He didn't accept being rebuked with actual facts from someone that understand things and so posted:

Quote:

Oh wow! Out of all the points I made you took the mature 3rd generation of Raspberry Pi and their current de-emphasising the ARM architecture, and because of that one point therefore @pgf_666's point is correct?

And that I corrected by showing that there were _never_ any emphasis on the ARM architecture from before it was released until as a mature product. The thing that pushed Raspberry Pi wasn't the ARM architecture, what pushed the Arduino wasn't the AVR architecture. What did was price, ease of use, and availability of open source software in both cases.

@megol

Quote:

The thing that pushed Raspberry Pi wasn't the ARM architecture, what pushed the Arduino wasn't the AVR architecture. What did was price, ease of use, and availability of open source software in both cases.

And clearly the Amiga met a measure of success largely because its numbers were dictated by its cost-effectiveness. The stalwarts will always maintain that technical specs were behind purchasing decisions but that's both inaccurate and self-serving.

The fact that we still have members of this community maintaining that they would never buy a (real) next generation Amiga unless it were based on some esoteric or long-abandoned CPU architecture is beyond cringeworthy. Sorry folks, you're not a viable market.

The Amiga needs to be reinvented.

I shouldn't lose my time posting here, getting tired of everything Amiga... But just to note that there's a huge opportunity being missed right now due to a couple of things that happened simultaneously:
1- The higher concern for privacy, specially given what big tech has been doing with user data
2- Windows 7 users (only recently win 7 was surpassed by win10 in userbase!!!) not wanting win10 and the "spyware" crap M$ is forcing on users nowadays. This is real, there are major youtube channels with huge viewer base talking about it (i.e. Linus Tech Tips ...)
3- Power9 being actually competitive on a per watt basis with x86, specially after the fixes for Spectre and Meltdown are applied.

If we had managed to get SMP with backwards compatibility (dreaming...) and half good memory protection I think we could have a very good chance of getting new users.

Forget the hardware, Trevor should have invested in a port to Power 9 and split the revenue with Hyperion.
I've said this regarding a Power9 port a couple years back in a Q&A at Amiwest and was almost laughed at, not laughing now sadly....

Last edited by Jose on 03-May-2019 at 10:31 PM.

_________________

José

@Jose

Quote:

Forget the hardware, Trevor should have invested in a port to Power 9 and split the revenue with Hyperion.

I first saw the Cyrus prototype boards 2 years before IBM announced Power 9.