@pavlor

Quote:

pavlor wrote: @cdimauro I hope you understand new PowerPC notebook is more probable than ALL needed improvements I mentioned. Some are trivial (RAM, GFX RAM, network), some not (compositing, 3D, FPU).

Yes. Nevertheless, they have a >0 probability, whereas a PowerPC notebook has 0.

@cdimauro

Quote:

Yes. Nevertheless, they have a >0 probability, whereas a PowerPC notebook has 0.

As usual, I don´t share your opinion.

@agami

Quote:

Nice try @bison. I like my feet without bullet wounds in them.

The reason I asked is that I suspect the situation would look quite different if we had to put it in writing (all of us, not just you). We would be dealing with facts instead of sentimental feelings of what was and of what may have been.

My list is rather short:

1. It has to be small. All modern operating systems are too big and complicated. Not just a little bit, but by a couple orders of magnitude. I am willing to give up some performance and capabilities for a better understanding of how the entire system works.

2. It has to be open source. I can't fix things -- or even understand them -- without the source code.

There is no number 3, since everything else is of significantly less importance. If an OS is open source and small, I can fix anything else that I don't like.

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

@pavlor

Quote:

pavlor wrote: @cdimauro Quote: Yes. Nevertheless, they have a >0 probability, whereas a PowerPC notebook has 0. As usual, I don´t share your opinion.

I know, but I invite you to think about the current situation of the post-Amiga market. It takes years to design, build and sell machines, which are VERY expensive. They do in small quantities, to be sure to have sufficient orders to cover the costs.

Now think at the effort that is required for a PowerPC laptop. How long it will take to design and build it? How much it'll cost? Are you sure that there'll be enough customers to cover the costs?
Last but not least, which hardware will be used to build it? You know that the current PowerPC SoCs do not integrate a GPU, because they were designed to be used for embedded and, specifically, networking, were a GPU is not required. So, an external chip (and memory) should be integrated to provide such features. It means that the board layout will be much more complicated, the costs grow, and power consumption too.
And what CPU will be used? An underpowered one?

So, yes, probability is not 0, but I think that a project like this will be quite not realistic.

@cdimauro

Notebook? That's just so 90's...

_________________
Dammy

@cdimauro

Quote:

Last but not least, which hardware will be used to build it?

Some of Freescale SoCs (eg. T1042) + Audio chip + MXM GFX card + whatever you need.

Sure, it would be expensive, but there were other designs even with integrated GFX (eg. uA1 or SAM440ep) - and their price was not prohibitive.

@damocles

Quote:

damocles wrote: @cdimauro Notebook? That's just so 90's...

@damocles: I've twos, but I prefer my desktop. Yes, I'm a bit 80's...

Last edited by cdimauro on 16-May-2015 at 07:47 PM.

@pavlor

Quote:

pavlor wrote: @cdimauro Quote: Last but not least, which hardware will be used to build it? Some of Freescale SoCs (eg. T1042) + Audio chip + MXM GFX card + whatever you need.

As you can see, you need a much more complicated motherboard.

The T1042 has 5 ethernets which are totally useless (Wi-Fi is the key for a mobile device, and you have to add it to the BoM) and the maximum clock is 1,5Ghz (so performances will not that good). Last but not least, it's a quad core, and no post-Amiga o.s. uses more than one core...
Quote:

Sure, it would be expensive, but there were other designs even with integrated GFX (eg. uA1 or SAM440ep) - and their price was not prohibitive.

They are very low-end. I don't think that they are useful for a laptop.

Anyway, dreaming is cheap.

@cdimauro

Quote:

and the maximum clock is 1,5Ghz (so performances will not that good).

1.4 GHz. Current SAM range uses 1.1 GHz, so it wouldn´be that bad.

Quote:

Last but not least, it's a quad core, and no post-Amiga o.s. uses more than one core...

With 10 W TDP, I don´t think this would be problem...

Quote:

They are very low-end. I don't think that they are useful for a laptop.

Radeon M9 (as used in SAM440ep) was until recently nearly best GFX chip supported by OS4... In this context anything better is bonus.

Quote:

Anyway, dreaming is cheap.

I see.

Last edited by pavlor on 16-May-2015 at 07:59 PM.

@pavlor

Quote:

pavlor wrote: @cdimauro Quote: and the maximum clock is 1,5Ghz (so performances will not that good). 1.4 GHz. Current SAM range uses 1.1 GHz, so it wouldn´be that bad.

If you think that it's enough, then fine.
Quote:

Quote: Last but not least, it's a quad core, and no post-Amiga o.s. uses more than one core... With 10 W TDP, I don´t think this would be problem...

That's for the CPU. Then you have to add the chipset, memory, GPU, audio, Wi-Fi. The consumption will increase for sure.

This will hurt autonomy for sure, but if you can live with a few hours, it's fine too.
Quote:

Quote: They are very low-end. I don't think that they are useful for a laptop. Radeon M9 (as used in SAM440ep) was until recently nearly best GFX chip supported by OS4... In this context anything better is bonus.

And supported.

@cdimauro

Quote:

If you think that it's enough, then fine

Fine.

Quote:

That's for the CPU

And memory controller and SATA and USB. Power requirements of other components (except GFX and probably also WiFi) are negligible.

Quote:

but if you can live with a few hours, it's fine too.

Anything above 2 hours is OK for me (note: I estimate it could be more than 5 hours in such design depending on battery).

Quote:

And supported.

Radeon HD drivers are now mature enough, that wouldn´t be problem.

@cdimauro

Quote:

@damocles: I've twos, but I prefer my desktop. Yes, I'm a bit 80's...

I rarely ever use my laptop, it's either my mobile device or the desktop. Only thing desktops have going for them is high horse power gaming that you won't find on cookie cutter console games.

But back on topic, to bring a OS back to mainstream will require it being on cheap mobile devices and that means a three finger salute to OS4 code.

_________________
Dammy

@damocles

Quote:

But back on topic, to bring a OS back to mainstream will require it being on cheap mobile devices and that means a three finger salute to OS4 code.

A fun alternative OS for mobiles is already available in the form of Sailfish on the Jolla. There is no sense developing AmigaOS for mobiles now. In the early to mid 2000s, yes that would have made sense. Now AmigaOS has to be ported to a reasonably priced laptop and be SMT enabled/have Gallium support on the launch of the X5000.

Mainstream in my view will have to be limited to one niche area of computing and with the Xena functionality in the X1000 and presumably the X5000 it would make sense to promote it as a control system for the 'internet of things'. That's the only 'new' market that I see the AmigaOne machines being able to compete in and have an inherent advantage in. It's time for A-EON to justify the inclusion of the Xena chip on the X1000/X5000 boards and prove the platform can benefit from having a IO controller chip built in! Other than that the only other selling points of AmigaOS are nostalgia and the built-in compatibility with 68k Amiga software and games.

Most probably none of the above is enough for AmigaOS to truly become mainstream again. However, a sustainable niche business could develop, big enough to sustain A-EON, ACube, Hyperion and AmigaKit? I don't see why this couldn't work out.

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

@damocles: I use my desktop or my smarthphone. Very rarely my laptops.

@BigD: Internet of Things means very small devices, which are able to connect/communicate with the outside world (and other devices also).

So, it's something very different from an X1000 or X5000 with its Xena.

@cdimauro

Quote:

So, it's something very different from an X1000 or X5000 with its Xena.

If the X1000 could be the hub and I'd prefer that than all this guff about using your smart phone to turn on your central heating. An X1000 running a surveillance system or sprinkle system or climate control for a greenhouse or in charge of streaming video to the fridge or messaging your smart phone when the fridge has run low on a commodity etc. Security wise I'm sure the X1000 would make a lot of sense. How many people would know how to hack an AmigaOne?

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

@BigD

Quote:

A fun alternative OS for mobiles is already available in the form of Sailfish on the Jolla. There is no sense developing AmigaOS for mobiles now. In the early to mid 2000s, yes that would have made sense.

There are a ton of alternate OSs out there already, why bother developing OS4 any further? Using your logic of course.

Quote:

Now AmigaOS has to be ported to a reasonably priced laptop and be SMT enabled/have Gallium support on the launch of the X5000.

Why does it has to? Obviously it's never going to be main stream in it's current design. What `$900 modern laptop are you talking about porting OS4 to? MorphOS already runs on old Apple PPC laptops and will be running on A1X5K, what is there for OS4 to achieve that hasn't already been achieved by MorphOS? Again, using your logic.

Quote:

Mainstream in my view will have to be limited to one niche area

One tiny niche area is mainstream? I don't think so.

_________________
Dammy

@damocles: I fully agree.

@BigD: Amiga o.s. is totally insecure, so it's out of question for such things like IoT, where privacy and security are key points.

There are also A LOT of cheaper devices which can do the job much better than any AmigaOne + Xena.

Take a look at the $9 computer, for example, but also at the Raspberry 1 & 2, Intel Galileo and Edison, Minnowboard, etc. etc.. They range from veeery cheap to very powerful devices (Edison and Minnowboards integrate a BayTrail x86 processor!), at less than $100.

@damocles

Quote:

One tiny niche area is mainstream? I don't think so.

The AmigaOne machines have to do one niche thing well. Just as the 68k Amiga was great at graphics (video editing, ray-tracing, bitmap animation, etc) so the AmigaOne machines have to do well in a certain area and carve out a niche. That should be the goal and mainstream success may or may not come out of that. The truth is the world doesn't need another mainstream desktop/laptop platform, it doesn't need another mobile OS and it definitely doesn't need another cheap tablet. The market is crowded and valuable AmigaOS development resources shouldn't be squandered on a slim outside chance of scrapping a tiny fraction of 1/100th of a % market share in a market where Hyperion can't hope to keep pace with Android and iOS.

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

@BigD: the world also doesn't need a $/€ 3000 (or more) machine. Especially talking about mainstream market.

@OlafS25

Quote:

OlafS25 wrote: @cdimauro I made the Lame test and outperformed all SAM models http://www.aros-platform.de/html/benchmarks.html

Sorry, I forgot to reply to you.

Honestly, I think that such page isn't useful until you add more detailed information about the used hardware and software. Please, if you can, update it.

Last edited by cdimauro on 16-May-2015 at 10:17 PM.
Last edited by cdimauro on 16-May-2015 at 10:17 PM.