With its memory, processor, and GPU limitations, I still favor an A15 over the RPi.

@amigadave

I agree with most of your points although driver support would be essential.

No 3D/video acceleration = fail.

I would prefer a port to x86-64bit hardware as it's better documented, easier to find, easier to replace by a faster machine and less limited. 256MB was very bad and 512MB is not so bad but I prefer having 8GB and being able to connect a pair of 2TB drives instead of using 512MB of ram, a sd card as hd, a slow ethernet port and not so fast usb2.0. It's fine as xbmc box but I would never use it as a desktop without proper driver support.

Come on, it even struggles playing videos using cpu that an old peg2 plays perfectly. If you want it to use GPU you have to buy linux binaries not easily useable from other operating systems.

Jim,
Quote:

Raspberry is the product of a charity, it's meant to be functional and cheap. That said time has moved on since it was started and you can now get Mac mini type power for the same price.

Type "iPEGTOP" into ebay ...No, I'm not joking.

You'll get a load of cheap USB computers with specs quite significantly higher than the RPi
Dual core Cortex A9, 1.6GHz. 1GB DDR3 Ram 8GB flash, USB, HDMI, WiFi etc...
They're quite limited but at that price who cares!

No A15s down there but once the Chinese start making them...


As for the whole x86 Vs ARM debate for MophOS I can kind see the benefits of both sides.
The closed platform and raw processing power of the Macs makes them a very good target but good Macs ain't cheap.

ARMs don't have the same level of processing power but see above.

If anything they should be cross platform, Macs for the high end, cloud sticks for the "mass market" low end.

As for the problem of every ARM from different companies being a different architecture isn't quite what it once was.
That's true at the high end where companies can afford to add their own components.

At the low end it's completely different. The low end guys compete on price so they just buy in parts and put them together. Well, it's actually nowhere near that simple but the point is they are not developing their own modules, they use standard parts most of which probably come from ARM.

As for how you would support 2 platforms well there's an answer to that but I suspect it'll be rather unpopular.
Use a Linux Kernel.
You get all the drivers, protected memory and multiprocessor support built in.
You customise the kernel for low latency, cut out all the legacy guff and build the MorphOS environment on top.

Won't this just make MorphOS into a Linux distro?
No, Linux distros are a hodge podge of open source components on top of one another.
Android uses a Linux kernel but it's nothing remotely like any Linux distro.

I suspect this will be an unpopular opinion but remember the kernel is only one component and it's completely invisible to the user. I suspect it'd speed up development quite significantly.

<runs away>

>"iPEGTOP"

Interesting, but the listing mentions a vague 1.2-1.6 GHz operating speed.
Still, slightly more then $50?
Pretty cool.

Crumb wrote:,
Quote:


I thought that what I wrote equaled exactly what you stated again in your reply above, except I did not mention my preference for x86/x64 for more powerful solutions.

In my opinion, the only thing the Raspberry Pi has going for it is the low price, but more important is the number of units already sold and likely to be sold in the future. That huge community of users is what would be attractive, because assuming we could interest even 2% of the people we could show MorphOS3.x to on the Raspberry Pi, 2% of 20,000 to 40,000 Raspberry Pi users (I don't know how many have been sold so far, but it is a large number), is much higher than 30% of the users of a much less popular product. Also, given the popularity of the Raspberry Pi, and the amount of media coverage it receives when something new comes out that can run on it, or a new use that someone finds for it, there would be a greater chance for wide spread coverage of an announcement of a port of MorphOS3.x to the Raspberry Pi, than there would be to any small production run alternative, even if the alternative was twice as powerful at half the price.

All I am saying is that "IF" the MorphOS Dev. Team decides to port to "ANY" ARM device, I think it would be smart for them to start with the Raspberry Pi, but only if, as Crumb and myself have tried to point out, they could leverage the Raspberry Pi's GPU as well as the Linux kernel currently does.

No drivers & no 3D/Video acceleration = FAIL! Which is why this port will probably never happen.

> 20,000 to 40,000 Raspberry Pi users (I don't know how many have been
> sold so far, but it is a large number)

News item from 2.5 months ago:

http://www.raspberrypi.org/archives/3011

amigadave,
Quote:


This is so true!!

Start with the RPi where volumes are and get volumes out and coverage. Then move on to the 64bit highend ARM CortexA57 (coming in a year or 2) to attract RPimos users to modern powerful but light RISC based ARM , as a legacy to the existing PPC RISC architecture.

It makes sense. i wish any of the development team would give their view on this.

amigadave,
Quote:

This is a very good point. Big captive market and cheap publicity.

..but don't think in terms of 1 or 2% or any %. Just think here's a potential market and how do we get to them.

What can MorphOS offer RPi users that Linux can't?

minator,
Quote:
I agree... MOS lags behind the other modern OSs in many ways. MOS is not a reasonable choose against other OSs.

amigadave
Why do you want to make MOS more popular?
I think, We - who used to have an Amiga computer in the past - find it interesting/fun to have an Amiga reincarnation OS.
(OK, it has some advantages also over the others, but there are many drawbacks too.)
I think, the most important thing is to increase the count of quality softwares for MOS.
If there are many asskicking sw, there will be more users too.

But, to reach this, we also shoud do something... Something more usefull, than discussing about the 'unclear future' of MOS.
(A small tool, a translation, writing a document, backgrounds, icon sets, etc.)

MOS developers are talented guys and we can let them to decide the question of architecture change!
I'm sure, they will decide carefully.


I am happy with MOS, and don't care with the others (who don't use MOS).
This PowerBook is a nice computer, even if not the fastes one.
It is fun to use an OS like this.

Quote:

a FASTER experience, in many ways. The raspberry Pi is nice when used without a desktop GUI, otherwise, even with the lighter linux ones, it's very slow, and painful to use.
That would make an obvious difference.

To be honest I found the Efika to be a right load of nails, how does the Pi compare?

> I found the Efika to be a right load of nails, how does the Pi compare?

Regarding CPU performance:

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=8369&forum=11&start=8

it wouldn't perform MUCH better, but at least with 512Mb ram that would make a huge difference. I wonder about the sdcard (class 10) speed in a raspberry in comparison with the (slow) ATA port on the efika though. Oh and it has USB2, the efika was usb 1 if i'm correct.
The CPU power isn't all, everything should be taken into comparison.




[ Edited by SoundSquare 22.03.2013 - 14:51 ]

SoundSquare,
Quote:

I am NOT using Linux without a GUI.
THAT sounds painful.

Hasn't anyone else looked at minator's A9 device?
Looks pretty neat.
And it ought to run circles around the RPi.

@Jim

Never heard of Minator A9 , a quick search on google gives a hit on your comment (!)

Do you have a link?

> Never heard of Minator A9 [...] Do you have a link?

https://morph.zone/modules/newbb_plus/viewtopic.php?forum=11&topic_id=8369&start=52

http://www.ebay.com/itm/Dual-Core-Android-4-1-MK808-TV-BOX-Mini-PC-stick-Thumb-Rockchip-RK3066-HDMI-NEW-/321076632237?pt=US_Internet_Media_Streamers&hash=item4ac1a89aad

dekanyz WROTE:,
Quote:

The answer to your question is in your own reply. More users makes it more likely that we will attract more programmers, to write more software for MorphOS3.x.

It has always been debatable which needs to come first, more users to attract more programmers, or more programmers to write more software that will result in more users. Sort of the Chicken and Egg conundrum, as which must come first.

I believe that some of those new Raspberry Pi users will also be programmers, so if we can grow the number of users and expand the amount of exposure for MorphOS3.x, we will in effect, at the same time, increase the number of programmers interested in writing new software, or porting old software to MorphOS3.x. Not huge numbers, but in our small community, we have seen that even one or two new programmers can make a difference. Just look at the amount of great work that has been done by just one or two of our experienced programmers (like Fab, Piru, and many others) who have written parts of the MorphOS itself, or given us great third party software like OWB/Odyssey.

If we can get lucky and find 2 or 3 new great programmers from the increased exposure we could get by porting to the Raspberry Pi (or any other means), it could be a great asset to our community one or two years down the road, if those programmers take a liking to MorphOS3.x and decide to stick with us and become active developers in our community.

That is why I am always trying to think of new ways to increase the number of users and programmers in the MorphOS community.

I'm curious... Is there any real intention for MorphOS to go beyond PPC?

Re: x86 vs ARM At this point in time, the expandability of x86 (multiple memory, sata, pcie and other slots) makes up for its legacy architecture. If ARM motherboards had the same expandability, I'd be using one.

> Is there any real intention for MorphOS to go beyond PPC?

https://morph.zone/modules/newbb_plus/viewtopic.php?forum=3&topic_id=7930&start=54
https://morph.zone/modules/newbb_plus/viewtopic.php?forum=9&topic_id=8121&start=18

Jim,
Quote:

Well, it's not my device but even one of those cores should be 4x faster than the RPi.
(The A9 is about 2x faster than the ARM11 at the same clock).

However as AmigaDave points out the RPi has a big community and actually it's designed to get people into programming - that's why it was built in the first place.

It's a relatively low performance device so I don't expect anyone to use it as a main desktop, it's more of a hobby thing. lack of memory protection and SMP wont be a problem.

MorphOS could offer a friendlier front end than Linux and is likely to be rather less of a resource hog.

Getting drivers could be problematic but this is going to be the same on any platform. Vendors are loathed to give out any specs, usually with good reason.

Without all the legacy stuff that ppc MorphOS supports, how many existing users would make the switch? Or is that not a consideration for the devs?

'nother thing about the RPi... If its weird, binary blob graphics driver can be accommodated by MorphOS, couldn't similar linux blobs for NVidia chipsets also be accommodated that way? Wouldn't that make something like ouya.tv a better choice? Or just a better target as other NVidia based devices could then be supported...

amigadave,
Quote:
Agree...

Quote:

It is just a gambling: Putting a huge work to the porting to have the chance of joining some new talented guys.
And don't forget: MOS is not free (agains Linux, which is more advanced; AROS, which is also an Amiga like system).
A MOS like system only could be inviting, when there are many great apps for the most common areas of usual computer usage. Unfortunatelly, MOS has many white areas yet. So, I'd be far much more happy with new apps, than having a new plattform.

I think, the 'Trying to have someone who will do the work' is a bad approach.
What I'm saying is our community is quite small, but... the headcount is over 1000. If everybody pick up a small 'task' then there could be more small tools, etc. created. A (pro)active community could be also a very positive thing.

dekanyz wrote:,
Quote:

I think we agree on more things than we disagree about anything. I also don't think we need any more different platforms to be supported at this moment in time, and that more software is the most important thing, but this discussion was about the Raspberry Pi and how it might benefit the MorphOS community. There has been lots of discussion regarding porting MorphOS3.x to the ARM architecture, or should we instead focus on porting to the x86/x64 architecture. It sounds like you agree with me that more software is more important now than more ports to different hardware.

That being said, I am very happy to see that the port to G5 PowerMac's is happening, because I want the increased speed, and I have a dual 2.7GHz G5 PowerMac tower here waiting for the port to be completed.

Quote:

I think that MorphOS3.x ported to the Raspberry Pi could be interesting to a few computer programmers/enthusiasts that are owners of the Raspberry Pi, but only if such a port proved to run faster and better than the existing Linux ports currently running on the Raspberry Pi (plus the other requirements we have both already discussed in this thread).

Quote:

Not sure how you figured out that we have more than 1,000 users in our community, as my own guesstimate is closer to 750 to 850, with most registered users owning 2 to 3 registered/licensed MorphOS3.x compatible computers, and some of us own more than 3 registered/licensed MorphOS3.x computers, while some only have 1. You may be closer to being correct than I am with my guess, or we might both be way off, but it does not matter. What matters is what each of us does to help advance MorphOS3.x into the future.

I do like your thinking that the community would benefit if more of the users were more involved in doing what ever they could to help out with small programming tasks to fill the missing tools needed to expedite more software production in the future. I am trying to teach myself how to program, but admit that it will be years (if ever) before I consider myself productive enough at programming to be an asset to the community. Until then, I will continue to promote MorphOS3.x when ever I get the chance, and support the existing Team and third party programmers in any way I can.