I don't quite follow you here. In what way is anyone preventing anyone from writing drivers? Isn't MorphOS as modular as AmigaOS ever was, i.e. were you ever required to actually have access to Commodore's OS3.1 sources in order to write a driver for it?
That's a *different* matter, now you're talking about bringing the OS itself to another platform, and doing so yourself, which of course will require the blessing from its owners (since it's proprietary). But this isn't the same thing as making your own OS modules, or SW for mobile applications, STB software, or anything else that turns MorphOS into whatever you need.
It would open up for endless forks which would drain all momentum from the "real" MorphOS (you suggested 6 forks yourself on top of the "original"; POSIX and None-POSIX times PPC, Intel and ARM).

If anyone wants to turn an Amiga OS into some POSIX clone, bring it to a different CPU architecture or whatever, feel free to do so with AROS.

AROS has been open source all along. And gee, what a success story that has been (BTW, I don't mean to insult the great efforts of the great developers that has contributed to it, but despite being Open Source and despite being x86, it hasn't made any impact to the world whatsoever.)

MorphOS was never a free and open source OS, and it has come miles further in much less time.

There you have the answer!
Haiku is an open source reimplementation of BeOS, a proprietary, commercial and closed source OS that entered the scene after the Amiga in practice was "dead". BeOS generated a lot of Hype and momentum, of which Haiku inherited some after BeOS went belly up. Hype and momentum is required for an open source OS to even take off. Simply going open source will not accomplish that, as we have seen in the AROS example. Unlike BeOS/Haiku, the Amiga had been too dead for too long for anyone outside the Amiga community to become interested in the AROS project.

It will be interesting to see how the new Anubis will make it. AFAIK, they have two developers, and "Dammy" as a promotion guy.
Anyone can write software, anyone can write drivers, and by doing so anyone can make MorphOS more suitable for example mobile devices like the LimePC ones (if that is your vision of a commercial context for MorphOS) or Set Top Boxes, or netbooks, or whatever. But in order to bring the OS onto a new hardware platform, you will of course need to negotiate a deal, sign some kind of contract, and pay up some money. MorphOS is a commercial OS. A lot of money has been invested in it, the Intellectual Property is owned by someone, and that someone is interested to get a return somewhere down the road (which is within their full right). If you don't like this *fact*, go use AROS instead, the great Open Source alternative.
Where did I argue that few users/developers would be better than many?
Try to follow the discussion before commenting. It was a reply to the fact that no developer will use a LimePC kind of device for serious SW development, not to praise the Mac Mini as some ultimate solution (although it's cheap, readily available, and offers a general performance about 2x the Pegasos 2 G4).
This is would be sufficient for developers and the true enthusiasts/power users, which probably are the only ones who will use MorphOS in a traditional desktop context. The true business based upon MorphOS will need to come from elsewhere. Devices like the LimePC would have been ideal, and it (or other cheap and low power devices) would be a setting where MorphOS could shine through its leanness and high efficiency.

A commercial purpose is needed for MorphOS, someone needs to identify a customer demand where this product has a potential to fit, and then start adapting it towards that demand.

That's how there will be a future, that's how OS development pace could increase. Open Source is not the answer, as AROS has clearly shown us.

Well, it *usually* helps to have the source of one driver to work on another.
No, it's not the same, but there are MILLIONS PS3 owners, some working on more than games and many hacking their way around Linux that might just be interested in working even in helping with porting MorphOS.
Not forks, ports. A fork is a separate source tree, a port however, usually shares the same source tree as the original. Usually.
I haven't seen AROS yet, but I have seen MorphOS, hence my suggestion to open an existing OS.
Apart from being the basis of some MorphOS components? (I didn't know that, I read it here actually, don't mistake me for a MorphOS user, I have little experience with the OS itself, I just used a bit on the EFIKA and liked it. And I find it is a pity to purposefully limit the number of users due the narrowminded developers.)
It could go even further. That's my point.
Haiku wasn't like that, until only this year actually. It takes time, not just marketing. BeOS was considered dead for quite a while and it took many incarnations (all failed btw) before Haiku could be recognized. Btw, I don't remember BeOS being marketed as the "next" Amiga OS. And I actually owned it and used it for a long time.
readily available != ebay, sorry.
I'm sorry, I still fail to be convinced. As long as I cannot have a say/influence on the OS I use, I refuse to use it. I wouldn't mind the price, but only open-sourcing it can really make sure it will have a future. Pity because it looks great and is very efficient.

Btw, I'm not a GPL/FSF advocate, any open-source license would do, even some kind of shared-source scheme would be better than the current situation, imho, at least.

In case of CGX the problem is that there's no ddk available. Except trial and error there's no way to develop a driver for CGX. It may work, but only with much luck and work. I don't know why there's no ddk. Closed source in general isn't a bad thing IMHO, but you need a good ddk. But I think it is available for poseidon and AHI at least.

I'm sorry but this is unadulterated crap, jcmarcos.

The Pegasos and Efika have been available from the Genesi store forever.
What?

_________________
Matt Sealey

It seems this statement from my part has made you very angry. Perhaps you think that I'm blaming Genesi for not providing hardware for MorphOS. Of course I'm not that stupid to imply that.
Genesi's role is not providing MorphOS with hardware. Although it has done before. In fact, MorphOS would be nowhere without bPlan's Pegasos and Efika. And we all know that.
We know that too. I'm just saying (and many might agree) that, nowadays, finding hardware for MorphOS is not easy. Nor for any of those PowerPC Linux distributions that Genesi sponsors. At least compared to other non-mainstream operating systems, that can be dropped into one of those "regular" computers at no cost.
No one is blaming nobody. MorphOS is exotic, one has to live with that.
I was just recalling history. More precisely, from 2003 and before. I don't think I'm too wrong here. Anyway, I'm not being very useful lately. But I really thought I wasn't saying anything silly.

Nevermind.

@jcmarcos

In terms of the EFIKA, we did provide hardware to eleven core MorphOS developers, including Ralph and Frank. We also provided complete systems or boards to a number of application developers, such as Grzegorz Kraszewski (Krashan) and Guido Mersmann. Grzegorz's EFIKA Project #163 was singled out by the original 5200B Product Manager at Freescale as being particularly noteworthy. Guido's Project #341 also helped create tremendous interest in MorphOS. Both Grzegorz and Guido have either hosted or been the star attraction in a number of PowerDev Meetings (for example, PowerDev Meeting #5) and have websites that draw further interest to 'the cause' (geit.de). We do provide a small amount of financial support for these events and we appreciate the efforts of the organizers very much. In total, through the EFIKA Projects Program more than 300 boards were provided to individuals that submitted projects. Some of these projects did not necessarily involve core development. One interesting project is Ulrich Becker's Project #100, a low cost, low energy digital picture frame. All these folks and *many* others contribute to the continued interest in MorphOS and we do what we can to support them.

And, Juan Carlos, we appreciate your interest and involvement here. Don't mind Matt. His bark is much bigger than his bite, so please stick around. As we all look ahead, what we need to do is get new hardware with MorphOS support released at the same time. This would surely make things better for all involved. We are definitely not making 'regular computers' so we all need as much cooperation and support as we can draw from each other if we are ever to be commercially successful with MorphOS, or for that matter the Power Architecture/PowerPC.

R&B :-)

_________________
http://bbrv.blogspot.com

That's what I was talking about. Well, actually, I was recalling the original Pegasos days. But perhaps that should be considered another company, not Genesi.
Please, don't confuse me for someone valuable :D . I'm not. I just enjoy these adventures, and sometimes, say something interesting. I like Genesi a lot, and feel happy every time you reach some milestone.
The thing is, I actually like Matt's style a lot, particularly when he bites. I consider him a very good professional.
But then, I also happen to understand MorphOS Team's sense of humour. I'm kind of a lost cause...
I'm sure many people feel very happy when reading this phrase. Let's see if luck is on your side!
Right!

Of course not, even though we did provide hardware for MorphOS.

I think you have it backwards, Efika was only delayed for 6 months due to a chipset bug we didn't want to expose. In the meantime we still sold Pegasos right up to the limits imposed by RoHS on July 1st 2006 (Efika was launched November that year).

It is not Genesi who did not have MorphOS ready for another 2 years. It is actually Genesi who have had the Efika on sale for 2 years now, and it's very easily available - log in to the store, click an Efika, click Checkout, enter your credit card details etc. and you have one on the way.

In fact Andre and Emil got a new store and website up and running which makes it even easier.

We even had an Efika on sale for $99 for 6 months last year.
I am concerned that you thought we made any money (is that what you mean by "big bucks"?) out of GIVING AWAY systems like Pegasos, ODW (and later Efika) to people. Do you know how much those cost to make? What the tiny profits are on computer hardware these days?

You can count on one hand (see Bill's post :) the number of useful MorphOS projects, and you'd need another 10 people to hold up their hands to count the freeloaders who wanted an Efika for free and never do any work (I actually put down a rule that MorphOS developers should not be approved, but Bill vetoed it - of course it took 2 years for MorphOS to arrive, so the only work done was in silence and secret by internal developers - nobody could see the AltiVec optimizations in Reggae until this summer, not that it matters too much on Efika)

As for the Linux guys, it's more people that did good work, but still a rather disappointingly high number of freeloaders.

We're going to ditch the 5121e and Coldfire projects pages today because it's really hard to run a projects site around hardware that probably doesn't exist. It was bad enough trying to explain the 6 month delay in the Efika.

_________________
Matt Sealey

Don't be concerned, it's just that you misunderstood me, or I didn't explain myself: I meant the days when Genesi had enough money to give away computers to developers. That is, exactly what you say.
Actually, I agree with everything you've said! Well, not everything: With your opinion about MorphOS Team, I don't agree, but I do understand you. At least I hope you don't hate each other too much...
Understood, and anticipated, mostly regarding ColdFire. It's a pity, because all this anticipation from Genesi can be seen as failed projects, and that doesn't sound right.
That's the price for taking risks, after all. Taking risks requires braveness.

So no efika2 for the moment?

Will you take new efforts into creating an efika followup? Something like an efika1 motherboard but with integrated graphics (2D) and sata?

I still like the efika's low power usage. For now it makes an efficient file/database/linux server. I believe no other hardware can beat the efika there.(except maybe arm boards)

Only the addition/need for using an external agp/pci graphics card 'destroyed' the power usage. And in the end it did not have the correct drivers to fully use the graphicscards 3d power. But maybe i was expection to much on that. And for office usage it was overkill in power usage.

The thing with morphos is that it lack serious office software. I liked its speed, but it wont be the os for day to day use, but that is a personal thing.

So in my opinion an new efika board with an integrated 2d graphics card with (already existing) open source drivers will be good. Should give good 2d performance and low power usage.

The lack of speed of linux i blame on the low memory provided on the efika. With more memory linux could be useable even for office usage, even on the current efika. (although i cannot prove that, but slimming down linux gives speed advantages).

My personal wish is an low powerusage board featuring opengl es. Like the beagleboard. But i yet have to see drivers for supporting that. So in the end it better to have good 2d driver support then no 3d support at all.

My Kill-A-Watt says 5W total with a 9250.

How is that destroyed? It's barely double until I run a 3D app (then it hits 7W :)

The power supply is efficient, the graphics card ridiculously low power (it can't dissipate more than 8-10W on that passive heatsink) in all consideration.

I don't think 7W is overkill for office usage. There isn't a comparable solution on the market (AMD Geode etc.) which works this well.

We're just missing ~384MB of RAM to make it all the more useful. Running in 1/16th of the RAM of usual office desktops is a real strain on the office apps.

_________________
Matt Sealey

Could you solder the Radeon chip and extra RAM to a new, redesigned Efika? Wait, sure you've thinked about this years ago. If you haven't done so yet, why is it?
I remember when you had the intention of making further models, with XGI graphics chip (a failed deal), and FPGA (now that would be fun). This could effectively be the Efika2. Instead of basing it on the "challenging" MPC5121e, which is so problematic, you could have a safe bet. The only disadvantages would be a bit more power drain, and slow USB. The disk controller in the current Efika is starting to go fast right now, I believe.

@jcmarcos

We have discussed this topic enough...:-)

*If* we do produce another version of the EFIKA it will probably have an onboard graphic chip. It will have more RAM.

XGI: "a failed deal"

From the outside it may appear that way, but you know if one does not play, one cannot win. The XGI driver code included sources that were not and could not be properly licensed under GNU GPL. XGI assured us this would be fixed, but never devoted the resources or made a serious effort to correct the problem. We did do what we needed to do. You can use the XGI card with the EFIKA using MorphOS (thanks to Frank) and Linux. XGI is not an NVIDIA or ATI/AMD. They are a small company. This is a much more important accomplishment:

http://ati.amd.com/products/certified/genesi.html

This is what is possible when you work with solid partners.

R&B :)

_________________
http://bbrv.blogspot.com

Yes, and I reckon that I might have upset you all in the past days, as I've been insisting so much on things that can't be twisted anymore.
I want to express my gratitude for the wealth of information and details you've been giving here and there. This place is perhaps one of the most informative in the world about freescale's chips!
That's what I was thinking exactly. I believe an MPC5200 + ATi Radeon MiniITX board could be considered "product". And it has to be easier to build than the proposed Efika 2, based on the MPC5121e.
Understood.