The Market research for new PowerPC system didn't prove as fruitful as originally hoped, but it generated a lot of interest all around. This inspired Power2People to create a bounty to help the design and production of the board!

So, I'm glad to announce the "Open Source Hardware Bounty"! More details on the bounty can be found here (scroll down to the open bounties and check the design steps). Genesi fully supports this effort as it is close to the spirit of their successful Pegasos systems.

So what is so special about this project?

ALL GENERATED MATERIAL WILL BE OPEN SOURCED! That is, schematics (PDF/Protel), PCB design (Gerber files), Bill of materials, etc. The license chosen is one suited well for similar hardware projects, and close to the spirit of GPL, the TAPR Open Hardware License. This means, after the design is finished, ANYONE can use the design to build a product based on it and sell it with no extra requirements. Also, the good thing about the license chosen is that it ensures there can be no patent claims from one board producer to another.

Software bounties will also be started in association to this bounty.

I want you people to know that I'm looking at this project very seriously, to prove this I already have donated 1000EUR to the first bounty, and I don't intend to stop there.

Discussion relevant to the development of the board will take place in the PowerDeveloper forum, but feel free to ask anything related here or by mail.

Hmm... I can not do it with OWB, but I'll make donation (100EUR) this evening.

2Bounty managers:

We have open, but not assigned bounties (~550EUR). Is it possible to shift this money to 8610 bounty?

An impressive effort, really! What a pity that everyone here has ran out of money buying MorphOS licenses
Seriously, one has to be very brave to go for something like this in these days.
By the way, after a quick read of the desired specifications, I saw you are targetting a "flex" sized board. Why not attempting the smallest possible board? Or would that mean ending in another non-standard compromise like the original Efika (which would have been great with a soldered graphics chip and no PCI slot)?
I'm not (yet another) guy saying "you should make a laptop board", just something small (flat!) enough to make it fit in every imaginable space. If its form factor requires a desktop case, many people will loose interest. And most of all, it's a SOC design, it has to be small!

Quote:

If the project is successfull, it might be easy to convert it to a laptop sized board with similar specs. There is this idea, the initial cost is the same however. But we have not dismissed the possibility.

@ mobydick

Quote:

1. At the moment, this bounty has no relation to MorphOS whatsoever.

2. There is no guarantuee that all bounty steps will be completed. So, it is possible that early donations will pay for the concept design bounty (step 1) but the PCB design will not be worked on because not enough money could be collected to continue. (So, only 15% of the work might get done with no refunds possible for any donations that paid for completed work.)


If you are interested in an open source mainboard (which is full of closed, proprietary components - just so everyone knows and there are no misunderstandings), do not mind running Linux and accept the risk that there might be only a partial mainboard design in the end, then by all means contribute.

But if you are only interested in this because you somehow hope MorphOS will run on a final production unit, think again.

http://www.morphos-team.net/faq.html

Quote:

MorphOS users who are interested in MPC8610-based hardware are best advised to delay spending money until a) the hardware is actually in production and b) the MorphOS development team has officially announced its support.

Quote:

You might do that, but I think a better suggestion would be to create a bounty to port MorphOS to this board. That way, you also give more probabilities that your goal -to run MorphOS on a MPC8610- becomes reality :D

Or even better yet, you could support both bounties :D

Quote:

It doesn't but if people don't get actively involved it will not even get a chance to have a relation.

Quote:

Same as above.

Quote:

Most(all?) open source mainboards are full of closed proprietary components, even the Beagleboard is. That point really doesn't stand. AFAIK, we're very far away from a truly open source hardware based on eg. OpenRISC or some other open CPU design. The best thing we can do is go for an open firmware. Btw, do you currently *own* an opensource mainboard -I do, a beagleboard.

Quote:

Yes, do that please. There is nothing in the bounty page that refers to MorphOS. If the project succeeds, it might just will be. If the project fails, it is guarranteed not to. Think that as well.

Quote:

After the design is finished, there will be built a number of prototypes. It's in the MorphOS users' hands to have one (or more) delivered to the hands of the MorphOS developers.

Quote:

Sorry, but IMHO it does stand. Look at the Pegasos design. ROHS came and devalued it. Thus besides 100k USD being an utterly unrealistic amount of money to be gained by a bounty, especially since not many people are left seeing any advantage in a PPC desktop, there will always be the risk that unavailability of one or more key components renders also this work useless (at least in parts).

Please don't get me wrong: I really appreciate your enthusiasm and your willingness to "put your money where your mouth is" - it's just that I can't imagine this undetaking to become reality.

Quote:

We have an assurance from Freescale that MPC8610 will be available "for a long time". Also, we're taking this in steps and there are corporate players that are interested in this -and we don't expect them to donate just ?100. It's not an unrealistic amount of money. It's a lot, but similar amounts have been gained in other projects (Open Graphics Hardware has received lots of donations in the past and now it's undertaken by Linuxfund).

Quote:

Well, if it doesn't, it doesn't. Let's see what happens first ok?

Well, hardware bounty... Very risky investment. I will eventually buy the board after seeing MorphOS running on it.

I really like the idea of a new 8610 board but it'd have to run MorphOS for me to part with any cash. I'm not sure I can afford to donate to the first bounty whith no gurantee that the whole project will be completed

I'd like to clarify a few points for people here. This is not entirely about MorphOS users trying to pay 100000 EUR for a board.

Yes, this is a high risk bounty, but the gains are pretty clear; after about 30000 EUR (which can come from the community or corporate sponsorship) there will be at least some open source schematic and perhaps a half finished PCB design to go with.

Genesi does not care who takes this and runs with it. After each milestone, the deliverables will be made open source, probably under something similar to a Creative Commons Share Alike license. If a company wishes to make modifications they should contribute them back; although they could also make a covenant with Genesi to not have to do this as long as they are not causing any detriment to the project (for example, ACube could produce it, or someone could stencil a huge butterfly onto it and make a glittery blue PCB and make a Morphboard! As long as they do not take it and make it into a VME backplane SBC that is completely useless for the community. There are enough of those on the market)

As long as no major modifications are made, it will probably be able to run the standard Aura firmware (which has a small per-unit license fee) or someone can port U-Boot or Redboot or whatever they like.

What does the community get out of it? The board they always wanted, of course. For high paying donators we may institute a money-off scheme so they can buy the board at a discount on the store. If the project does not go so well, and is aborted, we can still give them a discount on the Efika MX or whatever other products we have available.

At this early stage we're playing it safe to get through each step as it comes along. If the first step cannot be paid for after a certain amount of time (at our option) then we will refund the money. After Step 2, Step 1 can't be refunded: design is done on a payment up-front basis. If we get significantly far ahead we will start instituting bounties on a developer basis (i.e. take the projects from PD and let people donate to them) and if that money reaches a point where we can fulfil the last steps, we will use that.

Genesi will make all efforts to complete the projects with the funds at hand, may contribute far more to it than the community puts forward, and it should be made clear that Freescale (and other corporate sponsors) have expressed a genuine interest in the idea.

So please do not be disheartened by it. Contribute what you can, or what you will, at this early stage there is still a money back guarantee up until the first 10500 EUR has been committed to design.

BTW Creative Commons motherboard and chassis designs are already on the market - http://www.viaopenbook.com/ and we have a view to utilizing these specifications and designs. MorphOS Laptop? Or just a PPC Linux laptop...

[ Edited by Neko on 2009/11/16 14:55 ]

[ Edited by Neko on 2009/11/16 15:00 ]

Interesting bounty.

But 65,500 EUR for 4.5 months of work, that is 14,500 EUR per month (about half of what I make *in a year*), doesn't that seem a little expensive...?

I kind of agree with Senex. I would have considered donating 100 EUR (maybe more), but I doubt you will find 654 other people willing to donate the same amount (which is a hefty amount for a general consumer considering you won't actually get anything for the money in practice) without a refund policy and no guarantees, so I don't know...

Maybe if you manage to get some big donors to put down a solid foundation, if you manage to get donors from other communities, etc. Maybe corporate entities that may benefit from this in some way, would be interested to contribute? Maybe some other HW manufacturers (Acube? Individual Computers?) and/or system builders (like Fixstars/Yellowdog), or other entities that may make use and benefit from a design like this (the MorphOS team? Hyperion?) would be interested?

@ Neko

Thanks for the clarifications...

Quote:

These prices are more or less the norm in the hardware design industry -I cannot disclose the actuall bplan offer, but I admit it affected these prices very much. Sure a Chinese design would cost less. Would you pay for that?

Quote:

It's actually less, as I already donated 1000EUR :)

I read all topic, I want to believe! So, I've donated $150 (~100EUR).

Thanks! You will not regret this!

@Neko

This is not entirely about MorphOS users trying to pay 100000 EUR for a board.

Sure. But I prefer to buy a tested board with MorphOS already released for it. Or at least to have MorphOS port 100% guarranted. I have no use for such a board otherwise.

[ Edited by Krashan on 2009/11/16 18:50 ]

I have to agree with Krashan. If there is no any kind of a message from MorphOS Team that this kind of mainboard might be supported i prefer to leave that 150e quid in my pocket and spend it for MorphOS 3 which might be running on PowerBook in the future.

@ Pampers

Exactly!

@Krashan, I understand your concerns, but getting the MorphOS guys to commit to porting to a board is like trying to get a dog to commit to meowing, as you well know :)

I am sure if you see the results of the board you can make a great case for moving to it.

Konstantinos already has an MPC8610 so this isn't a personal project to get a great new board just for himself. I have an MPC8641D (1.6GHz dual core :) and I am perfectly happy with it here. Of course someone at MorphOS needs to see the real benefits of it, especially as we can just about guarantee the performance of a MPC8610 will beat the crap out of a 1.6GHz Mac Mini just on memory bandwidth alone, not to mention the addition of PCI Express, SATA II, 2-4GB of system memory, a well documented Freescale DMA controller.. allegory and analogy is of course not enough. We understand that.

Maybe we can get you an MPC8610 system from Freescale, so you can test out the performance (I am a little rusty on this but, maybe we already did that?).

We can go through and we can pick and choose controllers for the design which will work with MorphOS out of the box if you like, or we could choose controllers you guys will know will be able to be written fast (especially since there is no public SATA, multichannel audio or gigabit ethernet chipset support in MorphOS according to the docs) or we could commit to providing this as a firmware abstraction layer for the time being to support MorphOS - consideration given only that MorphOS will implement support for that abstraction layer.

If even suggesting that the MorphOS team will look at it when there is something to look at is out of the question for the team, we will go and design it for Linux driver support. After all, desktop boards must be designed from the software backwards, and this will be a waste, in my opinion.. it may cause you more work in the future if you decide that it is a relevant porting effort.

@Neko

1. I'm not denying MPC8610 performance.

2. I'm not the one in MorphOS Team, who decides which devices are target for ports. I have not enough knowledge and source access to do a port alone. I can write eth or audio driver but probably not much more. Then I'm not the proper person to talk about choosing controllers... From my POV any well documented ethernet and audio chip will do. But of course work on these drivers requires kernel and basic peripherial support to be done first.

Quote:

Perfectly understandable. In this case, it would help if one from the MorphOS team, would take the effort and just clarify their stance with regards to this project. It would be useful to the users and to us to know if the team will even consider such a port and under what conditions. As Matt said, it would be vital to the design process as it would make sure that components are selected on the criteria that they would be easy to port to -perhaps a BSD driver already exists that you could make use of, or something like that. If the team will not even consider such a port, there is no reason for us to make false promises to users who will expect MorphOS to run on this system, when there is no chance in that.

Maybe now time for understanding, need all of this for us or not in future?
What about "Red" side? It's will be nice, if Hyperion and MorphOS team release statement about support this hardware (in case of success development).
I think, this may increase chances for hw dramatically, with help from "red" and "blue" worlds.

it would help if one from the MorphOS team, would take the effort and just clarify their stance with regards to this project.

I guess you will need to send a direct e-mail to Ralph Schmidt to get such clarification. General point of view of the team to the new hardware has been written a few times. The key is "we do not port to nonexisting hardware". In case of hardware produced in low series, there is a high risk of hardware bugs. These bugs have serious impact on porting time and resulting port quality. After fighting with bugs in Pegasos 1, Efika, and - to some degree - Pegasos 2, and having in mind MPC5121e disaster, the lesson is well remembered. Then possible port of MorphOS to the bounty hardware may be considered only after the hardware is produced in some reasonable quantity (not a prototyping stage) and is proven to work reliably and according to specs.