Anonymous Will Award $200,000 for Xbox Linux

An anonymous reader writes: "The X-box Linux Project at Sourceforge reports today that an anonymous donor will award nearly a quarter of a million dollars to the individuals responsible for the completion of a two-phased effort to run Linux on the Xbox. One can't help but wonder if this will help or hurt the community. On one hand, it is likely to generate additional interest in the project, on the other, some people may be less inclinded to share their discoveries with money on the line. Then again, getting both Money and Glory sounds pretty good."

Re:hardware changes?

[David Price]

Reading the press release: a hundred grand is for a distribution of Linux for a modified X-Box, and another hundred grand for doing it to a stock console.

Re:Billg

[Xerithane]

Ok, the Xbox-Linux team knows who the donor is. It says that clearly in the site that the donor is known to them.

Do you think they would partner up with Microsoft for it? I don't... but then again, that's just my thoughts..

Re:This will hurt Open Source developers

[seanadams.com]

Won't the $200K reward encourage greedy developers to hide their work and end up reducing the amount of sharing that goes on?

Read the rules. Results have to be submitted to the sourceforge project, and licensed under GPL . This would be pretty pointless otherwise...

Full details and rules

[seanadams.com]

Note that the reward isn't all-or-nothing - it's partitioned into five distinct tasks, in two separate porjects. That also means that different people can claim the prize money for each task. If two groups solve the same problem, the "better" solution gets all the money.

Project A:
Task 1: Replacement BIOS - $55,000
Task 2: Kernel and XFree drivers
- 25,000
Task 3: Kernel logic: FATX and miscellaneous - 10,000
Task 4: XBE bootloader $10,000

Project B:
Run unsigned code on an Xbox without any hardware modification - $100,000

The goal...

[zeno_2]

The basic goal of the project is to find a simple and completely legal way to run Linux on the Microsoft Xbox.

The whole project is divided into two sub-projects, the first one consisting of four tasks.

Project A: Porting Linux to a modified Xbox:
Task 1: Replacement BIOS (software/hardware)
Task 2: Kernel and XFree drivers
Task 3: Kernel logic: FATX and miscellaneous
Task 4: XBE bootloader
Project B: Xbox hack without any hardware modification

A total of US$ 100,000 will be awarded for the completion of each of the two projects.

Well, I hope someone can do this, it would be very interesting to see, but I don't get how they can do all of this 'legally' as the anonymous donor wishes. To complete project b, are going to need to find out how the Digital Rights Management (tm) system works on the Xbox, and that would violate the DMCA as far as I know, but im no lawyer. I hope they clear that up soon, unless they expect this person to deal with microsoft lawyers to license DRM somehow, but i doubt they can if they have to disclose the reason why they want to license the technology.

Best of luck to all those capable of doing this though.

Re:legal expenses

[jfunk]

From the link:

The basic goal of the project is to find a simple and completely legal way to run Linux on the Microsoft Xbox.

Ahem.

Is An XBox Cluster REALLY Cheaper?

[MyHair]

A cluster of 1000 Xboxes would be mighty cheap computing power.

I'm too lazy to actually calculate this, but I have to wonder about the $$ feasibility of an XBox cluster. Okay, I could believe it's a better value to hack an XBox than to buy a PC for gaming, if you don't take future upgrades into account. (GeForce 6's and Radeon 12000's probably won't have a USB or ethernet interface; just a guess.)

But if you want to make the ubiquitous Beowulf cluster of XBoxen to crunch numbers, is it really more cost effective? Even if someone figures out how to put Linux on there without a hardware mod, you need to consider that the graphics and sound capabilities built-in won't be used in the cluster.

Don't compare an XBox cluster to a cluster of Linux gaming machines but to a cluster of bare-bones dual-cpu boxen or rackmount servers with no or minimal video, sound and i/o capability. Plus compare the power consumption, cooling and space requrements of the two since this becomes nontrivial with a cluster.

Plus, who with such high number-crunching needs would put up with the dearth of hardware support for Linux on XBox. You can't just swap out a motherboard, power supply or ethernet card on those puppies, at least not as easily as a desktop, tower or rack PC.

I don't think an XBox cluster is reasonably feasible beyond the geek in me saying "that's so cool that someone did that!" However for us Linux geeks and gamers I'd love to have Linux on XBoxes. (Not necessarily to own one, before you Linux Dreamcasters jump on me.)

Re:Billg

[BinBoy]

It can be done. It's already running a stripped down version of Windows 2000, as an MS engineer mentions in his "Inside the XBOX Launch" talk here: http://www.technetcast.com/tnc_play_stream.html?st ream_id=666

Fill your hard drive with music, movies and pictures while you sleep. [binaryboy.com]

Not likely

[kinnunen]

I didn't actually do much checking on the prices but they should be reasonably close.

Athlon XP 2000 - 150$
Cheap mobo with etherent 100$
128MB DDR SDRAM - 25$
Case and PSU - 50$
8GB HDD - 75$
----------
Total 400$

Yep, it's twice as expensive. But in a clusternode it's usually the the CPU that counts and XP2000 is 2-3 times faster than what is in an X-Box. A cluster node doesn't need a DVD drive or a top of the line Gforce4. You may not even need the harddistk. With 200,000$ you can get 500 nodes like this, or a linux distribution that boots on X-Box - but you still need to buy the 1000 X-Boxen to run that distro for another 200,000$. And of course 6 months from now the the Athlon config will be ~50$ cheaper, while the X-Box is steady at 200$.

Xbox BSD

[Above]

I'd rather see FreeBSD on it, I wonder if that
would be worth any money.

wha?

[jon_c]

No, no, the XBOX looks nothing like that, it's more like:

P-III 750 - $60
N-force mobo - $70
GForce3 Ti 500 - $250
64mb SDRAM - $10
8GB HDD - $75

the prices are made up, but thats pretty much the xbox, also one should concider the possibility that the gforce could also be used to crunch some numbers, sort of like a really fast MMX.

-Jon

Display Composite Images?

[MyHair]

What if the cluster is used to composite images.

I'm not sure what you mean. Do you mean something like a 3x3 arrangement of 27" TV's set up to make one big picture?

If you're proposing using 1 XBox per TV for such a display using 3D rendering I think you'd need to have the master 3D model on one box, thereby using only one XBox CPU for real-time modelling (of course all GPU's would be jammin'); I can't imagine offhand how to implement a distributed 3D model and distributed display, or even if that's possible that the inter-XBox communication would be fast enough. Without the need for a mod chip and with using real-time high-performance 3D rendering I might imagine this working out cheaper than PC's with high-end 3D cards. (Not counting if you grab used PII's with AGP motherboards.)

However, if you're just using such an array for a 2D display I would imagine--since I'm too lazy to investigate--there exists a VGA-to-NTSC adapter that would drive a TV array from a video card, and I would suspect that it would be cheaper than 1 XBox per TV. (9 Xbox * $200 = $1800. Throw in hub and network & power cabling.) Add PC's, dual-head card or 2nd video cards as needed to scale up.

I'm not sure what else you might mean by "composite images." I was imagining a collage or blend of some sort, but that can be done before feeding it to a display adapter [array].

Nevertheless, to do it with XBoxes would be a geeky cool achievement if not necessarily "the best way".

I'm curious about your project. Do you have a link or description?

The State of the Art

[warmcat]

I did not see anyone mention XBOXHACKER yet, which is at

http://www.xboxhacker.net/ [xboxhacker.net]

The BIOS hacking forums there is a focus of efforts to reverse-engineer the X-Box for the purpose of allowing Linux to run on it.

In the last few weeks we have successfully recovered the RC4 key used to encrypt the second bootloader in the BIOS, this has led to discoveries about the PIC chip that have allowed a minimal clean BIOS to run for the first time.

I also run a site at http://warmcat.com/milksop [warmcat.com] which has a variety of GPL hardware designs that are of use in getting the X-Box to run Linux (although they have many other applications).

On the prize, I worry it will change the ethos of people working towards this goal, which until now has shown the best side of people with a common, righteous purpose working together.