Receive a bounty of $10,000 For : Quake III at a playable framerate - Raspberri Pi
Bob Lawrence (VE1RLL)
Posts: 1,720
Propeller head gamers should have no problem with this.
Earlier today, Broadcom announced the release of full documentation for the VideoCore IV graphics core, and a complete source release of the graphics stack under a 3-clause BSD license. The source release targets the BCM21553 cellphone chip, but it should be reasonably straightforward to port this to the BCM2835, allowing access to the graphics core without using the blob. As an incentive to do this work, we will pay a bounty of $10,000 to the first person to demonstrate to us satisfactorily that they can successfully run Quake III at a playable framerate on Raspberry Pi using these drivers. This competition is open worldwide, and you can find competition rules here which describe what you have to do, and how to enter.
http://www.raspberrypi.org/archives/6299
Earlier today, Broadcom announced the release of full documentation for the VideoCore IV graphics core, and a complete source release of the graphics stack under a 3-clause BSD license. The source release targets the BCM21553 cellphone chip, but it should be reasonably straightforward to port this to the BCM2835, allowing access to the graphics core without using the blob. As an incentive to do this work, we will pay a bounty of $10,000 to the first person to demonstrate to us satisfactorily that they can successfully run Quake III at a playable framerate on Raspberry Pi using these drivers. This competition is open worldwide, and you can find competition rules here which describe what you have to do, and how to enter.
http://www.raspberrypi.org/archives/6299
Comments
It would be nice to have the added resources and the ability to have a swap disk partition of a SATA harddisk.
It would be nice to cost $35 and have 2 million units sold.
{sorry, not in a very ergonomic mood today!
We hear how tech jobs are going wanting for lack of talent...but never a mention of what employers are willing to pay for said talent.
One would make more money working at McD's at minimum wage than to pursue this challenge.
Based on a $ amount per hour maybe, however it depends on how talented you are and your area of expertise or interest. Of course not everything is based on money. A young programmer could gain a wealth of experience and industry recognition while accomplishing something that may find to be both a challenging and rewarding experience. It could also be a stepping stone to an exceptional programming career.
99% of what people get up to with their Propellers here is not done for the money. That's millions of man hours, billions of dollars of lost potential income.
And that is a drop in the ocean compared to all the other pointless, timewasting, non-revenue generating activities the rest of the human race pursues every day.
I guess one is going to have to provide a solution for a Linux OS environment... not sure that Android would be open enough for development.
It uses opengl.
It runs on Android, so in must be able to use GLES.
The Pi has GLES.
How come Quake III needs any work to run on the Pi?
Quake III already runs on the Cubie according to this forum post http://www.cubieforums.com/index.php/topic,582.0.html
I guess if has already been done, you could cash in for the $10,000 rather quickly.
What I do suspect is that Broadcom is hoping to get somebody to optimize their video for games in OpenAVR. And that is a lot lower in level than the LinuxOS.
Or is there another OpenAVR I'm not aware of.
Correct...they are looking for slave labor to turn a product line.
I agree...bragging rights are something.
But I am more of a "Show me the money" kind of guy...especially when a company will be finacially benefitting from my efforts.
That aside. How is this different than Parallax encouraging people to donate their Spin creations to OBEX for free under the MIT license?
Stupid Propeller people, why do they do that?
Which implies that Raspberry Pi is willing to help such an exploit as opposed to their claims of being primarily an educational mandate being computers to a new generation of learners... a non-profit foundation that really seems to be a marketing arm of a for profit enterprise.
The reason that the CubieTruck will handle it game is simply more resources. The generation of chip that the Raspberry Pi uses is earlier and less able to handle gaming demands... which have always been intensive.
HOW is it different? The scale of what Raspberry Pi is doing is much larger in terms of benefits to the private industry it is supporting AND it has claimed to be a non-profit enterprise which is going to assist education.
Parallax is openly for-profit, and OBEX contributions are much smaller in comparison. And OBEX contributions tend to be much more realistic in terms of offering new learners a useful tutorial example. Writing a video driver for a Broadcom chip is pretty much a task for advanced programmers, but beginners or intermediate learners. (if you are going to teach, students have to start out with the basics and progress. This contest is way beyond teaching the basics.)
Also, there is the fact that Broadcom wants someone to do an open-source version so that it can vastly expand chip sales by the vast richness of the Linux open-source community.
In sum, with Parallax... learning is very accessible. With the Raspberry Pi... first learn Linux, and the cross-compiling tool chain, and the assembly language for the chip AND then you can begin to do something with the Pi. It demands an awful lot from a naive new learner.
Hey wait a minute, all this talk of "slave labor" and "exploit" is not only misguided it's offensive.
The guys in the Pi foundation have been busting their guts for years and using their free time to bring their vision to the world. They all have full time jobs. They don't get paid by the foundation. Thanks to their efforts corporations are now easing up and releasing source code for their GPU's and such. Those guys do not deserve this kind of abuse.
Don't forget that the orginal efforts at building a Pi were not based on Broadcom devices at all. That was a happy accident that came much later.
Nobody said a charity cannot sell stuff and make money to further it's charter. Be sure the Charities commission and the Inland Revenue will be watching for any misuse of funds like a hawk. If a charities actions happens to include helping sales for a company, so what? Sounds good for everybody.
Come to that, nobody said a company, like Broadcom, cannot offer help to charities and the like that help their bottom line. What? Perhaps learning a little about Linux is part of the educational remit. Certainly seems like a very good idea. Nobody has to cross compile anything to use a Pi. Nobody has to learn assembler to use a Pi. Where do you get this stuff from? Perhaps. Certainly not anything like as much you are making out. This not true. The Raspi may have less horse power than a Cubie but it does have a good GPU. I believe it's plenty enough to run Quake III.
I was running Q3 on an AMD K6 back in the day, which had a lot less horse power than a Pi. Not to mention it had a much less able GPU.
Absolutely it is.
TIme for a little perspective.
There is a body of open code out there. This body consists of Linux, Droid, GNU utilities and development tools, countless libraries, drivers, and who knows what else. A very large fraction of that code was written by people without some direct compensation.
Slave labor? Fools?
Now think about the concept of use value. Code has a use value. Actually, it has a few kinds of use value. One can read the code to learn how to do something. Code as documentation / education. (and we need more of this skill being passed on to people in the world) One can combine that code in various ways, often adding a little bit more code, build it, run it and do something. That's code as application or software. Finally, there is that code empowering one to employ hardware in various ways, like operating systems, drivers, etc... With that last one, it's possible to fully understand what a device is doing, because the code as education use value means people can go and read the code and learn. They don't have to, but it's very important that it be possible for them to do so, if they are inclined to do so, or they pay somebody else to do so.
Now, why write code for free?
There are many bodies of closed code. These code bodies sometimes have use value in terms of empowering people to use hardware, or author their own programs, etc... but these code bodies have near zero, unless you've paid for an expensive source code license, code as education value, and that means they also have a near zero trust value. If you can't open it up and look inside, you really don't own it kind of thing.
Closed code is generally produced for profit, and the fact that we always use code in tandem with a license, whether we are just reading it, building it, executing it, whatever. There is a license. Open code has the basic uses licensed in ways that do not cost money, instead they cost additional code to be open in like kind. Closed code has basic uses licenced, if they are licensed at all, in ways that do cost money per user, cpu, execute count and every other metric you can think of.
Got it?
Now, let's say you want to do something. You can write all your own code to do it, and that's fine. You will have paid what it costs you to not work on other things, eat, exist, etc.. to produce that code, and doing that can be a project involving large sums of money. Because of this, and the natural limits of people and how long they live, we pay for code already written, and this is the basic value inherent in both closed and open code.
With closed code, you pay them to get it done and you run what you get from them. With open code, you don't pay anybody in particular, but must release code you add as open in like kind.
Either way you pay, but what do you get?
With the closed code, you get the well defined thing, task, function, you paid for, and that's it. Sharing that use value is difficult, often illegal, expensive, you name it. With the closed code, you get whatever well defined thing you are doing, plus anything else you want to do, and you can share it with the whole world!
Lots of use value there for those so inclined to make an investment in learning about open code. And that is an investment, and it's costs are on par with buying software. The difference is open code investments are yours to do whatever you want to with, while closed code investments value are linked to the economic interests of those who you bought the code from. Big difference there.
Now, software has a very interesting property! The whole is very often greater than the sum of it's parts! In other words, you very often get more out than you put in!
Somebody can look at the body of open code, realize a piece is missing to realize the use value they need. They can move on and buy closed code and spend that money, or they can go ahead and author the piece they need, combine it with the whole, and get far more use value than labor value they put in.
This is why somebody, somewhere will complete this project and it has nothing at all to do with wages, etc... and it has everything to do with improving the use value of the open code for whatever reasons they deem worthy, while the rest of the world benefits in like kind.
There is pretty much nothing else in the world like this, and it's very often misunderstood.
So then, the bounty isn't about compensation at all. It's an advertisment intended to make as many people, who may be inclined to realize that kind of use value as possible, aware of the need. And it's a nice "thank you" among other things.
re: "show me the money" Well, let's say you author that piece with the $10K bounty. And you labor long enough to result in some basic wage per hour. What did you get? Let's talk about buying what you can now do with open code. Game engine, development tools, drivers, sample game code, actual retail game code, graphics and art tools, documentation tools, modeling tools, rendering / texture tools, life sciences, visualization, and I could go on with operating systems, etc...
How much does it cost to buy all of that? More than you've probably made so far in life, and that right there is the value of open code put in real, hard, economic terms.
On a side note... Q3A FOREVER!! (Love that game way too much. Capture the Flag, anyone?)
I did a bit of checking, some of the Raspberry Pi installation has used OpenAVR to compiled. That is where I got into thinking it belonged to the OpenAVR community.
The processor is an A11, while Cubieboards use something different , an A10 or A20 from a different vendor developer.... much faster, more ram.
Very similar, but not exactly the same.
The rightous indignation of you Raspberry Pi supports is simply offset by my dismay as a real teacher that so many corporations tout education, so many non-profits wave the banner for a better world while banking the cash.
As yet, Raspberry Pi has yet to deliver of the vast majority of its original hype. While Broadcom and the makers of A11 and others have done very well in getting a heck of a lot of public recognition.
One failure here seems to be that it won't play some of the games it originally claimed it might.
+++++++++++
But what annoys me most is that Raspberry Pi has exploited the Paralllax Forums and likely convinced potential Propeller users that the Pi might be a better deal for robotics.
It is general purpose operating system with a GPIO. The combination will never perform robotics anywhere as well as a Propeller. The mutli-tasking overhead and the OS are not opimal for such.
And then there is the obivious fact that the CubieBoard outperforms the Pi in its strong suit of a SOC device.
Seems to me, you've got a Pi problem, as well as a basic misunderstanding of open code and the dynamics it brings to the table.
Of course Broadcom and friends are doing well! That's fine! Hardware that has high use value always does. And having open code run on that hardware delivers very high use value to anyone willing to learn. We want this far more than we don't, and somebody somewhere has to make hardware, and we all like hardware, so what really is the problem?
Hardware does not have the property software does. The whole isn't greater than the parts, unless it can run software!!
Buy a Pi for a few bucks, and then what? On closed code, you then pay for a ton of software, the cost of licenses dwarf the cost of the Pi to the point where really they give the Pi away when you buy licenses for closed code! This does nobody any good, but for those writing closed code.
On the other hand, buy a Pi and have it run the body of open code, and suddenly, that small amount of money, plus a personal investment in learning how to use open code, nets a person use value equal to huge sums of real dollars, and it happens all over the world for anyone who will ask and who can get their hands on a Pi.
See the difference? BTW: A real teacher really should. I've educated a few in my time, FWIW. This stuff matters, and failure to include it really doesn't do anyone any good. Just saying...
What can we say to get you on board here? Why all the negativity? Where do all these misconceptions come from? This is totally uncalled for. Basically untrue.
Eben Upton, whose idea this all was, was a teacher for a long time. He was teaching undergraduates at Cambridge University. It's there that he noticed that over the years the level of basic computing knowledge of the kids entering the university was getting less and less. Not only that the number of applicants had dropped dramatically. He had a theory about this:
In years passed the new university intake had been brought up with C64's and Sinclair Spectrums and the like. Many of them will have been hacking around programming since were 10 years old. Many of them knew more than one assembly language by the time they got to Uni. In later years kids only had games consoles. The family PC was not that cheap quick easy place to start exploring programming, indeed an out of the box Windows computer does not even advertise the fact that it is general purpose computer that it's owner could program in any way. As a result kids were not getting inspired to program or be interested in Computer Science.
Eben thought that possibly giving kids of today something of that out of the box "program me" experience of the C64 era might inspire some of them. It had to be small, cheap, disposable almost. Hence the Pi.
You know what? Kids love it. Although we have yet to see if it has any effect on increasing numbers of people who aspire to be software engineers.
A little story about Eben, as a kid he bought a mouse for an 8 bit computer, I forget what it was exactly. Of course the mouse did not have any software with it, heck that machine did not even have a GUI interface to use a mouse with. His father called the supplier and they said "If he is too stupid to write a driver for the mouse he should not have bought it". On that the young Eben wrote his first device driver in assembler.
"...starry eyed and naive.." I don't think so. Call it what you will but when untruths are spoken and abuse thrown around it needs to be called out. What has this got to do with the Raspi Foundation? Are you making accusations that someone is syphoning off the money, what little of it there is? Please don't do that. What hype? I don't recall any coming from the Raspi Foundation.
In fact I'm amazed they delivered far more than anyone expected. They doubled the memory for the same cost. They introduced a very cheap and easy camera module. Support has been terrific. And so on. What games? Who said it might? References please. I don't recall the Foundation ever claiming the Pi to be something it is not. You have missed a big point here. The Pi is not exploiting anything. They are not in competion. Many of us here, see Bill Henning's latest product, believe the Pi and the Propeller are complementary. When some of those two million Pi owners realize they need some help with real-time stuff they should be turning to the Propeller. That is a lot of potential sales for Parallax. You will not be helping by insulting Pi users when they get here. Yes of course. They are different animals. But if you can't see how a Pi and a Propeller would not make a great combo for a robot then, well, I don't know what to say... Yeah, so?
I'm using Pi's in embedded systems that only ever see 10% CPU load. Why would I want to pay twice or three times the price to get more horsepower? Why would I want to give up the wonderful world wide support network of Pi users?
P.S. I will be getting Cubie Truck for evaluation and fun. Don't really need it.
Now stop it with all this negativity toward the Pi or I will have to come to Taiwan and flog you with noodles
A. I reserve the right to dislike the Raspberry Pi. It has positioned itself as a very noble cause from the start, had people such as yourselves strongly depend it regardless of the realities, and it is just a tiny Linux machine at the end of the day.
B. Educationally, whatever the Raspberry Pi nearly anything that has to offer educationally can be offered on any computer that will run a copy of Ubuntu, Debian, or Redhat.
Why does it all have to be repackaged as Raspberry Pi for the sake of ARM A11, England, Oxford, Cambridge, and what-his-name? There is just a huge mythology driven machine behind what the Pi really is.
C. The Raspberry Pi struck me as being too small, bottlenecked on power, bottlenecked on USB ports, carried away with the idea that children were going to quickly grasp Linux and managing the GPIO for robotics just because it was credit card sized.
When I pointed out doubts about feasability, I get the mantra of Raspberry Pi's noble cause and great sacrafice. It sounds phony to me.
D. There are a lot of non-profit institutions in this world that exist just to make major corporations look good. And the people that serve on their boards, do reap benefits in other ways from doing so.
Just a bit of history....
In the USA we have 'The United Way' which is a big corporate supported charity. Bill Gates would have never sold MSDOS to IBM if it had not been that Bill's mom was sitting on The United Way board along with the chairman of IBM.. so she managed to get IBM to invite her son to try to provide something for IBM.
The reality is that an up and coming 'young man' might serve a stint heading a non-profit and later be rewarded with better opportunities. It seems to me that is likely part of the Oxford, Cambridge, Raspberry Pi, ARM A11 working relationship.
Since I am not part of the in crowd, I'd rather not participate.
++++++++++++++
But having said all that, let us get back to the Broadcom Contest and the fundamental question of what it has to do with the non-profit mission statement of the Raspberry Pi Foundation.
After all, the is a Raspberry Pi sponsored contest, Broadcom could have done the contest independently without Raspberry Pi.
It seems to me that it is a 'perfectly legal' publicity stunt for Broadcom that has very little to do with all that dedicated teacher hubris that Raspberry Pi puts out. Raspberry PI has the press release machinery in place to reach all 2,000,000 punters -- so Broadcom goes with the Pi network.
++++++++++++
It really doesn't matter that you evoke Oxford and Cambridge.
My experience in teaching English as a second language has shown me that both are more interested in driving sales of books, CDs, DVD, IELTS testing, and CET certifications for revenue to the U.K. than fundamentally teaching the language to Asians in the best manner possible.
And it could be Stanford and M.I.T, the same applies on the other side of the pond. M.I.T. offered an extremely popular free course on Artificial Intelligence that was vastly subscribe. It was easy for them to do so as the lectures were all on video. But the real windfall was the sales of the $200USD textbook and the share the university got for driving the added sales.
Non-profit merely means that in the event of ending business, all the assets must transfer to other non-profits. It has nothing to do with paying big salaries or selling profitable goods
++++++++++++
Of course you are free to not like the Pi. May be it's underpowered for you. May be it lacks features you want. Maybe you just don't like the colour. That is all fair enough.
What is a not fair is spreading untruths about it and the people that made it. Throwing abuse at them is just inexcusable. Straight away there you opened with "regardless of the realities" as if there were something wrong with the Pi deal. Yes "it's just a "tiny Linux machine", if I remember correctly when it was launched, certainly when it was conceived there was no such "tiny Linux machine" that normal people could buy at any reasonable price. At the time we were using ARM boards from ISEE, similar in many ways to the Pi or Cubie. At 200 euro a piece! It could be argued the Pi kicked off a little revolution as companies realized there was a huge untapped market there. That is true. At the time if you wanted to do that, assuming you knew there were such things, you needed a PC or Laptop to do it with. Hardly in the same league now is it? Making exactly that possible on a budget is exactly why there is a Pi. Not true. It does not. Out of the box Debian for ARM will run on the Pi just fine. The Raspian distro is the same Debian stuff recompiled for a bit of optimization if you want it. Which strangely almost enough everybody does. 99% of what runs on the Pi is available for the Cubie or your PC. It's all the same opensource stuff right? The 1% that is special to the Pi you won't want on other machines anyway as it is low level and hardware specific. No, there is not. The Pi specs and capability are clearly stated. "What-his-name" was named just above, why do you feel the need to put him down? Too small and bottlenecked for what? Works fine for what many want to do. Not true. The stated aim was that kids should be able to boot it up and immediately be able to start exploring programming in Python or Scratch or whatever. That may not be totally polished yet but it works for many. Not much Linux knowledge is required. As for the GPIO that's what libraries are for. A lot of kids have having fun with that. It's a reality. Why call it a "matra". Your doubts about feasibility are based on false assumptions, misinformation, and non-facts. They have all been easily refuted so far. Anyway the proof is in the pudding, the Pi concept has proven itself to be perfectly feasible. That may or may not be true. It's beside the pint anyway. But so what? Why shouldn't hard working, intelligent, worthy, "young men" move forward. You want them to be monks or something? You will then, of course, be throwing those Cubie's in the rubbish bin. The ARM was devised in Cambridge and still developed there today you know:)
There may be many reasons to complain about Oxford and Cambridge but that is way out of scope here. Isn't it obvious. The Pi Foundations charter is to get the means of programming into the hands of the young and to inspire them. Q3 is a famous and popular game. Getting it on the Pi is a big hook to make the Pi attractive and get kids interested. Thus furthering the mission. Not in England. A Charity has a charter. A "mission statement". Without that it cannot get it's charitable status. With that in place a charity cannot just do anything that happens to make no profit. It's activities have to align with it's charter. You are right about the "big salaries". The Pi Foundation has recently appointed it's first employee, I'm sure that salary is not huge.
I just watched Micro Men on YouTube last night, awesome good movie - probably ought to download a copy while it's available.
http://www.youtube.com/watch?v=XXBxV6-zamM
Contests are an often used vehicle by companies to get something for free.
Thanks! It was fun to watch.
Whilst you are at it do check out this other brilliant BBC production about Bob Noyce, the invention of the micro-processor. The start of Intel. The birth of silicon valley.
http://www.youtube.com/watch?v=-wO9Eu5I-hM
Be sure to find all the parts.
Can any one spot the bug in the program given for calculating the average age of the human population?
Why are you doing this?. Now I have to watch youtube instead of programming JavaScript.
Its all your fault.
Enjoy!
Mike
Sorry, yes, it's my fault.
I've been hacking away on JS all week. Check this out http://localhost:3000/media_test.html
So I need a little rest.
But there is more. Going back to the 1930's. You really have to see this: http://www.youtube.com/watch?v=qundvme1Tik
Conrad Zuse had the right idea. He said that for his computer for engineers floating point was essential. He had one number type, floating point. Just like JS. 32 bit floating point back in 1937 or so !