Despite all the excitement and expectation encompassing the RaspberryPi, the most remarkable thing about this low-power credit card-sized computer is its price tag: little more than £20 for a fully functional system capable of, among many things, 1080p video playback and hardware-accelerated graphics. The British-designed Pi is …
I don't know ...
I was editing video on a 386 about 12 years ago. Can't say it was particularly fast but it did certainly do it. And conversions from one format to another were pretty OK.
"you don't realise that your "open" Linux box is running a pile of closed proprietary binary code which the graphics card manufacturer has provided to support the minimum functionality that can be labelled "Linux video acceleration"
I think that if the video card manufacturer is providing Linux drivers for that GPU to work with Linux in the first place then its internal microcode cannot be in any way shape or form be described as "proprietary".
An awful lot of graphics code comes from a free and open source anyhow and yes this finds its way into "proprietary" hardware.
I think you'll find that most of the people working on video for Linux disagree with you.
More than possible. I admit I may be wrong -- I'm just assuming if you can see the driver source then you could make pretty accurate assumptions about what's going on in the hardware. That might be very wrong of me and I apologise if I've offended you in that assumption. No offence was intended.
Videocore IV supports H264 High profile 1080p30, along with MP4. However, those codecs that need an expensive licence won't be suppled. Don't blame the Raspberry Pi, blame the licence cost. Interestingly AAC is the highest codec cost of the lot, and yet is the simplest format.
What do you mean, "driver source"?
What such vendors do is provide a "blob" of executable code, possibly entirely proprietary firmware to load onto the card, and perhaps a shim to stick it where it needs to go. Sometimes that shim is written by some linux hacker and thus GPL and then it might be loading some binary filched from the windows driver distribution instead of anything "official".
That shim is the only bit you'd get to see the source for and isn't particularly helpful in fixing anything, nevermind writing your own. It's one of those things that sets the "tainted" flag on the kernel, and oftentimes won't on anything non-x86, or on the wrong version of the vendor-blessed linux distribution, or.... Vendors acting like linux can be supplied with drivers like it was windows. It's a bit of a mess, it is.
Of course, there do exist vendors that do supply open drivers for their hardware, and sometimes they're even decent to good. But those are few and far between. Most vendors and for that matter most of us are usually better off with the specifications published free-for-all; someone'll pick up the specs and write the driver for their pet OS. But of course with the increased threat of patent abuse there's a perverse incentive to make a pheripheral and then not tell anyone how to talk to it. Then again it does help if people contact vendors and ask for programming specs. If enough people ask for them, then more and more vendors will release them.
"or you don't realise that your "open" Linux box is running a pile of closed proprietary binary code which the graphics card manufacturer has provided to support the minimum functionality that can be labelled "Linux video acceleration"."
The proprietary binary code blobs i've used on a number of Linux boxes have provided an awful lot more than 'the minimum functionality that can be labelled "Linux video acceleration"' - don't be such a troll. There is a noticable, sometimes very large, gap in performance between, e.g. the AMD/ATi Linux and Windows drivers, in a number of areas - to be expected given the economics - but they are far from minimum functionality. The gap is also closing in some places, all from the same vendor.
You sound like you may know a fair bit about graphics, so why the mis-leading nonsense ?
Yeah but if a lowly Nokia N8/C7 with it's measly 680MHz ARM11 can record and play 720p then who are we to complain about not having enough oomph?
It's clearly a matter of optimising the OS. RISCOS on an 8MHz ARM2 equipped Archimedes A440 ran circles around my Atari PC4 (80286 16MHz) at that time (one o/t reasons I ditched my PC and Amiga with its 68030 to get an A5000). And later (early 90's) my ARM3 equipped 25MHz A5000 was a lot faster than the usualy PC clone 80386SX/DX. Things didn't change much untill the Pentium 2. Then with the arrival of dedicated graphics-accellerators and the much higher clockspeeds PC's began to overtake the Acorn machines in performance.
The Achilles heel of the Acorn systems was their abysmal floating point performance. Which didn't matter until the Pentium's reigned the market with their integrated FP-coprocessors after it became clear that 486DX sold a lot better than their FP-less counterparts 486SX.
RISCOS was (maybe still is) assembler-optimised which is the reason for it's performance. The same with current Nokia cellphones. Symbian is the most optimised ARM-platform to date hence the reason why Nokia can still use olde ARM11 cores and still get very decent performance out of them. I'm let to believe that Windows Phone 7 OS is also quite optimised (it should be if you consider how long M$ have used ARM-code/cpus).
The Nokia N8 can do 720p encode because it uses a Videocore 3 GPU - the predecessor to the Videocore 4 GPU in the Raspberry Pi.
Small world eh!
I hope this does well..
Consider for a moment that this hardware is perfectly capable of running a typical "office" application set, and there may be some serious interest from corporate purchasers - £200 a head for a PC, or £25 for this? (By way of a supporting argument, Microsoft are currently porting Microsoft Office to Nokia's Symbian3 phones which use a very similar hardware configuration). In these times of cutbacks, these would also make good student workstations for colleges - especially if you want to teach embedded and mobile development.
I'd like one^Wa couple, even if I have everything it does already.
Currently using a rather anemic old box (single core C7 at 1.5GHz) booted off of usb for the usual browsing and mail checking, which when booted from another usb stick gives me a browser with flash so I can watch video lectures*, take notes, and do some excercises. The former is FreeBSD**, the latter is tiny core linux. So likely the raspberrypi will actually be faster than what I'm training neural networks on now. Bit of a pity I'd have to buy something with hdmi to use it, but I just might as I need a better display anyway.
Also a bit of a pity the graphics chip is closed up, someone really needs to whack broadcom over the head for being so annoyingly secretive. But at least the mobile opengl is something. Now for opencl on that same hardware.
Thing is, the ability to stick something in your pocket that's a (near enough) complete environment to do some task is actually quite powerful, moreso than I expected. Even if you can't take the display with you also. We possibly might see the revival of a c64/bbc/amiga-esque "scene".
On the other hand, though, I'm seriously looking at acquiring some new, much faster kit to give me back a good development environment and a bit of oomph for running data analysis and such. Part of the price (some 500-odd ukp projected and damn those floods) is virtualisation support (with x86's notoriously poor historical support and architectural supportability that's noticeable, yes) so as to run various "OS containers" concurrently. I think that'll stay too, as sometimes the oomph and convenience of not having to swap out devices on the telly is worth something too. Now for some way to migrate those things back and forth from the cloud or even say a friendly local organisation's handy K super clone for real computation, to the desktop to keep it all in hand, to the laptop to keep tinkering abroad, and back again, preferrably seamlessly. There's much to innovate yet.
* The stanford online classes, all three of them. I really should stop slacking off.
** It could do flash with suitable emulation installed, but I opted for trying the different stick method instead.
That's Christmas sorted...
...for my son obviously ;-)
Still too complicated....
All this talk of tweeting, web browsing, linux nonsense - kids already have that crap on the PCs, why make this thing do all that?
Machines like the speccy and C64 inspired people because you turn it on and type some BASIC and things happened on screen. When you got the hang of it you move on to machine code and do better things.
And there is absolutely no reason why you cannot do this- perhaps to Python, or even BBC basic. It's just a SD cd swap to change the nature of the beast.
@ Anonymous Coward
Unfortunatly ppl demand some form of basic usability (which include websurfing).
On RISCOS you simply press F12 to get to a CLI. From there you type 'BASIC' to revoke the BBC Basic interpreter and from withing that BASIC you use [ ] (square brackets) to insert your Assembler code. Neat and easy :-)
The IMHO very best example of this kind of programming is the program Architec which was written in Basis/Basic-assembler. It was a real time 3D modelling package with real time texture mapping etc...
I've seen it's code. Not that I could do what these guys had done (I think it was Aspex Software from the UK whom made it) but it showed what could be done in these systems using the inbuild BASIC/Assembler tools.
Impressive feat of engineering, and I really hope it will be successful and help kids get into computing. Especially those who might not have any other chance.
I am a bit dubious though. How many people is it going to help?
People on this forum could just buy one for £20 and probably have enough spare kit to build a system.
For a household which doesn't already have a computer, in addition to the RPi they also need to buy a keyboard, mouse, USB hub, cables, power supply, SD card and probably a monitor. Starts to make a second hand laptop look more attractive. But they could have done that already, and haven't.
Then they need an internet connection as well, most likely.
For kids who already have access to a computer, what aspect of the RPi is going to inspire them to program? It's just a linux box. They will use it to go on Facebook.
I don't want to be negative. Its small form factor could make it a more sophisticated arduino. Maybe some schools will use it to kit out a computer lab with a private network.
Good luck to them though. It could be a game changer in the third world.
Thrid world home computers
I think the third world market is pretty well sewn up at the moment with keyboard equipped "Famiclones" - a copy of an 8-bit Nintendo system with BASIC, word processor and some educational software in ROM. These things come in many (hundreds of) different shapes and sizes and are churned out in huge quantities by many Chinese factories, and seem to retail at about $10 to the end user. There is a volunteer project to improve the quality of the software that gets bundled with them called PlayPower: http://en.wikipedia.org/wiki/PlayPower
For a similar sort of outlay to the Raspberry Pi, in the first world one can buy a Chumby One/Insignia Infocast which is an ARM based SoC equipped console with 3.5" touchscreen, built in WiFi and USB host. It runs a Macromedia Flash implementation on top of the Linux kernel, but it has been designed to be easy to get a remote console and do more interesting stuff with. It's obviously not as powerful as the Pi though, and lacks it's more sophisticated video capabilities.
Most reasonably modern tv sets have hdmi. A cheap mouse plus keyboard plus usb hub can be had for (looks up skinflint... 1.5+2+1 =) less than a fiver if you're lucky. Might already have a spare hdmi and/or usb cable, recent phones already get charged through mini-usb and thus come with a suitable wall-wart, and perhaps you can filch the old card plus microsd-to-sd adapter from a friendly nearby member of the family who recently upgraded the microsd in his phone. There's quite a bit of likely leeway there.
The big attraction is that it's not the family computer, but something you can drop in your pocket and use wherever there's a suitable telly to be monopolised. It's a great excuse to be doing "something educational" instead of watching cartoons or something, so sellable to the parents too. Schools need only bolt hdmi-equipped screens and usb keyboards+mice to the desks in the classroom (might even see them back integrated, how's that for a throwback?) and the kids will provide their own computing power and storage. It's theirs, they take it with them, and it'll work the same wherever they are. Newer versions will be brought in with newer classes, saving a lot of trouble upgrading -- presumably hdmi and usb will last a while.
Honestly, I think it's underfunded schools and the underclasses 'round here that'll benefit much more easily than the throwaway charitable goal of "the third world". You'd have to build a power grid there first, for one, and the tellies will be CRT type and will not have hdmi, for another. In fact power even to recharge mobiles there is often a couple hours walking away. An olpc with a humongous battery is what you'd need there. Forget about the third world. This thing is great for here, for getting more kids hooked on making things happen themselves, and through that get them interested in engineering. Which pretty much was the goal.
"I think the third world market is pretty well sewn up at the moment with keyboard equipped "Famiclones" - a copy of an 8-bit Nintendo system with BASIC, word processor and some educational software in ROM. "
"a Chumby One/Insignia Infocast which is an ARM based SoC equipped console with 3.5" touchscreen, built in WiFi and USB host."
Fascinating. This is (I suspect) a world most Reg readers know nothing about.
Besides, it has composite out for the CRTs, too.
But will the DoE be interested?
Can't wait to purchase one of these myself, but I worry that too many vested interests within Governemnt will simpy ignore. Today the coalition is talking about improving IT education and moving away from simply teaching kids how to open a spreadsheet but I am guessing this will just mean millions of pounds being spent on more Microsoft and Dell products as their lobbyists move in.
I think all of us here on these forums need to get information on this product into the face of their local mp's and make this a party political agenda before civil servants kill any chance the device has in education
Hobbyist machine, not a revolution, ffs
Come on people, this is 2011, not 1981. The "nightmare of frameworks and compilers" is what allows 90% of developers to earn a living programming. They neither need nor want to run "on the metal", or write "tight code". And neither should they; the democratisation of software authorship is predicated on machines having the power to run high-level systems and languages. This idea goes against that, and harks back to a time 30 years ago when only nerds programmed. It's a piece of nostalgia, plain and simple.
In case you guys haven't noticed, the marketability of low-level coding skills is diminishing year on year, and has been for over a decade. Sorry, but you don't expand the workforce to become coders by propagating an 80s myth about the level of elite skills required for the job, and going on about "coding to the metal" or how we did things when we were lads. Jesus. You may as well try to expand mining in this country by teaching kids to swing a pick-axe.
It's nice to have a cheap Linux board, but it's not exactly a revolutionary idea, and if cheap Linux boards are so desirable why weren't schools buying them in their millions at the £60 price point? We're being asked to believe this is a revolution, but the truth of that has to be based on there being this immense price sensitivity in the market, which doesn't seem likely given that the prices of bare-board systems have been dropping steadily for the last 10 years and they are still just hobbyist items. Is the next 50% discount really going to make them suddenly go mainstream?
I can't be the only person who thinks the hype around this is ridiculous, can I? Computers are all about SOFTWARE, it is software which makes them flexible, it is software where the cool stuff is happening that's relevant to a maturing market, and that software is becoming better because it leans on other software. And yet everyone creams their pants when a new piece of hardware comes out that is identical to a hundred other pieces of hardware, but at a somewhat lower price point. It's staggering.
Expecting to be downvoted for this, don't worry :) You do what you have to do :)
Way to miss the point
In school they are NOT ALLOWED to program the machine in case they bugger them up. They are an expensive resource that need to be maintained.
This device give them the ability to bugger it up to their hearts content. You just need to re-image the SD card and you are back to square one. No IT support needed. And even if you somehow break the hardware - its $35 so not going to break the bank.
You are right - it is just another Linux board. But it's CHEAP and therefore more accessible. But because its JALB (c) it can run almost any language you want for free. Unlike all those school computers running Windows. And there is the software angle. Yes, its all down to the software, but you need a device to write software on, and this is a good one.
And I think you are wrong about the market for low level coding skills declining. We are constantly needing low level coders, and there simply are not enough good ones. There is a constant demand for software for low level devices as so much stuff has some sort of SW requirement nowadays. There is also the fact that a good low level coder is generally better at high level coding and debugging because they have a better understanding of how the device works under the skin. A low level coder can work at the high level. Not the other way round.
of only teaching high level coding is bloatware; and don't we all love that!
"Come on people, this is 2011, not 1981. The "nightmare of frameworks and compilers" is what allows 90% of developers to earn a living programming."
You don't become a successful civil engineer or architect by being ignorant of basic materials science. Knowing the acceptable loading factors and stress ranges is kind of crucial: get it wrong and your building will fall down. You can delegate this low-level work once you've reached the heady heights of building major skyscrapers, but understanding the basic foundational elements of your line of work is most emphatically not optional.
Programming is the same: there's a big difference between knowing a handful of APIs, and knowing how a computer actually *works*. And the only certain way to learn the latter is to try doing it. In a low-level language like assembly language, or C. (I'm not aware of any CPUs that can do OOP at the assembly language level, so procedural it is.)
Knowing how a computer "thinks" is the only way to really understand programming at a deep level. It helps no end in debugging. It also comes in handy if you have to code for constrained platforms and embedded devices, as well as in situations where optimisation skills are useful. (E.g. in game development.)
If you'd suggested an environment like Unity (www.unity3d.com), I'd agree with you, but JS in a *browser*? Seriously?
"It's nice to have a cheap Linux board, but it's not exactly a revolutionary idea, and if cheap Linux boards are so desirable why weren't schools buying them in their millions at the £60 price point?"
Even at £20 you still need a display unless you're doing the kind of low-level stuff that you've already told us is unnecessary. So the cost is dominated by the display, anyway. Of course, there are likely to be lots of schools who are awash with kit and who would perhaps welcome this kind of thing as a discount, but then there are plenty of other reasons why they wouldn't go for it: demand that "business skills" be taught is the primary one, and the one which this initiative is trying to overturn. But apart from schools, this initiative is trying to make stuff more affordable for home use. There, with an existing display, a drop from £60 to £20 is quite something.
"Here's an idea - if you want kids to learn programming, demand they are given programming courses at school, on the existing kit."
This is already going on. Unfortunately, the environments are evolving to *avoid* teaching the basics. Not having to worry about how you represent numbers, say, isn't such a bad thing - we all prefer working with abstractions instead of worrying about the details - but there's a tendency for "junior programming" environments to become nothing more than animation software, and then the fundamental skills aren't learned properly by the pupils. Fancy animations and trial and error fill in for actual understanding.
I kind of agree with what you're arguing, at least around the hardware. I think some of the people involved have a fetish for the microcomputer era (with all the "Model A" and "Model B" non-originality), but then I think that they betray the spirit of that era by wanting to deliver a tools platform with many of the innards cordoned off: read the Partis thread with Alan Cox's comments for an indication of this. If they don't even expose the same stuff that you could play with in the early ARM computers, where the average person didn't even get to tinker as much as they did on the Beeb, then your arguments are strengthened somewhat and this looks a bit more like a vanity project.
The display can be a junk (donated) CRT TV, though, so not exactly expensive.
At least in the US, free or very cheap CRT TVs are a dime a dozen, because of the digital TV mandate here.
NB a hardware hacker's board
It's a board with a collection of digital I/O lines that you can interface homebrew electronics to. A PC is notably lacking in this department. You used to be able to bodge some things onto a parallel port, but now PCs don't have a parallel port. Otherwise you choose between USB (too complex) or buying a digital I/O card (not cheap, especially if it's a notebook PC rather than a desktop).
Schools may or may not be keen to let their students do real programming, but they surely draw the line short of attacking a PC's motherboard with a soldering iron!
Absolutely great but...
Absolutely great. The price is amazing. An Arduino board with a single attachment cost more than this. The down side of course compared to the Arduino is its size.
Definently getting one if they maintain that price.
I do agree with some above posts that put into doubt how effective this really will be in the classroom. It just isnt that easy to get started with programming now. For all its ills, BASIC which was designed to get you started in programming, was so much more accessable.
PS: COULD SOMEONE EXPLAIN WHY FOR A UK CHARITY THEY KEEP REFERING TO THE PRICE IN $??????
Its because all the parts are sourced in dollars, so pricing the part in $'s means no issues with exchange rate buggering your margins.
It'll run Linux and Python. In my opinion if someone doesn't take to Python like a duck to water, s/he'll never be a programmer. What more could you want?
(Java? Perl? Ruby? Occam? Fortran? C++? Not my choices for a first language, but those as well, and more).
$$$ because they are thinking globally. Outside the EC most people probably don't have the faintest idea what a £ sign denotes (assuming it even displays right!), and little more idea what the GBPUSD or GBP/local exchange rate is. Global trade works in US$. Even the BBC World Service converts prices to US$.
This reminds me of...
... the Acorn Microcomputer (AKA System 1) which I played about with in school learning a bit of 6502 machine code!
See http://speleotrove.com/acorn/ for details.
I think you mean 720MHz Mr Writer!
Both products are great in their own way, BeagleBone has a powerful general core with a NEON vector processor and lots of low-level I/O functionality that you'd normally have to use an intermediary chip for like PWM, ADC etc. The chip is commercially available and the design is simple and so easy to produce custom models. The unit is designed to be expanded.
RaspberryPi is cheap but difficult to reproduce for commercial purposes (cpu can only be purchased in whopping quantities and stacked memory is difficult to manufacture). On the other hand whilst the main CPU is a bit weak it has a very powerful graphics core + DSP and so some dedicated time on coding for it could yield surprising performance. Also has HDMI out. It's more of an Application processor like what you'd expect to find in a smartphone (was originally designed for camcorders).
Re: 1GHz BeagleBone?
Ah, you're right - not sure where I got 1GHz. Fixed.
We have something similar for an OU course I am doing. though thats more a hardware and software combo. Principle is the same. Something to cut your teeth on and learn some fundamentals that doesn't need any real support.. Often the simplest solutions are the best, and this is pretty simple.
"not much chance of directly poking registers in the hardware,"
This will put off 99% of owners
The results from the other 1% should be more interesting.
What's wrong with running an 80s machine in an emulator and learning that way?
This seems to be a great tool for those who want to learn low level development. But for teaching kids they just need to learn BASIC, Logo or some other language. Going into hardware access and embedded development is for the more advanced student.
You can do both!
It's a linux box, you can load any number of language compilers on to it. QT works, GTK etc. C/C++, Python, Lua, and any number of other languages available for Linux. (See the Wiki)
I doubt more than 1% of people buying the board will be interested in the HW access side of things. And perhaps some of the rest will eventually go bit lower.
My plan it to take over the eldest's BigTrac. 16 steps just ain't enough, I want wireless control and on board video streamed back to the PC. Bwahhhahahahahahah
Does the beagleboard need any other comment than
"the days of scamps typing in REPEAT:PRINT "Hello, world! ":UNTIL FALSE on a machine in a computer shop and legging it are long, long gone."
Hey, I went into a large electronics retailer a couple of months ago, found an unlocked Mac, opened a terminal and typed
cat /dev/zero > foo
and legged it*. But then I'm 25.
*well actually I nonchalantly sidled away so I wouldn't draw attention to myself :)
Just tried that ..
on my aging laptop and got a 1GB file in ~15 secs.
How fast can you sidle ?
I for one..
welcome the creators of our robotic masters
All this talk of scampish behaviour in computer shops reminds me of when I typed a few lines of 6502 assembler to print "Rosie was here" (or something similar) to execute upon hitting the Beebs ctrl-break/break keys. Oh how we laughed as the Brent Cross WHSmith geezer had to powerdown the display to get rid!
I pray that the Pi facilitates such precious learning. I'll do my best to promote it - never did me any 'arm.
Now I don't pretend to understand all the tech in this thing
which is the point this project intends to address, but when these become available I'm going to have a word with the IT teacher at my son's school and see if he'd like enough of these for a class.
If so, either I'll pony up what I can, and scrounge the rest from other parents and the PTA, or I'll suggest a "buy two, get one" scheme where hopefully enough parents (and it only takes 30 or so out of a whole school to sort out the IT class room unless you've severe teacher : pupil ratio issues) can come up with the £44 required, a whole class is equipped and those who keep the 2nd one at home can see what their little darlings have learned when they come home and demonstrate. Its certainly better than them just learning "how to use Office".
Perhaps other readers might consider something similar for any school, youth group etc that could provide the necessary tuition, and with which they have an affiliation. Or even (careful now) if they dont.
Been following this for a while
And I'm really looking forward to it, will be picking up a few I hope to try tinkering with electronics again - at that price point I'm not too worried about my shitty soldering skills cooking something.
But the really nice bit is their charity status - the profits from the ones I buy help get more of these into schools.
Along with everybody on the reg of course I'm going to get one - and I have a few arduinos lying around.
I want one. I would put an OS on it that can handle emulators, and some sort of easy-to-navigate-with-a-controller interface, and every emulator and ROM that I can imagine, so that I can finally play things like GoldenEye64 on the TV again. (not to mention all the Mario and Sonic I can get)
We'll just say that I'm living in a nation who's copyright laws allow this sort of thing.
Low cost cluster?
At US$35 a pop you could build a respectable (low performance) cluster to experiment on.
Hadn't thought of that. And you've got hardware I/O lines to play around with for low-latency non-packetized inter-board communications. And you won't need air-con, even if you are playing around with many tens of them costing no more than a workstation-class PC.
Fun with DIY computer architecture research? Or massively-multi-monitor VR walls?
not-for-profit cutout of Broadcom
Follow the money.. Founder of RaspberryPi, Eben Upton, told slashdot..
"our dream scenario is that someone in China decides to copy our design and start knocking out millions of clones. Remember we’re a not-for-profit organization under English law, and all our trustees have other jobs".
Indeed! For his day job, Eben in an executive at Broadcom. The firmly closed source Corporation will be in a dream situation if Chinese cloners take the bait.
There are plenty of better development boards than this one. Some may not be quite so cheap, like the boards based on the MIPS-compatible Loongson CPU, but you do get you what pay for.
P.S. It's distinctly bad P.R. (and poor etiquette) to close a forum thread rather than use the focus to defend your decisions. Is that going to happen every time someone questions the RaspberryPi Foundation?