Placeholder Image

Subtitles section Play video

  • ADAM BEBERG: My name's Adam Beberg.

  • I work here at Google.

  • I'm very excited about the Raspberry Pi and happy to

  • introduce it here.

  • So 30 years ago, I learned Apple Logo and we had the

  • Apple Turtle, which is a little bowl-like thing.

  • And you tell it forward 10 and it moves 10 units forward, and

  • you tell it right 90 and it goes right

  • 90, and other things.

  • And that was an amazing experience for me as a kid,

  • because it was a robot.

  • You could see the motors in there and little servos, the

  • serial ports and all the fun things.

  • And that got me really excited about computers, and that was

  • kind of what started me on, well, the

  • path that I'm on now.

  • So I'm really excited about the Raspberry Pi and the

  • potential it has for education in my own kids and other kids,

  • and getting people to understand what is involved in

  • a computer.

  • I mean, you look at a tablet or a phone, it's pretty much

  • literally a black box.

  • So I'm very excited to welcome Rob Bishop.

  • He's one of the early engineers in the Raspberry Pi

  • Foundation.

  • And he's going to tell us all about the Pi and the

  • foundation.

  • ROB BISHOP: Hey, guys.

  • So yeah, I'm Rob Bishop.

  • I'm here over from the UK touring a number of computer

  • science departments and hackspaces, talking about the

  • Raspberry Pi, talking to the community, seeing what

  • projects are being made, making links to see what we

  • can do for education.

  • So as was introduced, I'm one of the earlier engineers for

  • the foundation.

  • I'm one of very few developers we have

  • supporting this project.

  • Talking about the foundation.

  • We're a registered charity back in the UK.

  • We're a not-for-profit organization.

  • And at the moment we have no paid employees.

  • So when you think, we currently have half a million

  • of these devices in the world and there's maybe five or six

  • engineers supporting this in their own time, that's

  • probably why we're probably a bit slow on responding to some

  • issues on GitHub, right?

  • So what is the Raspberry Pi?

  • For people who've just turned up and know nothing about it,

  • essentially it's a credit card computer.

  • It's a Unix box with GPIO and an HDMI out which is $35.

  • I mean, that alone kind of sells it to a lot of the

  • hacker crowd, sells it to a lot of people in here.

  • The point is that cool robot thing you've always wanted to

  • make but you couldn't quite justify spending hundreds of

  • dollars on the brain to go and do it, now you can go do your

  • Unix development, now you can make physical

  • computing for $35.

  • And before Christmas, we're going to release the Model A--

  • this is the Model B, named after the

  • tradition of the BBC Micro--

  • which is going to be the same board but without the

  • networking chip and connector, and that's going to be $25.

  • So why do we produce this?

  • How come a bunch of engineers are giving up their evenings

  • and weekends to come and make this thing?

  • And why am I here talking to you about it today?

  • So it all started when the founder, Eben Upton, was

  • working as the director of undergraduate studies for

  • computer science at Saint John's College in Cambridge.

  • And while he was there, he realized that the quality of

  • the candidates he had apply for computer science was

  • dropping on a year-by-year basis.

  • And he's getting fewer candidates with lower skills.

  • And he was really concerned as to why this was.

  • And he realized that it's bad for our industry, it's bad for

  • our economy.

  • It's bad for geeks like me to have toys to play with if we

  • don't have people who are talented enough

  • to go and make them.

  • And it's something that resonated with him.

  • Now he's working at Broadcom.

  • He finds it very hard to hire good engineers.

  • And the problem is because we're sort of suffering from

  • people not growing up with the low-level skills

  • that he grew up with.

  • So why doesn't my generation have those skills?

  • Well, if you think about it, my generation grew up with two

  • kinds of computing device.

  • We grew up with a games console and we grew up with a

  • shared Windows home PC.

  • Now starting off with the games console-- that's a

  • phenomenally advanced bit of kit, right?

  • The silicon in there, in terms of FLOPS performance, is

  • better than the 1990s' Cray supercomputer.

  • You think, like, DARPA and NSA could build

  • supercomputers out of PS3s.

  • The problem is, is that they're completely closed.

  • And as an educational device, they're a complete dead-end.

  • Whereas Eben, if he wanted to go play a game, would go to

  • the newsagent's, buy a computing magazine, flick

  • through the pages till he found a game he liked, go home

  • and actually type in the source code that was on the

  • page before he could play it.

  • We just get downloadable content for "Call of Duty" and

  • go play with phenomenally advanced graphics engines, but

  • we can't even see the source code for that even

  • if we wanted to.

  • You don't sell a games console based on what silicon's in it.

  • You don't sell it based on what FLOPS performance is.

  • You sell it based on the games titles and the

  • kind of closed packaging.

  • And the problem is, as I say, that's a dead-end platform for

  • learning how to use.

  • So what's the other kind of device we grew up with?

  • I mean, when we grew up back in the UK, we did these things

  • ICT lessons, Information Communication Technologies.

  • It's kind of rare to find a school that teaches computing.

  • Certainly I went to a good private school back in the UK.

  • But there weren't any computing lessons available,

  • even if you wanted them.

  • And these lessons were essentially sales pitches for

  • Microsoft Office products, right?

  • I mean, does my generation really need to be taught at 16

  • how to go and use Microsoft Word?

  • It kind of doesn't make any sense.

  • And the thing is, while it's great that we have this

  • proliferation of computing devices, while nearly every

  • home now has these Windows PCs, there's a barrier--

  • either an effort or a cost barrier-- to going and

  • developing on them.

  • If you want to go develop on your Windows PC, you have to

  • go and find the development tools.

  • You've got to go and have source Visual Studio and

  • either download it for free as a student or

  • go and pay for it.

  • You've got to go make the effort to go to be able to

  • develop on it.

  • And you've got to invest the time and the money

  • to go and do that.

  • Whereas alternatively, you could open up Chrome and go,

  • [INAUDIBLE].

  • Right?

  • And this is a problem.

  • In this instant gratification society that we grew up in,

  • when there's no barrier to content consumption and there

  • is a barrier to content creation, no wonder we're

  • having a sort of Lost Generation of people who

  • didn't bother becoming hackers because there was HD video

  • content that they could easily get to.

  • And what we realized is that, sort of taking a step back,

  • when I grew up, talking about hacking on PCs, I remember

  • taking apart my parents' computer, their Windows box,

  • and getting shouted at by my dad because he needed to check

  • his email and the motherboard's

  • on the floor, right?

  • It's no good having these machines and saying, let's

  • tinker, let's hack, let's make, if you need to then use

  • them to do your

  • word-processing for your homework.

  • And it's no good having these game consoles that are

  • incredibly advanced if when you do go and hack it and run

  • Linux on it, you get sued by Sony Corporation.

  • So what we realized was what we needed was another device.

  • We needed an additional device that removed the abstraction,

  • that removed the barriers, and that was just there as a toy

  • for tinkering.

  • We wanted something you could switch out your Xbox 360 with,

  • put in a low cost barrier, a low effort barrier, and it was

  • just there.

  • It was straight in with the development tools.

  • It was straight in to go and make things happen, to go and

  • build robots.

  • And we realized there was two keys to this.

  • The first was price.

  • It needed to be cheap enough that even parents who didn't

  • understand computing, they were happy to buy it just

  • because it was cheap enough.

  • And then we needed to make sure the kids had ownership.

  • We needed to make sure that this was a

  • device just for tinkering.

  • It was their device.

  • They didn't have to worry about whether or

  • not they broke it.

  • They didn't have to worry about whether or not it was

  • still in a usable state to go and do their homework on it.

  • This was a machine for play.

  • It was a machine for hacking.

  • And so Eben always had this dream, but it was only when he

  • was working at Broadcom developing processes for

  • mobile phones that he realized that actually, we have the

  • technology at the price point we needed to

  • go and fulfill this.

  • And so essentially what happened is we took the

  • development board that was being used for the Broadcom

  • applications process on this board and turned it into this,

  • turned it into the Raspberry Pi.

  • And that's essentially where this came from.

  • In many ways, this is a cell phone without the base band,

  • without the radio.

  • And the idea is that this has HDMI connections, it has

  • components.

  • You could hook this up to your CRT TV.

  • You can put it in place of your Xbox 360.

  • And all you need is an SD card out of your camera you might

  • have lying around, keyboard and mouse from an old PC you

  • might have junked, and a micro-USB charger you might

  • already have for your BlackBerry.

  • These are things that are just lying around.

  • So you just need this $35 investment to go and have a

  • toy to play with.

  • What's great about this device is it's accessible but not

  • necessarily easy to use.

  • The point is that rather than covering all the kernel

  • booting on some kind of nice graphic, we print out all the

  • steps on the kernel booting.

  • Because what we kind of hope is that

  • kids will ask questions.

  • You know, we think you learn by seeing.

  • You learn by asking questions.

  • You learn by being inspired, wanting to make things and

  • having to overcome obstacles in that goal.

  • When you boot this up it goes into command prompt.

  • If you want a GUI, you have to launch a GUI.

  • And we don't have a nice sugarcoated button saying, you

  • know, Launch GUI.

  • You have to type in startx. x.

  • And then we get these 10-year-old kids going, well,

  • what does startx mean?

  • It's like, well, you're starting an X server.

  • And they're like, well, what's an X server?

  • You're like, well, actually, you know, this is how

  • operating systems really work.

  • The Start button isn't an integral part of your

  • operating system, contrary to popular belief.

  • And so the point is that it's all accessible, it's all

  • immediate, and it's a great platform for developing.

  • So where are we going as a foundation?

  • So I talk about the fact that we're very interested in

  • education outreach.

  • But what we realize is that ultimately, we're a bunch of

  • engineers, a bunch of low-level software kernel

  • hackers, ASIC engineers.

  • We produced this chip.

  • And what we wanted to do was make sure that we made the

  • tools for the educators, we made the tools for the

  • outreach projects already there, to go and write

  • resources for, to go and teach with, so that they had a cheap

  • platform to go and do it.

  • And so why am I here?

  • Why am I talking to you guys?

  • What I'm saying to you guys is we went and made a Unix box

  • that's $35.

  • We went and made a computer but you can buy for $35, that

  • you can give to your kids.

  • It comes pre-loaded with Scratch.

  • It comes pre-loaded with Python.

  • It's a great learning platform.

  • We have GPI out for physical computing.

  • And what we're saying is, please go

  • and do awesome stuff.

  • And please help us get these in the hands of kids.

  • And please help us teach.

  • We're not yet ready to go into schools and say, this is a

  • finished product that you can put in your

  • schools and teach with.

  • And what we need is your help to polish the OS, to refine

  • the various bugs, and to make those resources so that we do

  • get to a point where we can go to schools and say, hey, we

  • have a whole computing package for you that's $35.

  • Here's some free resources.

  • Here's some case studies by dedicated teachers that have

  • already been using it.

  • This is the tool you need to go and do that.

  • So how do we see education working with the Pi?

  • If you're out here and have kids, you have cousins,

  • siblings, and you're like, yeah, I'm inspired, I want to

  • go teach some stuff.

  • How can I do that?

  • So firstly, we really like Scratch.

  • We think Scratch is a great way to get kids introduced to

  • programming.

  • One of the things we like about Scratch is that it's

  • teaching data flow, it's teaching algorithmic

  • development, without ever needing to say those words.

  • It's a graphical programming language.

  • They're dragging and dropping control blocks.

  • We'll probably have a demo next door or

  • possibly pull it up.

  • And you're creating short programs, you're making things

  • happen, just by dragging and dropping these boxes.

  • And we've seen, like, seven, eight, nine-year-olds make

  • games for the first time using Scratch.

  • And the point is, they don't really understand the computer

  • science behind it.

  • But when they've grown up around computing devices,

  • where all they've ever done is consume apps made by other

  • people, the joy they have in showing their brothers and

  • sisters a game that they made, that's really awesome.

  • Ultimately, programming computing is a creative tool.

  • I know it's tempting, sort of academics among us, to say,

  • you know, we want to optimize things for the sake of

  • optimization.

  • We want to research things for the sake of science.

  • But ultimately, it's a tool.

  • And it's a tool for creativity.

  • The best engineers are lazy people, right?

  • It's a way that we can go and do things that we might not

  • ordinarily be able to do very quickly, very easily.

  • And we want to make sure that it's not just--

  • it's not just the people who know

  • they want to be engineers.

  • It's not just the STEM students.

  • It's anyone who had a crazy idea to go and make a robot.

  • Anyone who wants to go and fire NERF guns remotely.

  • I mean, we've seen some great projects.

  • When I was over in New York, I met a videographer from Milan

  • who was there covering New York Fashion Week.

  • She came to the Raspberry Pi talk because she wanted to use

  • the Raspberry Pi to show videos she'd recorded of

  • catwalks, runway stuff.

  • And that's awesome.

  • We met some, at NY Resistor, we met some people who'd made

  • this huge tent that had a 512-point FFT around the tent

  • on LEDs that they took to Burning Man and made as a

  • dance tent.

  • I'm willing to bet most people dancing in that tent didn't

  • really know what an FFT was, but the point is, it's cool.

  • It's a toy.

  • It's a way to go and do awesome stuff.

  • And we think, if we're going to inspire kids, the way to do

  • it is to go and make these cool projects, show them the

  • cool projects, and then get them to want to learn so that

  • they can replicate them.

  • If we go and teach programming for the sake of programming,

  • we go say, yeah, it's important.

  • You should learn this.

  • Yeah, this is good for science.

  • That's not going to be as effective as saying, dude,

  • this is a robot.

  • We made it using programming.

  • You know?

  • That's the way that we make stuff happen.

  • I mean, someone over, early on in tour, at Maker Bar, I

  • believe, made a wearable computing set-up for, like,

  • under a couple of hundred dollars, just by using

  • Raspberry Pi, a Wi-Fi adapter, a small

  • display, and a coat hanger.

  • Sort of like a very cheap Google Glass, right?

  • And that's really cool.

  • It's cool that we can go and do that stuff, that when we

  • were all growing up, like, we wanted to do, but you know,

  • maybe couldn't justify it to our beer budgets to go and buy

  • the toys we'd need to make the things.

  • So let's say-- so back to the learning.

  • So let's say we've inspired them with Scratch.

  • They've made their games.

  • They've shown their cousins, and they're,

  • like, that's awesome.

  • And what they want to do now is they want to go make some

  • motors move.

  • They want to light some LEDs.

  • They want to go and make a robot.

  • We really like, as a first programming

  • language, we like Python.

  • We like Python 'cause it's a great language

  • to get stuff done.

  • It's human-readable.

  • If you want to go do "Hello, World," it's one line, as

  • opposed to like 600 in Java, right?

  • [LAUGHTER]

  • It's a great language for just getting stuff done.

  • It's why the scientific community uses it.

  • And what we like about Python is the fact that you couldn't

  • introduce kids to it just using it through the

  • interpreter.

  • They can use it as their desktop calculator, right?

  • They can just do their maths homework on it.

  • They can write simple lines.

  • It's one line for "Hello, World." With our libraries for

  • GPIO, it's one line to go and turn on the motor.

  • It's one line to turn on an LED.

  • And if you want to go use JSON to make something happen as a

  • result of someone tweeting the word "Raspberry Pi," that's

  • still only a few lines, right?

  • That's the joy of Python.

  • And what we see is once you've introduced them to syntax,

  • they've had that immediate success of typing something

  • and seeing something happen, you can then put those lines

  • together and compile them.

  • And that's your first procedural program.

  • And there you've written a program.

  • You wrote some code, you made something happen--

  • that's awesome.

  • And then obviously as people know, Python's an

  • object-oriented language.

  • The best way to write Python is object-orientated.

  • But rather than learning Java, where you kind of need to go

  • read those textbooks before you even go write your "Hello,

  • World," you've already picked up the syntax.

  • You've already picked up data flow and algorithmic

  • development from Scratch.

  • You've already gone and compiled your first program.

  • You're in a good place to go and learn about

  • object-orientated methodology, learn about class hierarchies.

  • And then that's a good point to go and write your

  • object-oriented code.

  • Once you've done that, stepping over to Java or

  • something's pretty easy, right?

  • Because you know how do to that kind of design.

  • It's just another set of syntax.

  • We also really like teaching the low level.

  • So one of the things we've done is Cambridge University

  • Computer Science Laboratory have given one of these to

  • every fresher coming into this year's set of

  • computer science undergrad.

  • And this is really great for two reasons.

  • The first is that we can hopefully see a whole load of

  • projects by undergraduates wanting to prove themselves,

  • make a name for themselves, and going and making awesome

  • stuff just because, you know, they have time and they want

  • to go do it.

  • But also because it means that the academics are going to

  • start writing teaching material that's

  • tailored to the Pi.

  • So there's already a course out there called Baking Pi

  • that's-- yeah, great name--

  • that's made by the academics for a computer science

  • laboratory.

  • And that's a course on how to write your own operating

  • system in assembler.

  • Like complete with frame buffer.

  • And when you think, you know, one of the problems is that

  • we're lacking those low-level skills--

  • my generation didn't grow up on BASIC.

  • We didn't grow up on Spectrums.

  • We didn't do command-line stuff.

  • We didn't learn machine code.

  • And so the point is that you can go and write an operating

  • system in assembler.

  • When you then go move on to C and help us with kernel

  • development, your C is going to be a lot better, having

  • written an operating system in assembler first than going the

  • other way around, right?

  • And we think this is great for not just the

  • computer science learning.

  • We also think this is great for physical computing.

  • As I say, I'm a EE grad.

  • I think the best way to get people inspired in programming

  • is to show them stuff happen.

  • It's one thing to try and teach them why some

  • optimization or some bit of code's cool.

  • It's another thing to show them something happening and

  • going, that worked because we had a computer and we had some

  • code to make it do something.

  • So really I'm here asking you guys to please keep making

  • cool projects.

  • Please let us know the cool projects you're making.

  • Helps inspire more kids.

  • Help us get these in the hands of kids.

  • Show the kids you know, your cousins, you kids and things,

  • and how to go program in Scratch.

  • Introduce them to Python.

  • Introduce them to the joys of physical computing.

  • And also, it'd be great if there's

  • any educational outreach--

  • I know Google does lots of education outreach out there--

  • if you could help us write material for the Raspberry Pi,

  • help us get the lesson plans and the structure we need in

  • place so that we are ready to go to schools and say, here's

  • a complete package at a low cost.

  • This is the way we think you should be teaching computing

  • and introducing computing in schools.

  • I'll probably hand over to a Q&A now.

  • I can answer more technical questions.

  • I know these kind of talks, we get a wide variety from

  • teachers turning up, saying, I've heard about this

  • Raspberry Pi thing, what's in it for me?

  • Through to guys saying, you know, why doesn't my

  • particular bit of split-transactional USB work?

  • So we can hopefully try and cover that range.

  • And I'll try an answer what I can.

  • Awesome.

  • AUDIENCE: So thanks for the work you've done so far.

  • So I'm working on an open-source project, and we

  • honestly can't get our hands on enough of these things.

  • People want them.

  • One thing we have had a lot of problems

  • with is the USB stack.

  • ROB BISHOP: Yeah.

  • AUDIENCE: We have a bunch of USB interfaces that vary

  • between, like, locking the device up and rebooting it and

  • all the rest of it.

  • Do you know where people are on the USB stuff?

  • ROB BISHOP: Yeah.

  • So initially we had a problem whereby the endpoints in the

  • microframes were--

  • the allocation was fixed.

  • So you're limited to seven endpoints you could

  • service in one frame.

  • And that meant the first few devices which used up those

  • endpoints--

  • bearing in mind that one goes to bolt transfer anyway, one's

  • used up for the networking, and then most devices have two

  • or three endpoints anyway, which meant that if you were

  • using multiple USB devices, that didn't work.

  • We've now got pushed out a microframe scheduler fix that

  • dynamically allocates endpoints.

  • So we can now service lots of devices.

  • We've also pushed out a fix with the interrupt masking so

  • that we service the USB first.

  • And we've reduced the CPU overhead that was incurred in

  • doing that.

  • We're still actively working on USB.

  • Any engineers in here who have worked on USB, that's probably

  • the hardest thing we've had to develop for this.

  • You need these 100K analyzers to go and work on it.

  • And it's a systems problem, because you've got to

  • understand everything from the state machine and the RTL

  • through to using a logic analyzer to see what's going

  • on on the frames.

  • Through to the other problem with USB, is the ubiquity.

  • So the problem is that the general perception is with USB

  • devices is that they should just work.

  • You know, they're USB, right?

  • But there's a very loose understanding in the industry

  • of what the USB spec actually says.

  • We've seen lots of devices that don't adhere to the spec

  • and don't work with us.

  • Or do adhere to the spec but don't act in a nice way.

  • So we had some issues with USB serial converters, and they

  • just kind of flooded us with packets.

  • They didn't perform in the way we expected.

  • So part of what we're trying to do now--

  • and as I say, we expected to ship 10,000 of these in our

  • first year.

  • We're probably going to ship a million in our first year.

  • We shipped half a million already.

  • On there's maybe five or six engineers working part--

  • well not even part-time, in their evenings and weekends--

  • on supporting this.

  • So we're working on it, but the USB is slow

  • because it's difficult.

  • And if--

  • I know some people say that one of the problems with the

  • foundation is we're not transparent enough in what

  • we're working on.

  • It's probably just 'cause, you know, these are all engineers

  • who are working at Broadcom on their day job, going home and

  • trying to fix USB in the evening.

  • They don't really have much time to go and write a blog

  • post as well.

  • You just kind of have to trust us that we want to see this as

  • finessed as possible.

  • We want to get the best performance out of USB, best

  • performance out of the processor.

  • We're going to work on those things.

  • We're going to do our best.

  • We'll kind of push out updates as we do it.

  • But we'd rather do the development than necessarily

  • spend lots of time kind of blogging about it.

  • AUDIENCE: A little more transparency, though, might

  • get you a lot more help.

  • ROB BISHOP: Yeah.

  • I agree.

  • It's something we're working on.

  • We do recognize that.

  • And it's just something--

  • I mean, obviously, we're all engineers that, we're working

  • in a sort of corporate environment.

  • We're now working in an open-source project.

  • It's quite a big headspace switch to go move across to

  • doing everything sort of in a kind of structured way to

  • suddenly doing things in the way that you're asking the

  • community for feedback and you're getting involved.

  • I think part of the reasons we don't look transparent is just

  • because we're so overwhelmed.

  • We're desperately trying to do as much as we can and we're

  • spread so thinly that we don't have time, necessarily, to do

  • all the things we'd like to do.

  • But definitely, as things are stabilizing, as the foundation

  • actually starts to have engineers working on it

  • full-time, we're certainly going to move toward more

  • transparency.

  • One of the things we've done, we've opened a Twitter account

  • called @rpf_dev_updates.

  • It's linked to our GitHubs.

  • And what we do is every time there's a commit, it get

  • tweeted onto there.

  • Also every time we push anything new into the

  • repository, every time we push an update to the firmware, we

  • tweet about it.

  • And that's just a really quick way so that the developers

  • among us can kind of keep track easily on what's going

  • on without necessarily having to go to GitHub and just see

  • what the development is.

  • But yes.

  • It is something we're working on.

  • AUDIENCE: Are you guys thinking about any hardware

  • widgets to add on to this?

  • Or are you mostly focused on the software side?

  • ROB BISHOP: Yes.

  • So we can talk about that.

  • So I should have mention that in the talk, actually.

  • So the way the foundation works, ultimately--

  • Pete Lomas, who designed this board, did a great article in

  • "Wired" recently--

  • I don't know if any people read it--

  • which is where he was saying about, we have to sell out a

  • little to sell a lot.

  • We realized there was going to be a lot more demand than we

  • could ever raise capital to go and produce ourselves.

  • So what we decided to do was to go and approach a number of

  • multinational companies.

  • We went with RS, who trade as Allied in the US, and the

  • Farnell group, who have a number of business units--

  • Element 14, Newark, MCM, the have here in the US.

  • And we licensed them the design of the PCB so that they

  • could manufacture it for us and handle distribution.

  • As a result, that meant that we couldn't release all

  • Gerbers and things on the outset, which we wanted to do.

  • The problem is that to get people to invest money in the

  • infrastructure and producing these, we have to make sure

  • they can protect that investment.

  • And so one of the things we talked about is we are very

  • much believers in the open ideology.

  • It's something we want to do.

  • We can honestly say this board is as open as possible.

  • If we could make this more open, we would be doing it.

  • There are things we are doing right now I can't necessarily

  • talk about to try and make it more open.

  • The problem is that what we thought was more important was

  • to ship than to sit around worrying about the ideology.

  • Ultimately, yes, the GPU on this is closed.

  • Yes, we haven't released the Gerbers.

  • But we can get UNIX boxes in the hands of kids for $35.

  • And that's our goal.

  • And the important thing was in doing that, was in delivering

  • that, and then kind of secondary, making sure we can

  • fulfill all our own personal beliefs on the openness

  • ideology and things.

  • And the other point is we're going to be in a much better

  • position when we've shipped a million or so units to go and

  • talk about the open debate than we would be if we were

  • sitting around saying, hey, we're just going to wait till

  • we can do an entirely open board for less than $35.

  • So as an engineer, I understand

  • the frustration there.

  • But ultimately, the point of this is to get cheap computing

  • devices in the hands of kids.

  • And we're going to make sure we do that first, and we're

  • going to make sure that if we can do that, we do that.

  • And we make it as open as possible, but you know, that

  • stuff will come.

  • AUDIENCE: I guess I was asking more about add-ons.

  • ROB BISHOP: So we'll talk about that.

  • So we are responsible for the design of this board.

  • We're responsible for the kernel.

  • Like that's what we do as a foundation.

  • There's then a lot of add-ons which

  • are made by the community.

  • So the first thing you'll notice is we

  • don't produce a case.

  • So there's a couple of reasons for that.

  • I mean, firstly, we're not graphic designers.

  • We're hardware engineers.

  • We're software engineers.

  • We thought it was cool to leave it open to the community

  • to go and do that for us.

  • Certainly with the rise of 3D printing, there's designs you

  • can go and download, go to your local hacker space, and

  • just print there.

  • And there's a variety of cases.

  • Like this one, we really like this one.

  • It's called the Pibow.

  • It's a multilayer case.

  • Comes with really nice faux IKEA instructions.

  • And these are made by the

  • community and openly available.

  • And it's great that we can kind of encourage the Maker

  • community and give them ways to sort of raise some of their

  • own finances by making projects that

  • they can sell alongside.

  • The other reason we don't have a case is because when I put

  • this in the hands of kids, they go, hey, what's that?

  • What's this bit do?

  • And that's awesome.

  • We've grown up in a generation where we think electronic

  • device have all kind of black slates with

  • rounded corners, right?

  • And it's important to say, no, no.

  • This is what a computer looks like.

  • This is a computer.

  • And it's great to be able to answer those questions.

  • It's great to be able to show them a PCB.

  • I mean, I quite like going up to CS grads and saying, hey,

  • can you name the capacitors on this board?

  • 'Cause you know, it's amazing how much we've lost that

  • knowledge of computing, that when we had Spectrums and

  • Apple IIs, that was a bit more well known.

  • There's also hardware made by Broadcom engineers.

  • So there's a guy called Gert van Loo who's a really smart

  • engineer, did a lot of the ASIC design for the sock

  • that's on this board.

  • And he wanted to go and make really big things move with

  • his Raspberry Pi.

  • This is 3V3 digital logic.

  • The current draw is obviously all powered

  • by the power supply.

  • So it's shared between the processor, the USB devices,

  • and the GPIO, so you're kind of limited there.

  • But he wanted to go make big motors move.

  • So he went, OK, I'll go design a board to make that happen.

  • So I have a thing called the Gertboard.

  • It's available from Newark.

  • It's $40.

  • It comes as a kit.

  • You can solder it together yourself.

  • It's not on sale yet, just because we're going through

  • the last few bits of FCC testing and things.

  • It'll be on sale as soon as we can sell it.

  • And this allows you to drive things up to 4 amps with

  • 5-volt logic.

  • So you can just slip this is in the same places as you

  • would do your Arduino sensors.

  • It's the same 5-volt GPIO.

  • You can go make some great, huge motors move.

  • There's physical fuses on that.

  • That's kind of cool.

  • And it actually has an [INAUDIBLE] chip on it.

  • So if you want to go use your existing Arduino

  • microcontroller code, you can run it on this board.

  • And this isn't made by the foundation, but it's a part of

  • the fact that our mission is to focus on what we do well,

  • focus on what we can do for the community that might

  • otherwise struggle to raise the capital or get the

  • engineering talent to go and do.

  • We made the Unix boxes.

  • We're kind of letting the rest of the community come and help

  • us out with add-on gear.

  • Oh, and particularly, we really like my

  • friends over at Adafruit.

  • They have a learning website called learn.adafruit.com

  • where they have a whole bunch of tutorials and products that

  • have been well tested for the Pi, everything from GPS

  • receivers to GSM receivers, wireless keyboards, small

  • displays, all of these things which tutorials,

  • and they're for sale.

  • And what's great is that they were waiting for something

  • like this to come along.

  • They were waiting for someone to make hardware hacking

  • accessible and cheap enough.

  • And that's what we feel we've been able to go and do.

  • The one thing the foundation is going to produce, in terms

  • of add-on hardware, or certainly is going to in the

  • immediate future, is--

  • as I mentioned, this is basically a cell phone without

  • the base band and the radio.

  • The other thing most cell phones now have is a camera.

  • So we're going to release a camera board which works over

  • SPI data, I2C for control.

  • It's going to be $25.

  • It's going to hopefully be done before Christmas.

  • People often ask me, when's this going to be done.

  • I've got to go back and write the software for it.

  • So as soon as I finish that, it'll probably go on sale.

  • It's a 5-megapixel smartphone sensor.

  • And you're going to be able to record 1080p video with H264

  • encoding in the hardware.

  • You're going to be able to hopefully use the JPEG

  • hardware encoder to get quite good frames-per-second encoded

  • JPEGs off the 5-megapixel sensor.

  • We're also going to hopefully give a raw bitstream that you

  • can put across the network, put into OpenCV, go and do

  • cool robotics projects.

  • And that's all going to run on OpenMAX media streaming ware.

  • And that's all going to be in userland.

  • And talking about the open, we are wanting to open-source

  • everything that runs on the ARM, all of the userland.

  • We are trying to be as open as possible.

  • The problem is, as you guys will well appreciate, that

  • involves talking to a lot of lawyers, which is A, not very

  • fun, and B, very time-consuming.

  • But we're taking on that pain on your behalf, right?

  • So you should be very grateful.

  • [LAUGHTER]

  • So--

  • So yes.

  • Cool.

  • AUDIENCE: Hi.

  • Have you found that for the adults that are working with

  • kids using Raspberry Pi that it's better for a certain age

  • group or certain other age groups?

  • ROB BISHOP: So one of the great things about Raspberry

  • Pi is that with the help of things like Scratch, as I

  • said, we've seen seven-year-olds produce games.

  • So on our blog, there's actually some videos of some

  • games that some seven-year-olds have made,

  • that we saw and we just though, that's awesome.

  • This seven-year-old's made a game and is really

  • excited about it.

  • I quite often meet engineers who have been teaching their

  • kids, and you've got these kids under 10 who are so

  • excited that they made something themselves.

  • It's a game that they made.

  • When you think-- when you're seven, you have all these cool

  • games you want to make.

  • The fact that can actually go and produce something someone

  • else can play, that's awesome.

  • And it's the right way of encouraging programming,

  • because we're showing that it's a tool for creativity.

  • It's showing that there's a way that you can go make those

  • awesome things you want to do.

  • And I'm pretty sure that's the reason why

  • most of us are here.

  • We grew up making awesome stuff in LEGO.

  • We grew up wanting to build robots.

  • The point is, we've made something cheap enough and

  • accessible enough that the kids can go and do that.

  • And that's awesome.

  • Sort of a moving up from there, we've got

  • undergraduates using the assembler course to go and

  • write operating systems.

  • We have the kind of, you know, lifelong hackers making all

  • sorts of awesome stuff.

  • And Python kind of fits nicely in between for physical

  • computing all the way through to web development stuff.

  • And you can use all the web libraries for JSON stuff to go

  • and make web apps.

  • And it's great what the community's done with the

  • Raspberry Pi.

  • And one of the things that I was really excited about, back

  • at home there's a computing magazine and it had a review

  • of media centers.

  • And the Raspberry Pi was reviewed as a

  • media center option.

  • And you think, you know, we produced some

  • hardware for education.

  • The community's gone, hey, we can make a media

  • center out of this.

  • And they went and got XBMC polished enough that it was

  • good enough be reviewed in a commercial magazine as a

  • commercial product.

  • And we think that's awesome.

  • It just shows what the community can do.

  • I think one of the great things about this platform is

  • that if we do sell a million in our first year, there's

  • going to be that wealth of people making projects for it,

  • that wealth of people on the forums answering questions,

  • having user groups.

  • And that's going to be a really great

  • platform to learn on.

  • Because if there's something you want to do, someone's

  • probably already blogged about it, and that's really cool.

  • I've been touring hack spaces, because we really like

  • supporting hackspaces.

  • We think hackspaces are a great place where, you know,

  • artists can turn up and say, hey, I've always wanted to

  • make this ridiculously awesome thing for Burning Man, but I

  • have no idea how.

  • And then there's guys like us, saying, yeah, let's do it.

  • And you can share those skills and inspire by doing rather

  • than inspire by teaching and inspire by academia.

  • AUDIENCE: So I have a question.

  • When will I be able to order more than one Raspberry Pi?

  • I currently have only one.

  • ROB BISHOP: No, no, so you can order more than one right now.

  • Yeah.

  • So you can do that right now, if you want you.

  • So there's two manufacturers or distributors.

  • There's Allied and Farnell.

  • Farnell currently have a lot of stock in North America.

  • You can go and order from MCM Electronics.

  • And it's just shipping, it's 3-to-5-day shipping.

  • They have stock right now.

  • And there's no order limits.

  • So you can go do that now if you want to.

  • AUDIENCE: It takes two days to get here.

  • ROB BISHOP: OK.

  • Two days.

  • So there you go.

  • So you can have 100 in two days, probably.

  • My only worry is I'm going to give this talk at some point,

  • and then everyone's going to get their iPhones out, ordered

  • them, and then just give me no stock by the time

  • I finish the talk.

  • But that hasn't happened yet, right?

  • So hopefully there's still stock right now.

  • AUDIENCE: Do you think [INAUDIBLE]?

  • Because I ordered in June, and I still haven't got mine.

  • ROB BISHOP: Yeah, so the problem is that the chip on

  • this board is a custom ASIC, application specific

  • integrated circuit.

  • You can't buy it off the shelf.

  • You have to order them from Broadcom, an and that's got a

  • 23-week lead time.

  • So the problem is that once they sold out, it takes quite

  • a long time to get your orders through for the chips to go

  • and produce more boards.

  • So that's going to stabilize once we have a better

  • understanding of demand.

  • But as I say, we expected to sell 10,000 units, right?

  • I mean, we crashed [INAUDIBLE] our website

  • on the day of launch.

  • Both of our distributors, they-- you kind of see when we

  • launched on their share price.

  • [LAUGHTER]

  • ROB BISHOP: And so--

  • and so we've been kind of overwhelmed.

  • And they've been a bit overwhelmed.

  • We think before Christmas the stock situation should be

  • stabilized.

  • We should be a good place.

  • As I say, there's plenty of stock in North America through

  • Farnell right now.

  • So you can get them in two days.

  • AUDIENCE: Do you know about Arduino?

  • How do you feel--

  • are your audiences the same or different?

  • Are your ambitions the same or different?

  • ROB BISHOP: Yeah, so people quite often ask about

  • competition.

  • So we're not a start-up, where we went into this to get rich.

  • No one gets paid yet.

  • I'm probably going to be the first employee of the

  • foundation, paid employee.

  • Eben's wife Liz is full-time doing the PR at the moment,

  • doing the blog.

  • You read most of her postings if you go on the website.

  • But certainly as the first engineer when I get back.

  • And we didn't do this to get rich.

  • I don't have any equity right now.

  • I'm not being paid to do this.

  • We didn't say, hey, let's go do this thing.

  • Let's make a start-up.

  • We're a bunch of engineers who had the drive to go and make

  • something for education and had the facility, had the

  • technology to do it.

  • I think a lot of people say, why did you go

  • for a Broadcom chip?

  • And it wasn't that we sat down and said, hey,

  • let's produce a platform.

  • We had a platform, and we said, hey, this would be great

  • for that thing we've always wanted to do for cheap

  • computing for education.

  • It's that way around.

  • This is a start-up that was born out of necessity rather a

  • desire to go and have a start-up.

  • And I think with the competition--

  • we didn't, certainly as far as I'm aware, we didn't sit

  • around and go, where's our competition?

  • What's out there?

  • We just kind of went, we think this is a good thing to do.

  • We think we need this.

  • Let's go and do it.

  • We don't want to go and compete

  • with these other companies.

  • We think it's great other people are working in the

  • hardware space.

  • I mean, we often say, if someone was to come in and

  • produce a higher-performance board or a board that was

  • somehow better for the community that was cheaper,

  • that's great.

  • We'll go back to our day jobs, right?

  • Mission accomplished.

  • We're doing this because we think it should exist.

  • And we're hoping that we spawned it.

  • But if you read the tech blogs, you'll see the tech

  • blogs are full of "Raspberry Pi competitor," you know,

  • "Raspberry Pi-like devices." And we're kind of proud of

  • that, because the point is that we've shown that there's

  • volume in doing cheap computing devices.

  • We've shown that this is a device that people want.

  • And hopefully we can kind of get those being created.

  • And that's our goal.

  • Our goal's not to have a massively successful business.

  • Our goal is to get these in the hands of kids and to make

  • something like this.

  • And so we think there's still room for the Arduino.

  • I mean, the Arduino is a microcontroller.

  • It's a lot better platform for really cheap sensing projects,

  • so anything where you want a basic microcontroller.

  • But soon as you want anything with networking, as soon as

  • you want anything where the development's going to be

  • quicker in a Unix environment than it would be writing for

  • microprocessor, then this is where this really wins.

  • I believe an Arduino is a similar price.

  • But then by the time you buy the networking shield and get

  • all the stack working, that starts being

  • over $100, I believe.

  • That's what people have told me.

  • Whereas this is $35 with USB, with networking.

  • So that's where we see this being useful.

  • We see it as living alongside an ecosystem, not necessarily

  • being a replacement.

  • AUDIENCE: Could you talk us through a little bit?

  • If somebody wanted to make a simple project that just

  • controls a couple of motors on a robot or something, what's

  • involved to go from there up to that?

  • ROB BISHOP: So I think what you do is if you go buy one of

  • these from MCM you'll get a nice box--

  • I don't know if anyone has a box with them-- but a little

  • cardboard box with one of these devices in it.

  • Doesn't come with a power supply.

  • Doesn't come with an SD card.

  • It literally comes as a bare board.

  • So the first thing you're going to want to do is source

  • some kind of power supply.

  • We recommend one that's rated up to an amp, 5 volts.

  • One of the things we found is power supply manufacturers do

  • vary greatly, and we found at least one manufacturer that

  • sold a range of power supplies that just were all 7 volts,

  • regardless of what it said on the label.

  • And it is worth making sure that you can supply enough

  • current, because one of the problems we have is that

  • people plug power supplies in that maybe only give 500

  • milliamps or less.

  • And said it's enough for the board to boot, but soon as the

  • CPU load gets significant, soon as you plug-in your Wi-Fi

  • dongle, it restarts 'cause there's not enough power.

  • So Adafruit sell a 1-amp, 5-volt supply.

  • If not, just scout around.

  • We found the iPhone ones are pretty good.

  • So you can get them.

  • So get ahold of one of those.

  • Get ahold of an SD card.

  • Get ahold of an HDMI cable.

  • Plug it into your TV.

  • Yeah.

  • Yes, as we couldn't do today.

  • Apparently this TV only takes VGA, which is why I don't have

  • a demo behind me.

  • So that was a pretty poor example.

  • And it doesn't have component either.

  • Because I mean, the argument is that for the developing

  • world, you have component.

  • So if you want to kick one of these out for free, you go to

  • your local electronics recycling company and you say,

  • hey, the next time someone's chucking away a CRT, or

  • Freecycle or Craigslist or whatever the US equivalent is,

  • you can source CRT monitors for nothing these days.

  • People are trying to get rid of them.

  • The same with keyboards and mice.

  • You go to your local bank or whatever and say, hey, next

  • time you're reprovisioning your IT, you know, and you're

  • throwing away those perfectly good keyboards--

  • we believe you can stock these out for nothing with a little

  • bit of effort.

  • But yes.

  • So you get your SD card.

  • You go on our website.

  • We recommend an operating system called Raspbian.

  • So Raspbian's a fork of Debian.

  • Obviously we only forked because we absolutely had to,

  • you know, who wants unnecessary forks

  • of operating systems?

  • But Debian have a version which supports V7 instruction

  • set with hardware floating-point.

  • And they have one which is kind of the catch-all

  • distribution, which supports V4 with software

  • floating-point.

  • So the problem was because we're V6, we were losing all

  • of that performance by having the build for V4 with software

  • floating-point.

  • So the community went away--

  • we're very grateful to them for doing that--

  • and rebuilt Debian, rebuilt the packages, with hardware

  • floating-point with V6.

  • So you get significant performance increase.

  • So that's why we have this Raspbian.

  • But it's essentially Debian.

  • And so you can get a Debian disc image.

  • DD it onto your SD card.

  • Plug it in, boot it up, and hopefully this'll demo next

  • door since I don't have it up on here.

  • It boots up to command prompt.

  • It says, if you want to GUI, type startx.

  • You type startx.

  • LXD starts up.

  • You know, it's what most people would

  • recognize as a computer.

  • Scratch is pre-installed on image, so if you want to go

  • into Scratch, you go straight into Scratch.

  • Python's pre-installed, so you can open a Python terminal.

  • I mean, obviously, if you're doing Python development, we'd

  • say don't waste the CPU overhead of

  • going into the GUI.

  • Just go straight into it in terminal.

  • And that has a library which has already got

  • all the GPIO control.

  • So you go on a wiki, the eLinux wiki, you look up the

  • necessary Python lines.

  • It's all well-documented.

  • And it's, like, a line to go internal GPIO.

  • So it's 3V3.

  • So you go get a head of pin, get some breadboard, connect

  • it to something LED.

  • Connect it to ground on the connector.

  • Type your line of Python.

  • LED turns on.

  • So those are basically the steps.

  • But it is-- it's well set-up.

  • I mean, the great thing is, anything you want to do, if

  • you go and Google it, someone's probably done it or

  • done something similar.

  • There's all sorts of wrappers.

  • There's a wrapper called WiringPi which makes the GPIO

  • like the Arduino, I believe, makes it very simple.

  • But also, all of the GPIO and the LED control, they're all

  • mapped in the correct place in the Unix file system, so you

  • can do it with a bash script if you want to.

  • So I actually did a great workshop where we got a bunch

  • of complete beginners and we went and turned on an LED on a

  • breadboard using bash scripting.

  • And we did it not because that was the easiest way, but

  • because they actually understood what was going on.

  • We had a multimeter out, and we were also bash scripting.

  • How often do you get to do those two things in the same

  • project, right?

  • We were reading a resistor value and we were also

  • explaining what a pipe did.

  • That's awesome.

  • And that sort of teaching computing on the low level,

  • it's teaching computing by doing, and it's getting those

  • low-level skills, which we're losing, to people who aren't

  • bothering to learn assembler, aren't bothering to really

  • understand how a computer works.

  • AUDIENCE: You said you were planning to go into

  • schools with this.

  • Have you thought about how to measure the impact?

  • Or are you just going to throw it out there and see what

  • people do with it.

  • ROB BISHOP: Right now we're kind of throwing it out there.

  • So right now, we're targeting the STEM groups, the outreach

  • groups, who maybe are already doing stuff with Arduino,

  • maybe are already running electronics classes.

  • They don't need any lesson plans or resources from us.

  • They just hear they can get hold of one of these for $35

  • and that's all they need.

  • And so for those people, we're ready for you to go and be the

  • trailblazers, do the case studies, get it in the hands

  • of kids and see what they can make.

  • We also have another set of educators, which are the

  • teachers, saying I'm not an IT specialist, I'm not a

  • computing science specialist.

  • I want to teach computing.

  • I hear what you're saying.

  • But I don't know how to do it.

  • To those people we say, hold off yet.

  • We are working with big government groups.

  • We're working with educational groups to get those resources

  • made, but that takes time.

  • We're not arrogant enough to believe that because we

  • understand it we can teach it.

  • We want to talk to specialists, get

  • their skills involved.

  • And that's on its way.

  • I mean, the travesty would be if we pushed this hard into

  • schools generally now and sort of had them sitting getting

  • dusty on a shelf just because they were cheap and they

  • seemed like the in thing to be.

  • No, we're very much focused on making sure that we have a

  • package, we have a board and an operating system and a set

  • of resources that are polished enough, ready to do that.

  • We're not there yet, but that's our goal.

  • That's what we're working on.

  • Right now we're saying, let's go make those cool projects

  • which inspire people to do it.

  • Let's go and get them in the hands of kids like we probably

  • were when we were growing up, who didn't need a lesson plan.

  • We just needed to be able to have something

  • to go and play with.

  • And let's make sure that we get them out there, get them

  • tested, get the feedback, so when they do go into classes,

  • it's polished, we've got the inspiration, and they've

  • hopefully already seen cool things other people have made.

  • AUDIENCE: Power management and low-power [INAUDIBLE].

  • ROB BISHOP: Yeah.

  • AUDIENCE: I've tried to shut down this and it still seems

  • to absorb about one watt.

  • ROB BISHOP: Yeah.

  • So I believe, on the Model A-- as I said, the networking

  • consumes about 50% of the power consumption.

  • It's running at 700 megahertz on the ARM.

  • With the Model A, I believe with underclocking, you can

  • get down to 250 milliamps.

  • The update we've pushed out has this governor.

  • So you kind of have this historesis effect where we

  • monitor the CPU load and we step the core voltage and

  • frequency based on the CPU load.

  • And then we have another threshold determined by an

  • on-die temperature sensor, where we kind of ramp that

  • down again.

  • Which hopefully means that you get the extra performance when

  • you need it without needing to run overclocked all the time

  • unnecessarily.

  • And you don't need cooling for that.

  • Often, as soon as you say "overclocked," people go, but

  • there's no active or passive cooling.

  • It's like, there isn't in your smartphone.

  • You don't need it.

  • AUDIENCE: But on the Model B, the internal [INAUDIBLE],

  • they're going to--

  • ROB BISHOP: Yes.

  • So on the Model B, sadly you can't--

  • I don't believe you can turn it off in software.

  • So that's frustrating.

  • I mean, I don't see any reason why you couldn't desolder it.

  • But the Model A's going to be out before Christmas.

  • And you can underclock this board.

  • So you can go into settings and reduce the clock speed if

  • you wanted to and hopefully drop the power

  • consumption that way.

  • But yes, you're always going to have the overhead of the

  • networking on the Model B.

  • But the Model A is coming.

  • The problem is we can't produce enough Model B to

  • satisfy demand, so we haven't done the Model A yet.

  • But it's on its way.

  • AUDIENCE: What are your plans for improvement?

  • I mean, what are you working on improving?

  • Maybe reducing the size or the speed of the processor?

  • ROB BISHOP: Yeah, we're committed to continuous

  • improvement.

  • Much like, you know, with software, with kernel

  • development, you're kind of improving it as often.

  • Commit early, commit often.

  • We're trying to do that similarly with the hardware.

  • We are going to seamlessly keep updating the PCB as

  • necessary when we find bugs, keep fixing things.

  • We just released this Rev 2, which fixes some of the

  • earlier bugs.

  • And hopefully sort of improve that way.

  • I mean, right now what we want to do is, as I say, get this

  • polished enough to be ready to give to kids, to have a sort

  • of education-ready product.

  • And that's our focus.

  • We don't have a road map for any wild new stuff right now.

  • So yes.

  • It's going to be minor improvements.

  • It's going to be bug fixes.

  • We're probably not going to change the form factor, just

  • because there are so many people producing cases and

  • other things for it that we're kind of stuck by our own

  • success, and the fact that we probably don't want to change

  • this form factor or the pinout for things, just because that

  • will be frustrating for the community.

  • But we've added mounting holes as well in Rev 2, which is

  • something that people wanted.

  • Yeah.

  • That's pretty much--

  • pretty much it.

  • It's going to be continual improvement.

  • AUDIENCE: [INAUDIBLE] without the graphics?

  • ROB BISHOP: So the problem is it's on-die, right?

  • So the sock in this is a graphics processor

  • up with an ARM core.

  • So if we were to--

  • we wouldn't, as Broadcom engineers, have access to

  • another sock to replace it, 'cause, you

  • know, Broadcom engineers.

  • Also, to get the sock redesigned and in new

  • [INAUDIBLE], you're talking millions of dollars.

  • You're not going to be able to do that for $35.

  • You know, this is cheap because it's an existing part.

  • And the GPU, it's very powerful, it's very good for

  • doing things.

  • I mean, So we have Open GLS, 2.0 API.

  • So for example, you can run "Quake 3" at 60 frames a

  • second fairly consistently using the overclocking.

  • It's pretty cool.

  • You can do 1080p video, encode, decode.

  • We've got a whole host of media codecs

  • you can use in hardware.

  • Yes, it's frustrating it's closed.

  • But it's the fact that that bit's closed that allows us to

  • sell this this cheaply.

  • And ultimately, we do get people saying--

  • we just won a Makey Award for Most Hackable Gadget.

  • And someone was saying on Twitter, how can you be the

  • most hackable gadget if your GPU is closed?

  • And it's like, well, I'm not sure I'd teach kids GPU

  • programming as their

  • introduction to computer science.

  • For what we want to do--

  • you know, there's ARM-JTAG.

  • The V6 instruction set's well-known.

  • Yes, you need a binary blob to boot it, but as a learning

  • tool, for price and for availability, we don't think

  • there's anything better.

  • And as I said, this is as open as we can make it.

  • We're actively making it as open as we can.

  • But we're limited to the chip that we have.

  • Cool.

  • OK.

  • Well, if there's demo questions, I believe we're

  • setting up some monitors and some Raspberry Pis next door

  • so we can kind of play, have a bit of a workshop.

  • Come and have an introduction to Scratch.

  • Have a look at Python.

  • Have a look at the Gertboard.

  • And yeah, Hopefully if you guys have been working on

  • projects, you can show us those projects too.

  • Great.

  • OK.

  • Thanks, guys.

  • [APPLAUSE]

  • [MUSIC PLAYING]

ADAM BEBERG: My name's Adam Beberg.

Subtitles and vocabulary

Click the word to look it up Click the word to find further inforamtion about it