John Resig: Redefining the Introduction to Computer Science at Khan Academy
ejohn.org
ejohn.org
Compared to my high school experience of learning in a C/C++ mismash with a teacher who insisted on introducing pointers very early, before ever presenting me with anything interesting to do which would make me want to figure out the details like that (and which scared me off programming for four years), I think it sounds like a dream.
I would be very happy if these kinds of irrationalities could be ironed out in a future version of the language, even if it meant backward incompatibility.
If the language is thoroughly rationalized and there is a simple, beautiful API for graphics then I think it becomes the beginner's language to end all (since client-side web is such a nice platform to target). But for now, it isn't there... I'd be happy to hear that it was a priority.
Perhaps something like Dart will pick up that quest.
I think things like dynamic "this" in javascript will be more confusing for newcomers. But, once you start talking about "this", you should probably already have the basics down.
Edit: With all that said, I'm not saying Javascript is a great language to learn programming with. But I don't honestly know of a single language that I would be perfectly comfortable telling people to learn with. Most languages have their own quirks.
var a = {} + [];
var b = [] + {};
var c = String({}) + String([])
All three are equal. The order does matter in general because its doing string concatenation which obviously matters ([1] + [2] == "12"), but the fact that js always does string coercion and concatenation anytime you have + without numbers on both side isn't really that crazy.
There are plenty of parts of JavaScript like that, but these gotchas are just red herrings. The core language is great.
If you learn from good tutorials it can be as good as learning with scheme.
Notice that I'm not saying these things are bad, but they are certainly more than a beginner needs to learn basic syntax and become familiar with control flow.
I don't think you disagree with me at all... I said that javascript feels strange to me because I didn't start with it. I think it's a fine language to start people on, just an uncommon one that should yield interesting results.
I certainly wouldn't suggest teaching with more "rigid" languages. I was trained on Java, and looking back I can see what a terrible way of doing things that is.
FWIW though, it's been my experience that for all but the most trivial tasks in Javascript, you quickly find yourself getting into somewhat advanced territory with e.g. scoping, closures, and anonymous functions. Again, I think these concepts are mostly weird to me just because they're different than what I personally was trained on, but they're certainly not as easy to grok as e.g. variable assignment and conditionals.
In that way, I certainly see what you're saying. It'll be interesting to see what new problems emerge if this continues as a paradigm for teaching how to program.
But in terms of JavaScript specifically, I think first timers could do without being exposed to lexical scoping and truthiness.
I like that Javascript gives you encapsulation without necessarily object orientation, which is probably a good place for beginning programmers to start.
Case in point, recently some chick from highschool asked me where she could learn to code, so I pointed her to some Scheme tuts, nothing too complex since she requested something as easy as possible.
....then she writes back asking if "this is what jquery uses". I say "no", she gets upset and ask me "what the hell is this for then?"
Long story short: for a lot of people out there is not about learning how to code but learn-X-to-do-Y no more no less...
If you're going to ask people to download something, I still think Smalltalk is a nice alternative -- and it has been for a while:
It's simpler than most OO languages. Not sure why you think it doesn't scale as well.
The reasons why JavaScript is thought not to scale well are very-well known. They include:
Minor syntax issues:
- Deeply indented callbacks, passing a function to a function to a function and the resulting complexity of flow control.
- Issues with "this" become more likely with larger codebases. I know that JavaScript's "this" WTF is very old news, but it's the gift that keeps on giving.
Major design issues:
- JavaScript is not an OO language, it is a language that provides an OO construction kit. The OO patterns are layered on top of the language, the best idioms (e.g. self-invoking anonymous function, revealing prototype pattern) are not obvious and took a while to become common, and since none of them are mandatory, they can be ignored, mixed and matched or just gotten wrong. This is not a prescription for trouble-free scaling of the code.
- Lack of a built-in packaging/module or linking system. Requirejs is filling this gap, but for a long time most JavaScript code just wasn't partitioned, it was put all in one file. It is unreasonably hard to build large systems this way.
Granted the stuff you made was nothing like the 3D shooter you played in the evening, but it was something _you_ made that made it really cool. JavaScript works really really well in that regard in that you can easily make stuff appear on the screen with very little effort with html (and move onto canvas / webgl) and they don't look half bad. Unlike other programing languages graphics are "built in". Python is great, but compare drawing on the screen in C++/Python/Java v.s. JavaScript and you get my point. As a first language it is a perfect choice as it is something you can get excited about.
As opposed to the classroom model of: [lecture -> assign work -> grade & return with static feedback], students can actually play with the subject matter during the lesson, instead of turning pens into projectiles or doodling in their books because they're bored to death (not that I know anything about that). Then they get immediate feedback, whether right or wrong, by seeing how the instructor would have solved the problem. That kind of hands on learning, where the student learns through their own trials and errors is much more fun than sitting through a lecture that has to accomodate the varied learning paces of a classroom of 20+ students.
The video posted here unfortunately focuses on the live editing aspect of the app, but you can see the interactive lesson function more clearly in this early prototype video[1].
This kind of interactive learning is the same thing that Sebastian Thrun is working on at Udacity[2]. For all the promise of making university courses available for free online, I think this is the truly disruptive stuff going on in online education, because it's way beyond just filming lectures and throwing them online, it's a fundamental leap forward for education.
[1] http://www.youtube.com/watch?v=nvaaude_1hk [2] http://www.youtube.com/watch?v=75TP3hoPA8U
Kind of removes the need for a "lesson" and teacher, in favor of a facilitator that can go around and help those learning to discover things they hadn't considered.
This type of teaching has a formal name. It's sometimes called course delivery reversal or course flipping, with the idea being that, essentially, the "lecture" is done outside of the class, and the "homework" is done within class. In this way, you can have a subject matter expert present at the time when he or she is actually needed.
I've tried it in my own courses (recording in Camtasia and having students watch the lecture outside of class, where they can rewind, pause, etc.), then dedicating the entire class time to allow the students to actually work problems. More generally, active learning has been gaining popularity in the classroom.
- http://www.programbydesign.org/
Those folks have tons of data on what works, what doesn't.
I get a bit of a NIH feeling from this effort, that I hope is unfounded.
>_>
http://www.wescheme.org/We are home schooling and always looking for new and interesting stuff to do with him.
Thank you, John.
Sorry yours wasn't the same!
Anyone remember Morphic, Squeak, Lisp, etc? Emacs?
Lively Kernel is pretty darn cool too.
I'm not saying it's bad, I think it's cool that this style of programming is making a come back (and about bloody time I suppose).
The only thing I'm not certain of with this course is actually calling it, "Computer Science." Perhaps that's because I tend to agree with those who question whether it is indeed a science at all. However after watching a few videos and trying some of the exercises it seems to me to be a great way to learn how to program.
Everything is programmable: the display of a bot (Canvas 2D Context API), the buttons used to to control it, the interactions between the bots, etc.
One issue I am having here in Chrome in Ubuntu on my computer is that some of the videos are giving me minor playback problems where some visual features are not visible. For example, in the intro to drawing tutorial the rectangle seems to turn into a line because three of the sides are not visible. This is probably just what I get for using Linux though.
Anyway, I was wondering: have Mr. Khan and the rest of them put any thought into how the new interactive programming environment might be applied to learning math or other concepts taught by Khan Academy besides Computer Science? Like specifically taking some of the math or other lessons and presenting applications to tweak that would demonstrate those concepts. Maybe lessons in those other areas could sometimes include a link to a programming experiment.
Also, other question: are there any plans to try to cover a broader range of computer science or maybe even software engineering topics, for example things like Objects/Classes, components, unit testing, QA, feedback loops in general or for example between the developers/analysts and the users or between the developer and his test suite?
My one daughter is going into 8th grade and has had a total of two half-semesters of "business and computer science" where the most they do is to play around with PowerPoint and Excel. Computers are so integral to every-day life now. It's positively disgraceful that every child shouldn't be educated on what they are, how they're built, how they work, how to program them, etc.
This CS learning platform brings programming education back to its simple roots, back when your first program was as simple as drawing a circle in BASIC. It also leaps it forward, borrowing ideas from Bret Victors responsiveness talk was brilliant and I hope sets a precedent for programming education moving forward.
I really admire what you've done, Mr. Resig and the Khan Academy team!
Interactive platforms and approaches have been constantly introduced by CS educators to teach introductory programming. Educators are very aware of the value of interactivity and reactivity in teaching and learning. The only new thing here is the utility of cool new tools equipped with latest web-based technologies. Further, their approach is limited to programming, which is a large part but not the only approach to Introductory Computer Science. For example, they can't touch approaches such at MIT or Rice (I think), among others, which employ cool physical robots to teach coding, thinking, programming and robotics.
So this is not so much redefining, but rather a little enhancing CS education with cool tools.
Also, what components exactly are you planning to release open source? The other work in this area that I've been able to find is very tightly coupled to the rendering library (d3 or PJS). Is the Khan Academy system open to alternate rendering libraries?
Still it's better than c++ or java.
I have my own reservations about JavaScript the language, but after seeing what John had built when I started, I believe that this is undoubtedly the right way to go.
If we wanted to use another language to accomplish similar goals, the options would be:
1. Run the code on the server. This dramatically increases latency, introduces more security concerns and would dramatically increase our costs as this would inflate our server load dramatically. Really, its the latency here that makes this a no-go. As you use the number scrubber, it's re-executing several times a second, and that's just not going to work out with client-server communication (even with websockets).
2. Run a different language that compiles to JS and runs in the browser. As it stands, the overhead of doing this is also too much. It would be shipping a compiler or parser down to every client. CoffeeScript may be the most reasonable option there, but it has many of the same warts as JavaScript anyway. Any parser/compiler needed introduces an extra step in the pipeline and would inevitably reduce responsiveness.
While JavaScript the language certainly has its problems, it's what it lets you do that counts here.
Not at all--it takes some work, but especially if most programs are less than a few hundred lines long, the overhead is far less than you think!
1) You want a language/environment which encourages creativity and offers visual feedback as a reward. This is where Brett Victor's demonstration becomes remarkably powerful. Instant feedback is important not only for the creative process, but for maintaining interest. Some children will never enjoy programming, but for those who will, how can you excite them? Languages such as JavaScript can be a good tool for this. Children spend most of their time in an IT lesson playing web games when your back is turned. Teaching them how to make those games helps foster an interest, and JS is a good language for doing this.
2) Do you want a language which teaches the basic building blocks of CS, or do you want a language which is of practical engineering benefit? You could argue that functional languages such as Haskell encourage children to really think about what computation is, or conversely you could propose that teaching JAVA prepares them better for the real world. Both have benefits, striking a balance is difficult. Deciding what's important can often be subjective: functional or OO, strongly or weakly typed, concurrent or not? There are so many factors at play that no-one will ever agree on a perfect learning language.
3) Simplicity of syntax and semantics. Wrapping your head around abstract concepts can be a challenge when young. Maybe Haskell's functions are too unimaginable to grasp, or Smalltalk's object model too mired in strange syntax. Is a type system containing Integers and Floats meaningless to a child who still finds fractions puzzling?
4) Finally, compiling and debugging are pains which few children are going to endure without losing interest. The requirement for code to run within the browser, I believe, is paramount: no installation, no mangled libraries and dependencies. Instead it can be run on a school computer just as easily as it can at home. Codecademy has done this really nicely, and it looks like John and his team at Khan Academy have done the same.
JavaScript ticks a number of boxes in my opinion. The number of tools and libraries available means children can build interesting, visual programs. Functions as first-class citizens helps introduce students to concepts within CS, and execution within the browser is a huge benefit.
Nonetheless, I don't believe it has it all. Its object model makes the teaching of basic OO concepts a nightmare. Its untyped and furthermore its type coercion can cause strange behaviour (this IMO can be extremely damaging; things work then all of a sudden they don't). Operator overloading is odd. Its just too quirky! Personally I don't believe it is an appropriate first language. It does some things nicely, but I believe it leaves itself open to a lot of bad habits if left unchecked. Ultimately I believe a first language should enable the student to pick up any language they want thereafter, and feel comfortable within it. Primarily it should teach pure concepts, not syntax and style.
Shameless plug, but having reached my own personal conclusion I decided to write my own language for the browser as an MSc thesis (http://joeroot.com/wardrobe/try/ | https://github.com/joeroot/wardrobe). Its still far from complete, but I think it serves as a demonstration of what I feel is important in a language. I like many was inspired by Brett Victor's talk, and thats the eventual, unmet goal for the environment. But for now my focus is on the language and implementing it within the browser. It is strongly typed (but fewer types), has first-class functions alongside a pure OO model, and less syntactic pluralism to name a few of its traits.
As I said before however, programming languages are a hugely subjective topic, and I think John's project is a big step forward. Personal experience of teaching children would suggest that this looks like a far more involved experience than what is currently available.
Some of these warts are surprisingly easy to work around: for instance, they always use "===" and don't even tell the student about "==" until much later.
In other cases I just think you're wrong. For instance, I think that the prototype system is simpler and easier to understand than classical objects, while at the same time being more powerful.
You also want "real typing". Do you mean strong typing or static typing? In either case, the language will fight you all the way, being intrinsically weekly/dynamically typed, and so it is somewhat orthogonal to prototypes vs classical OOP.
IMO, if you want to teach type theory, you ought to be doing it in a language like Haskell anyways. Which might not the best thing to expose an absolute beginner to. ;)
Ubiquity! Ubiquity! Ubiquity!
I learned to program with BASIC in the 80s. Somehow I survived, though old-school BASIC was like a clay tablet compared to JS. (Get this: No functions, all global scope. It was like assembly language but with less power.)
But BASIC was awesome because every microcomputer had it built in. Walk up to a machine and start typing BASIC. I must have typed in the same program 20 times.
I remember spending 45 minutes with this random Wang terminal thing at the local library, trying to find BASIC. I never did, of course, so I remember concluding that it was probably broken. I'm still not sure I was wrong, exactly. ;)
Ie:
var draw = function() { point(random(0, 400), random(0, 400)); };
onDraw(function() { point(random(0, 400), random(0, 400)); });
Explaining variable assignment is much simple than explaining callbacks and/or higher order functions.
On a more practical note, this is the convention that processing.js uses, and thats what we're using to render the canvas.
"Welcome. I'm super excited about computer science. it's my favoriate thing in the world... except for pot".
QFT.
You are right though; since it's aimed at high school, I doubt it's pot. ;(
I learned a few things:
1. We seriously need to work on her typing skills. We practice typing once a week but I think I need to increase that.
2. I need to go out and buy a smaller mouse to fit her hand.
3. She immediately understood functions and parameters, which, honestly, was all I really wanted her to get out of this.
4. The sliders are critical. She has never seen a coordinate system, but the sliders allow her to play around with location without completely understanding what the numbers represent.
5. At this age, she was engaged because she could draw things (she had a house worked out in about 15 minutes, which was also the limit of her attention span). I'm not always a fan of gaming things up to make them more palatable, but it definitely worked here.
The XO (One Laptop Per Child) and now this.
Please stop giving people tinker-toy environments with tinker-toy problems that have nothing to do with the real world.
If someone's not motivated to learn, yes, they may need this kind of hand-holding. But if someone genuinely wants to learn to program computers for a living, this is - in my opinion - not what they need at all.
Drivers Education is taught in a real car, not in Mario Kart on a Wii. I'm sick of the Mario Kart version of computers being spoon-fed to otherwise intelligent people as though it has anything to do with reality.
I learned to program using Turbo Pascal. I was not so far removed from reality that everything I learned was almost completely useless. I have not given this much attention yet, but so far, it feels completely useless.
PLEASE correct me if I'm wrong.
Yes, but you're removing NECESSARY complexity, in order for them to learn something barely worth learning.
"I feel that purely textual coding environments (or coding environments that favor textual output over visual output) serve as a highly uninteresting introduction to programming."
There are people who get to use non-textual coding environments, and there are people who get to work with very visual outputs. I feel that by giving them these tools, you've castrated their ability to understand anything meaningful.
"Obviously, right now we’re not teaching very much in the way of traditional Computer Science content,"
Yeah, approximately none of it.
"We decided that it was much more fruitful for us to tackle the most challenging problem that exists with Computer Science education first: Getting people excited about programming."
Of the people who are AT ALL interested in learning about programming, I have encountered exactly zero of them who aren't already excited about it. Why do you think that GETTING them excited is the "most challenging problem"?
I absolutely applaud the effort, and I sincerely wish you success.
I just personally think you're doing it all wrong.
How is this tautology at all helpful? Of course the people who are excited about programming are excited about it. For every one of those, how many view computers as entertaining magic boxes? Most non-programmers I know regard programming as a kind of impenetrable dark art. There is such a monumental disconnect between "Hello World" and "I can do whatever I want!" that most people just assume that they'll never be able to achieve any meaningful result. Being able to create something visually interesting that you can share with your friends is a big motivator, and I think you're too quick to deride it.
Also, I don't get why you view this as something "barely worth learning." The "real world" doesn't run on Turbo Pascal anymore, and Javascript, like it or not, is a big part of modern programming. Do you really think it will be so difficult to translate skills acquired generating interesting games into skills that can be used in a modern web shop?
Anyway, you're welcome to your opinion, I just personally don't understand it. Luckily for people that share your opinion, there are plenty of traditional CS resources all over the web and elsewhere to learn programming in the "real world." For those whose "real world" consists of playing games and wondering how they work, this might just be the thing that motivates them to look under the hood.
He doesn't just claim that it's a problem with Computer Science education. I agree that it's a problem, and I value tools that try to help.
No, he's claiming it's "the most challenging problem." I strenuously disagree.
I think the most challenging problem with Computer Science education is retention. I know a bunch of people, personally, who for whatever reasons, got excited about programming - they wanted to learn. They had their own intrinsic excitement for programming, and wanted to move forward.
A few things hindered them, in my opinion: they had no idea how to get started, and even if they did get started, they soon lost interest because it wasn't obvious to them how the tools they had access to were ever going to get them anywhere they wanted to go. Given a goal - say, how do I make a game? - they did not believe there was a finish line they could cross.
I can accept the criticism that I'm too quick to deride it. I'm sure this is a valuable tool for some people. Given the people that I know, who already were motivated, I do not believe this tool will help them at all, and that disappoints me because I had hoped for more.
I think it's barely worth learning because I know hundreds of professional software developers, and I know approximately two of them who got to play with HTML5 Canvas at work - myself and a friend at another company. Both of us did it in our spare time, for fun, because our bosses would never have asked us to do it. If you know more people who are professional HTML5 Canvas developers, then your anecdote trumps mine.
I wish all trolls applauded effort and wished people success.
I may not be expressing it well, but I honestly think calling this "Computer Science" is absurd. Feel free to call people who disagree with you a troll any time it's more convenient than actually having a conversation with them.
I wish we could discuss the wild success of One Laptop Per Child, which tried a similar approach. But we can't. And I personally believe it's because of this "tinker-toy hand's on" approach that One Laptop Per Child has failed so spectacularly. I fear that this approach will shove Khan Academy's "Computer Science" section into a similar footnote of history.
I also think all of this is too important to do wrong, so it pisses me off to see a group so prominent make such a misstep.
http://scratched.media.mit.edu/resources/harvard-cs50-introd...
According to slide 27, most students prefer it that way:
[0]: https://khanacademy.kilnhg.com/Auth/LogOn?ReturnUrl=/&nr... [1]: https://calenglishbr.appspot.com/ [2]: https://uc3m-ka.appspot.com/login?continue=http%3A%2F%2Fuc3m...
I can't find any info about KA moving to closed source.
They even make their ec2 instances open source!
My worry with this is that there are many, many youngsters out there who are intimidated by math but who would absolutely love to write code. This is especially worrisome since such a huge amount of coding involves no math at all.
"One question is obviously: How much “Computer Science” do we actually teach? Obviously, right now we’re not teaching very much in the way of traditional Computer Science content, we’ve placed far more focus onto the platform and feel that a lot can be gained from just what we’ve already built. We’re still very early on in the production of content and will be producing much more, of increasing complexity, over the upcoming weeks and months.
We decided that it was much more fruitful for us to tackle the most challenging problem that exists with Computer Science education first: Getting people excited about programming. If we can get people excited about programming, and build (or point them to) the resources they need to learn more then we will have been successful.
Eventually I think we’ll work to effectively replicate the materials one typically finds in a “Computer Science 101″ but in a way that is far more engaging and self-paced than what you would find at a traditional university."