I just finish my first attempt https://www.youtube.com/watch?v=BiwWT2CERA8
I wanted to ask the HN, what other ideas or what type curriculum I can do to make it more fun and interesting for them. All ideas are welcomed.
Are the kids remotely interested in programming, and does it make sense to try teaching them if they show no interest? As a child, I remember pushed towards taking piano lessons and hating every second of it.
And now as a parent with kids, I wish I got a chance to take piano lesson (or any musical instrument) when young. I make kids do things that I know I didn't enjoy when young, like studying, being responsible, facing consequences. And just like me, it's a decade(s) long process.
Obviously there should be reasonable limit on how much you will push your kids to do things they don't want to do. But I believe it's every parent's responsibility to help/push kids experience/learn new things when young. Even if they don't enjoy it (if so, it should not be pushed too hard or for long), it's a success because everyone involved gets to know what the kids enjoy or hate.
I'm not suggesting scheduling every waking hour of a child for an organized activity/study. But more/various activities for young kids are generally a good thing imho.
But, I think programming is a creativity tool, and as such kids can use it for creation. Sadly our tools adds a lot of complexity and doesn't help to make this creativity flow easily. Many kids will be frustrated by the number of details they should learn to move a simple object on the screen. This is why I recommended Scratch Jr. and Alice 3D, while not perfect you can do some advanced stuff with simple programming. If kids are not frustrated or annoyed by the tools they will engage in creative projects in art and/or science because this tool enables incredible new ways of expression.
"Would you like to go to @codeclub"
"Yes!"
"Ask HN: How should I teach code to kids?"
Getting kids an opportunity to become interested in @codeclub is a related, but separate question.
No one knows the details of your experiences and psyche, I'm not saying that you should continue your piano lessons.
However, parents are primarily responsible from guarding children from their own poor choices among other things. I wish I didn't stop the guitar lessons when I was a child although my mind was primarily on girls, consoles and basketball.
The first contact I had to actual programming was messing with Pawno scripts to make gameservers in San Andreas multiplayer when I was 12... Spent hours and hours programming with it. It started simple placing coordinates where the cars should spawn until I started to mess with more 'complicated' stuff like gates that open when player get near.
I would get some game that is popular among kids that age (minecraft?) and teach them to make simple modifications, if they like they will try to push themselfs to make harder and more complex stuff, the reward/work ratio is bigger than make a game from scratch on HTML, where there is a lot that needs to be done before getting to the "fun part".
Many of us who grew up around MS-DOS had to figure out ways to get around the 640kb base memory limitation and deal with conflicting IRQ ports. Did we want to do that? No, we just wanted to play Dune II. But through that, it got us tinkering with installing new programs, tinkering with configuration files, and eventually learning BASIC, putting together your own computer, etc.
Unless the kids are extremely mature and patient, the best way is to focus on an appealing end product. Don't present it as "oh, let's learn about boolean operators" today, as much as "hey, want to build a [something cool] together?"
And you know what? It may not stick; maybe they'd rather be outside, bake some cookies, or beat on some drums, and that's fine, too.
https://en.wikipedia.org/wiki/ZX_Spectrum
My father was not an engineer and had no idea what to do with it. I also was completely uneducated (I was 10 years old boy in 1994) and basically ignored school classes (or prefered to sleep there). So I also had no idea what it is.
He bought me games and bunch of books about programming on ZX Spectrum.
I just started writing code from these books and see the result on the screen. Then I started slightly modify code. Then I started writing my own little programs.
After 6 years, I managed to learn assembly language and even write my own version of Snake game:
https://en.wikipedia.org/wiki/Snake_(video_game)
I had no help whatsoever from my father, from school teachers or anyone else.
Then my father bought me Pentium I 166 Mhz with installed Windows 98. Unlike ZX Spectrum I couldn't see what's going on under the hood at all. I tried some available programming languages. Their IDEs were huge, bloated, complicated and confusing. I was so frustrated that I couldn't understand what's going on under the hood so I dropped programming for next 7 years. I returned to programming when I switched to Linux and started writing programs in C using Vim as editor (and I'm still stick to Vim!).
So here are my advices based on my story:
Buy them something similar to ZX Spectrum, a small, simple, programming friendly micro-computer where kids can start programming straight away. Give them some simple programming books. Kids will learn themselves!
I personally can't imagine how I would start programming on modern computer. Especially under Windows OS (such a messy unfriendly OS for beginning developers!).
I would definitely not teach kids web-development! It will create a mess in their heads!
It really was the perfect intersection of complicated enough to be a fun system to explore, and you could do Real Projects with it, and simple enough that you could 'hold the whole machine in your head'. I still remember how proud I was when I figured out how to write a TSR driver for a mouse cursor using the tape buffer for storage so that I could move a sprite around with a joystick at the same time as being able to use the shell. At the time, it felt like magic.
ZX Spectrum+ from my father in 1988, self-learnt after that.
Made a few games like you, but had no hard drive or disks. Storage mechanism for my programs were pen and paper.
Learnt machine code programming in tenth grade.
Later got a PC with Windows in 1997 and learnt C.
Worse, teaching under-12 year olds to think "logically" like computers (of any kind) may mess them up for life.
We'll probably hear about it in a few years, given the current fashion. After the consequences become apparent and the victims complain, if they can.
(Fortunately, humans in general are quite resistant to instruction.)
It's a 3rd person view where you control an astronaut on different missions. It has some RTS elements as there are various robots to control.
However, in addition to being able to directly control the robots (which is fun and accessible), you can write code that it'll execute and automate some simple tasks, like fetching resources, defending a base, etc. It's very high quality and well made, and incorporates step by step lessons, starting with basics of programming. Your creativity is the limit.
What I've found out recently is the game was eventually released as open source (it was originally commercial) and picked up by a community of people working on making it better, and it's suppported on modern systems.
It's called Colobot [0] and I highly suggest you try it out, see if you think the kids would find it interesting.
* You don't need any complicated stuff for it, just printing to stdout and reading from stdin.
* Kids love making up stories and games
* When they understand the basics you can gradually increase the complexity: Add if/then/else conditions (do you wanna go left or right?), functions (e.g. to parse answers to questions), variables and simple arithmetic (treasures found, monsters fought, ...) and libraries (e.g. for adding randomness to the game).
And if they're hungry for even more afterwards, you can add some graphics programming into the mix. By then they should be motivated enough by their story that they actually want to learn something more complex and challenging.
I'd recommend a scripting language to get started, as it removes the need to compile code, which (IMHO) just adds unnecessary complexity.
Not sure if this is the best approach, I tried it once and it worked great though!
Or use a REPL (eg repl.it)?
Start off by following the tutorial on the right panel when you create a new project.
Once you're done with that, your cousins will probably be exploring on their own. Help their exploration by printing out some Scratch Cards (https://scratch.mit.edu/info/cards/).
After they go through those, check out the Harvard Scratch curriculum (http://scratched.gse.harvard.edu/guide/). There's full of exercises and activities in it.
By now you can tell if your cousins actually wants to learn to code. If they do, they'll be building things in Scratch on their own and remixing other people's examples.
I can go further as to next steps, but this should be plenty of material to get started :)
Better because portable, works offline, and the language itself is more sane (although I haven't checked recent versions of scratch)
Thanks for the reminder! I forgot that another neat thing about Scratch is that it connects with different modules such as littlebits, arduino, etc. This allows them to integrate hardware with the games they build and paves the way to circuits, etc.
Edit. Adding info about Scratch extensions.
I wrote a book to teach 7-11 year olds to code in Python and Scratch and teach them some computer science along the way - I read a few other books out there first, and there's a lot of "just copy out this code and things will happen", which is exactly what I tried to avoid in this book.
The reviews: http://www.goodreads.com/book/show/28232614-coding-unlocked#...
The book: https://www.amazon.com/gp/product/B013R4OFVA/ref=x_gr_w_bb?i...
I was impressed with how quickly he went from "just tried this at school" to "leave me alone, Dad, I know what I'm doing here".
Seems like a good intro for kids.
the rest of this is just me criticizing things I used and didn't like, so feel free to ignore
I tried reaching a friends' kid programming with Scratch and making any kind of more complex stuff was actually too hard because of the "simplifications" built into the language. Can't remember the details, but you had to use globals instead of passing parameters for something basic, like constructing objects.
[0] processing.org
But if you want to teach coding to increase your own esteem in their eyes, or to validate your own life choices by making them follow yours, then step back a bit before proceeding. Get to know them - and yourself - well before embarking on this quest.
Also moves into HTML, Python, etc.
I had great fun teaching using their curriculum. They're rolling out worldwide and all their materials are CC licensed.
A lot depends on how much time and interest they have. With some kids, you show them a few things and act as a resource to them and they are hooked and off to the races. With limited time (many kids are overscheduled), you need to adjust your approach accordingly.
I am a big advocate of Scratch as a first language. The payoff is very quick, "syntax errors" and fussiness over indents, capitalization, and punctuation are pretty much nonexistent, and there is a vibrant community with millions of examples to inspire and "remix".
There are a surprising number of online edutainment/learn to code services that flat out don't work for 7 - 9 year olds because the software developer threw up their hands at the prospect of complying with COPPA and just set their terms to be "must be 13 years or older". I think it is a bad start to setting a child's moral compass to tell them "just lie about your age to access this service".
https://www.amazon.com/Lauren-Ipsum-Computer-Science-Improba...
For younger kids I only recommend Scratch Jr. since it is natural, has less friction, and doesn't require reading.
I am also a fan of hour of code ( https://code.org/learn), it uses popular games and Disney characters and typically no project takes more than an hour.
CoderDojo in an NGO that has spread globally. I wholeheartedly recommend it.Just don't leave your notebook lying around, or some curious kid might take it apart.. (no joke)
Working with games and graphics is a good way to give a goal to the learning.
It comes with PocketCHIP, which is a cool cheap way to get handheld computers in kids hands, or you can get it standalone and run it on a PC/Mac/Linux.
We've recently started to teach kids with the use of Aidlab (https://www.aidlab.com/developer) to show them some basics about heart, lungs or motion data. They are able to measure themselves, develop simple apps thanks to Unity engine, and make use of those data (ex. building games that use respiration level to control game character).
It's basically a UI for writing/sharing/leading code exercises.
It should obviously be using a database and have better security, but that wasn't a major priority for my original purpose.
Creative Technology Club: http://www.ctc.ooo
He is starting this in small in-person batches for kids and adults (Ahmedabad, India), with plans to take it online in the future. Would love to hear thoughts on how to best scale this model.
My 6 year old daughter loves video games and constantly wants me to help her make her own. We're currently working out how she wants it to work. I'm going to help her with the art (she'll show me what she wants) and then I'm going to code with her.
I'm not forcing her but she loves the idea so far. I plan on writing up each step on my blog so people can see our progress and, if it's success, maybe it'll help others repeat it.
If she ultimately hates it that's fine. It's mostly me coding and explaining stuff to her because she's so young. I think it's good for kids to have an idea how it works but if they don't want to be a developer like me then that's fine.
One recommendation is https://www.nostarch.com/ - No Starch Press. There are a number of good titles in the kids section. They were also very decent in giving us a classroom discount for materials.
We used ChromeBooks as that is what the school had for kids. It is OK for Scratch but less than desirable for Python. We used repl.io for an IDE and it wasn't bad. It is better to use IDLE if possible...
I've had an idea of teaching a class where the beginning part is to get a Rasp Pi up and running - 1 per kid or 1 per pair. After that use the Raspberry Pi to learn Python / Scratch.
The moral of this story is that it helps to focus on the interest of the children you are teaching, identify desired outcomes that will excite them and deliver the outcomes while secretly teaching them coding and computer science basics. It's like hiding the dog's pill.
Second, depending on their personality, it may give them more motivation to use a language actual software developers use.
So all in all, good ideas are:
* Python because of the load of libraries
* Haskell because it's very human and has a great REPL
* C because it's pretty easy, and gives them the feeling of something "serious"
* Scratch because it's playish
(This is from a perspective of an 18 year old who started at 11 with C and went through a lot of languages and various programming categories. I think those are things that would've motivated me even more if I had them.)
I've got two nieces, one is 4 and the other 2. For the four year old I tried using different games that are similar to ScratchJr. She's a lot more arty/dancey so even though she likes playing games on the tablet, these sequential typed games really didn't do it for her. She would just get a bit bored. The other kid, never plays with toys(ever!) she only draws and plays dress-up, but regardless, I bought her a code-a-pillar (https://www.amazon.co.uk/Fisher-Price-DKT39-Code-a-pillarTM-...) which is for ages 3-6. So even at the age of two she started playing with the code-a-pillar (obviously with the help of her mum), and managed to make it snake around all of her other (unplayed with) toys. So she clearly showed an interest an inclenation to sequential thinking. (Next I'm gonna try get her using ScratchJr, then getting her a Kano or moving to Scratch.)
Personally I would have loved getting them both involved, but ultimately I think only the two year old show's a preference to that "type of thinking".
Sample to move 1 step:
function move(steps) {
$("#lamb").css({
'right': steps + 'px',
'transition': 'all 1s'
});
move(100);
becomes
move(500);
I don't know haha, do they even want to code in the first place.
Doesn't even consider requestAnimationFrame or proportional scale. Also I'm starting to realize jQuery is pretty big even the min version especially on slow networks.
As they get more interested I switch to https://tryruby.org then scraping with nokogiri then asking them about the problems they want to solve.
For the second year, we are teaching AP Computer Science, so we are using Java. I think the students had more fun with animation, simple games, and web development than they have with AP exercises. I find that there are too many concepts involved in teaching Java to make it a good first language, which can be demotivating. Tons of time spent debugging syntax errors as well. While these things are important eventually, some kids get frustrated and give up too early without ever seeing interesting results.
The mbot lessons slowly work them up towards a "sumo wrestling competition" where they have to push one other opponent's mbot out of a ring.
If anyone is interested, we made a little repo to hold our code. It's a little outdated, but we'll get it updated soon.
Teach them to experiment, problem solve, to be creative, and that failing is okay.
Coding is useful today.. none of us know what will be 10 or 20 years from now.
The middle kid learnt some scratch at school and loves it, and has taught themselves loads about it on the web (youtube is by far their favourite way of learning). They have done a fair bit of Python with me too. However they are interested in 3d games so are teaching themselves Blender and Unity (and therefore some C#) aged 11! Now it seems they enjoy 3d modeling as much as coding. Interestingly they are also fascinated by working with the command line for anything, so they are learning some basic bash too.
The youngest is an avid Minecrafter, but at age 8 they have had a go at Scratch junior. However, they have decided they are too grown up for it and are dabbling in 'proper Scratch' now.
I don't push them, but try and create opportunities for them to see new things
I imagine as LOGO was developed for kids, it should work :-)
The game is pretty amazing in that it teaches pre-reading kids how to "code" by presenting it as a puzzle that you add pieces to. Really well done.
I'm a huge fan of exercises instead of frontal introduction. I've had success with the Khan Academy program ("introduction to programming") and then giving them simple web exercises.
I just following the steps one-by-one without asking WHY. Getting the PHP forum source, upload it, change few lines. Then, started to try to change color, adding plugins. (Thanks God! In the old days, there's no good MVC separate framework, I must dig into source to change things to add plugins or change styles. When I did more and more, I started to understand the pattern. And then I tried by testing the code.)
So I believed, maybe started with making a WordPress site rather than just HTML & CSS. Child got an end product to show their friends and able to drop comments in the blog.
Get thing done first -> Then, to understand the basic.
It eventually became PopMath (which I reimplemented in ObjectiveC for the iPhone) but she understood the concept of loops, random number generation, collision with sides (so x coordinate had to be reversed) and event handling (function to get the user clicks).
I also did the same with my son when he was 9, with another game where the player would need to use arrow keys to navigate a maze.
Also in JavaScript because they can see the results right in the browser, can easily add graphics, and even sound effects.
I think simple games are a great teaching tool.
They must be motivated to explore and teach themselves. If you try to dictate their interests or lessons then it will just be a chore and they won't learn how to teach themselves.
They will naturally emulate you if they see you programming. They will also naturally lose interest in programming if there is no quiet time set aside where neighborhood stick-ball or whatever cannot interfere and they are distracted by their friends.
Basic skills in computing is understanding how things works in order to delegate your tasks to them.
It will be a skill working with human and computers alike.
You can start out with a drag-and-drop programming interface and work up to JavaScript or Python. The BBC and their partners have created a wide range of educational resources based on the micro:bit. The boards are really easy to attach to other things and have bluetooth on-board.
I agree with the other comments in that they need to be interested first. Well, how do you get them interested? That depends on their interests and hobbies. So what I would do is create a short video that ties one of their hobbies to programming, and try to discuss what they thought of the video. Once you get them hooked, they'll be ready to learn.
If English is no barrier then https://code.world/ is also pretty awesome, and in the browser.
Disclaimer: I'm one of the devs of KTurtle.
That's why I decided to do as much unplugged activities as possible. The concepts are so much easier to grasp I think when you're in a familiar environment (the physical world). Of course it's important to mix though.
It looks quite good, but does anyone here have any first hand experience with it? I'd love to hear an honest review.
It gets the basic mindset right (loops and what not) while making things visual so that the kid's mind can follow what's happening in an intuitive manner.
If you launch them straight into "real" programming they'll get bored imo.
please don't
I'm not a teacher, but I did begin learning to code when I was about 8. Allow me to offer a bit of anecdata.
I first started to learn to code on the Commodore 64. The first thing I remember making all on my own was a text-based lucky number program. I made various other very simple text-based games, but without someone giving a bit of guidance, I didn't build much that was complex.
When I really started to learn to code was when my parents bought me a modem, and I got into some rather nefarious activities on AOL (phishing, cracking passwords, mass mailing warez, etc). I was on a Power Mac 7100 at the time, and I started to write tools to assist with my endeavours. I mostly used a tool called OneClick which allowed for easy text capture from on-screen elements, recording of macros from user input, etc. I remember my motivation at the time was part competitive (I wanted to create better tools than other people in the community), and part driven by acquisition (I wanted the latest games/applications, etc).
Finally, the third stage of my childhood programming activities came when I switched from Mac to PC as a young teenager (13 or 14). I pirated a copy of Visual Basic 3.0, and started learning network programming. I remember building a very simple multi-user chat server and client program, getting into ASP a bit in order to build web pages, etc. Honestly I'm not sure what my motivation was at the time, but I remember distinctly thinking that this was going to somehow make me special. That by learning to code I'd be approaching a status which is something like a mix between James Bond and the hackers you'd see in movies like, well, Hackers.
I can also say that while my parents and teachers didn't push me (quite the opposite, really - they always wanted me to quit 'playing computer' and do whatever it was that I was supposed to be doing), they definitely encouraged that feeling of "this will make me special" in subtle, unintentional ways. Once certain teachers started catching on, they too helped encourage the idea, but in much less subtle, more intentional ways (you're gonna be the next Bill Gates, kid). I think more than anything else, that's what lead to my successful career in software development.
So, if I can offer any advice, it'd be to focus on motivation of learning first, and facilitation of learning as a close second. Keep the goals simple and attainable, make sure the activities are aligned well with the kids' sense of fun, enjoyment, and/or utility, and I think they'll lap it up quicker than you could ever provide it.
Edit:
One last thing - I think your video is great, but it's not substitute for one-on-one time. As a kid (and as an adult, really) I remember learning well by watching others who were patient enough to let me interrupt and ask questions. I think videos will be great for self-lead learning once the appetite is whetted, but early on one-on-one time will give the kids a much better chance to learn at their own pace.
I have the kids learn Python between age 9 and 11 depending on when they are interested. We write games and learn the fundamentals of CS by following the curriculum of Rice's "Introduction to Interactive Programming in Python" class designed by Rixner and Warren. It has a Python interpreter and game library that compiles in-browser into javascript, which is a very helpful design for a first class, and for sharing their results with myself, their families, and each other.
Afterwards I encourage those who really enjoy it to learn Java, also using web sources primarily. I intercede when they are stuck but generally they help each other out.
Some of them used to use the Java to write Minecraft mods but that has become a bit more tricky.
There was one student who started to learn Java at age 7, but in general that's too young. If you push things too earlier they don't get a head start. They get burned out and don't want to have anything to do with it any more. You really need to be sensitive to what they are interested in and let them lead.
Most kids when they see others making their own games want to try their own hand at it themselves, so I have managed to create a self-perpetuating cycle.
Learning languages to make games makes sense. Pushing languages to force concepts will be resisted and is counterproductive.
The games approach is nice too because they rapidly learn trigonometry, basic physics, and linear algebra by age 10 or so, and are teaching themselves Calculus on Khan Academy by age 12, as well as researching optics, writing their own shaders and generally doing what most would consider college or graduate level mathematics. Their Java skills also often lead to writing for Android. Some selling their games and other software at a certain point, which helps their families.
It helps that on the reservation I have a great deal of autonomy in what I can do. I've worked in public outside schools before. None of this would be allowed there.
I don't really care for the idea of dumbed down toy languages that can't really be used for production software. Python and Java are well designed enough that they are understandable and usable from the age where they are able to do programming thought. Python is a simpler one to read and use and so we start with that. But Java is very similar and the lessons of Python directly transfer. Java ends up being much more useful for distributable and saleable software, though it is harder as a first language.
>distributable and saleable software
Do you or they make the Python code into an EXE for sale, using one of the available tools, or sell as source?
They also offer free teacher and student accounts with the ability to limit time. Can't say enough positive things about the service.
Python is dead simple—I recently had to take on a project written in it having never programmed in Python before and I essentially did not have to learn it to get the feature done.
I might give Python an edge for a number of reasons (a shell is a simpler environment than a browser in a number of ways, some of the language features are cleaner), but to say that it's much harder doesn't bear out with my experience. Capable first-timers become productive at about the same rate.
I've found it's mostly experienced developers who trip over JavaScript, because they're carrying a set of expectations and are weirded out when it does something different. Turns out principle of least surprise means different things depending on people's backgrounds.
Feel free to choose to start someone with Python if you prefer it, or if they're likely to want to work in a domain where it makes sense (say, data science), but it's probably not going to change your learn-to-code outcomes much.
Comparatively, other languages, like Python, will throw sane exceptions and a meaningful stack trace when something goes wrong.
Please do not make your cousins learn to code.