There were challenges. There always are. [laughs] The challenges weren’t so much related to language, not among the partners anyway. Working with kids all our partners agreed the only possible common language could be English. Obviously that’s not ideal but in the absence of automatic language translation there is not much else you can do.

Also, and I’m not sure how relevant this is to you, but in a project like this there are problems with different educational cultures, different research traditions and different styles of methodology… I think there is a bigger message here which is that if you consider this list of countries, it’s pretty diverse. Thirty years ago there would have been some countries on that list – even 10 years ago – that we wouldn’t have been able to have a joint project with, Bulgaria, for example.

There was a lot of learning that had to be done and I’m sure that none of us, I’m not necessarily talking about WebLabs now, have really got the answer of how to cross these cultural boundaries. I guess we’re learning slowly how to do it.

Well, the project from start to finish was three years. But it is important to say that it built on, from both a technical and educational point of view, a previous three-year European project called 'playground' in which the focus was on kids building their own video games. The WebLabs project was very explicitly about kids learning maths and science but we borrowed both the technical infrastructure and the lessons we’d learned about activities and methodologies and so on from the playground project (http://www.ioe.ac.uk/playground/) which had been over for a year before WebLabs started, so if you take from the start of playground to the end of WebLabs, it’s over seven years.

If you’re trying to do something really ambitious you know it takes time.

WebLabs is based on two big ideas. One is that a really powerful way to build your understanding of mathematical and scientific phenomena is to actually write programs that do what you think they should do. So, rather than learn how things react when a force is attached to them, to write a computer program that really does it, and think about what it takes to build that program… building stuff on the screen in order to build stuff in your head.

That’s number one big idea. Number two big idea is if you share your programs with other people, if you build models and share them and say “don’t just read what I think, here, run what I think”, that’s an even more powerful way to learn.

WebLabs has got these two pillars. On the one hand, kids building things for themselves test out the way they think the world works; on the other we’ve built this very substantial web-based system called WebReports (http://www.weblabs.org.uk/wlplone/) which is for sharing… not just chatting about what you think, but exchanging models of what you think and allowing people to say “no, no that doesn’t work, hey, look at my model it works better than yours, have you thought about this, and so on”. In this way you can share real evidence of what you think.

I’m going to tell you about one that’s actually in a clip and then you can check it out for yourself on our website (http://www.lkl.ac.uk/kscope/weblabs/video.htm). Okay so we have a Lunar Lander, landing on the moon. Some of the kids [in England] build a model of how it lands and what its speed is over time and that generates a graph and then they send a challenge to some kids in Bulgaria and say “Can you land your Lander like ours?” So, the kids in Bulgaria have to build a program that does what they think is the same thing and the English kids can get it back and say “well actually, no, your graph is different from ours… maybe our graphs are actually the same but they have different scales” and they have a whole conversation about whether two graphs are the same if the scales are different. It’s a really nice point to discuss actually.

Or here’s another one… now this was our most successful task and we didn’t predict this. We thought it would be dry and boring even though we were trying to make it playful and creative and in the end, like most things, we succeeded by accident. We got these groups of kids who got obsessed by writing number sequences. You know so 1, 4, 9, 16, 25, if I say that sequence to you, you can say “Oh, I think I know what that next number is”. Pretty soon you can do really complicated ones. And so the way it worked was this: say kids in England generated a sequence by having a rule and then they challenged the kids in say Italy, asking “Can you find the rule for my sequence?” Well you could do that without writing computer programs but then it’s just kind of your opinion against my opinion. So what the Italian kids would have to do was to write a program which generated the same sequence. But how many terms do you need to show that your two sequences are the same? I mean will five do, will five hundred do? One of the really lovely things in ToonTalk is that it is completely effortless to generate ten thousand terms and get in your helicopter and fly along the screen looking at these numbers and so on. This really set the kids on fire. They were so excited to see if they could get more and more complicated sequences that the kids in the other country would never be able to crack the code and so on. There was a lot of really creative thinking by the kids.

Building things for yourself is a very exciting thing to do. Anyone who has ever written a computer program and ever made it work it doesn’t matter how simple knows this…

I’ll answer the second part of your question first. Over the three years, as you can imagine, there were a huge number of different strands of this project and so many partners doing all kinds of different things. But the most time consuming thing, the thing that really makes the difference between successes and failures with kids being creative and really thinking in new ways, is not the technology, or sharing across the web. It’s getting the activities right. And, it’s surprising because, of course, once you’ve got them right it looks like it was easy. And that is fantastically time consuming.

I wanted to say that because while the technical side is obviously really, really necessary, it isn’t enough. You can get the technology absolutely perfect and still have nothing creative for people to do with it.

To answer the first part of your question, you can find out how WebLabs works by getting yourself some copies of ToonTalk, which is ridiculously cheap, I think it’s about 25 dollars, and then downloading all the WebLabs tools.

Well at the moment they aren’t. It’s not clear whether they are going to be freely downloadable, because our publisher partner Logotron is going to make a WebLabs’s edition of ToonTalk, the programming language the whole thing is based on, and if he might wrap the tools with his version of ToonTalk, which might be the best way to get them, then everything works on a CD and so on.

But anyway you get a copy of ToonTalk, you get all the tools, you read the way we did and off you go. I mean it’s not terribly difficult.

[Editor's note: The WebReports technology is open source and freely available, including the design documents.]

Exactly. Exactly. There are lots and lots and lots of delicate issues. Kids want their photos on the web and you have to make sure it’s only the backs of their heads, you can’t imagine, well I’m sure you can imagine, it’s terribly difficult. So technically that’s that.

Well we didn’t do that anyway, we didn’t do that for either of the projects. Well, coming out of playground we actually had a piece of software it was called Pathways in it’s first instantiation and in the end Magic Forest (http://www.logo.com/magicforest/manual/) and completely surprising to us it turned into a best selling piece of software for the publisher, so that sometimes happens. You have a product and it really works and it works so easily and it is so easily packageable that people can go to a store and buy it.

Now, I don’t think that will be the kind of thing that WebLabs will do. I think what will happen with WebLabs is, what will make me happy enough, is if the ideas of WebLabs help to breathe life into someone else’s project, maybe using a different programming system than ToonTalk… I mean since we’ve first conceived of WebLabs the idea of a shared web-based system for exchanging programs and ideas and so on has become common place but in 1999 when we first had the idea of WebLabs, and we got the project in 2001 I think it was, you know, it was still quite a revolutionary idea to get it right.

I think if people pick up the big ideas of the project it may be they run with it and it comes out looking quite different and presumably quite better technically because we’ve got all these years of experience now and I’d be quite happy about that. I think that often does happen actually I think if you look at the number of things that Logo inspired 35, 40 years ago where people are still being inspired by Logo even though it doesn’t look like Logo anymore.