Takeaways are accurate but also add in the importance of more flexible/responsive user input particularly via Storytelling: limitation to text input and buttons is very restrictive.

Yeah, I updated the message with this after my initial post. That feels a little bit like cheating (other folks on the thread don't get a notification of the update) - but it's useful for posterity and if you want relevant information in one place.

How hard is the code-tracing idea to implement by the way?

I have not given it very serious thought, but it might not be as easy as you think. For example, with line-by-line interpretation (which, coincidentally, popped up in my head yesterday when I saw your comments on tracing!), you can't trace into functions.
Where did you find out about the interaction between vars and line-by-line interpretation? This used to be the case many moons ago. Now, you need to put a #line-by-line comment in your code to force line-by-line interpretation.

One reason I want to document everything is so that I don't forget how it all works myself!!

I have about 8 years of experience editing Wikipedia so wiki-editing and building tutorials there would be within my comfort zone.

Sounds great.

To get things rolling, I have done the following:

• Sent you an invitation to become a member of the Kojo Wiki (on Wikidot). In the future we can set up a link so that people can invite themselves (Wikidot supports a lot of member onboarding options).
• Set up a new Wiki page for Educators: http://wiki.kogics.net/sf:educators. As you mentioned in your comments, Educators have access to very little targeted documentation in the Kojo world, and this sounds like a good place to start the content ecosystem effort. Feel free to start putting in whatever content you want in there. I have some links at the bottom of this post that will help you with authoring content on Wikidot.

Some thoughts on Educators (this might help in structuring the content of the Educators Wiki):

• Two kinds:
  • Interested in Programming
  • Interested in Math/Science
• Multiple Levels:
  • Interested in authoring their own content
  • Interested in tweaking existing content
  • Interested in just using existing content

[Update]
Once the effort gets going, I will invite other folks to contribute. I can think of a few people who might be interested in contributing to the Programming section of the Wiki.
And I'll be pitching in too, of course!
[/Update]

Useful Wikidot information:
http://www.wikidot.com/doc:wiki-syntax
http://www.wikidot.com/doc:quick-reference
http://www.wikidot.com/doc:templates
If you have any questions about the Wiki syntax etc, please feel free to ask me. I have done enough with Wikidot, off and on, to be able to answer most questions, or at least point you in the right direction.

I'll send responses in another post to other issues raised in your (previous) post on this thread …

The wiki could certainly do with some work. As far as I can tell the ScalaCodeRunnerBuiltinsCommands page (sorry, can't link) is somewhat out of date too? I have about 8 years of experience editing Wikipedia so wiki-editing and building tutorials there would be within my comfort zone. I also tend to excessively comment my sample code, which is ugly but hopefully more newbie-friendly than a lot of the existing code examples.

Takeaways are accurate but also add in the importance of more flexible/responsive user input particularly via Storytelling: limitation to text input and buttons is very restrictive.

How hard is the code-tracing idea to implement by the way? I noticed something you wanted to do already was to see types in the Script Editor via mouse-hover which would be nice - to the untrained eye, seeing current values by hover-over while the script is running doesn't sound much harder. If I understand correctly, when the code contains a "Var", the script has to be fed in line-by-line and if that's going on, highlighting the active line and annotating return positions doesn't sound too difficult either. But I might be being over-optimistic here :-)

The Storytelling window is definitely Kojo's USP by the way - because of the way it allows an educator to structure a learning activity, it directly resolves a number of issues that had been bugging me about how to present the material using other software. If this was your own bright idea, please feel vindicated!

That's really good to hear!

Takeaways:

• Richer interactive controls are needed within the Storyteller (added on update)
• Browser support is high priority
• Quizzes have lower priority than Interactive Learning material within the Storyteller (because Quizzes can be built in VLEs)
• Code-Exchange improvements are needed.
• Support for tracing of programs would be great.

I'll use this list to populate/sort the backlog.

From a Content Ecosystem point of view, I suggest contributions to the Kojo wiki, with you leading the effort (with my support as needed) as the first step. This will involve putting in Lesson Plans, Interactive Lessons, Tutorials, Troubleshooting guides, etc. etc.
Are you Ok with that idea? Is this something you would be interested in contributing to? I understand that this will take time to flesh out, but we can make a start as soon as we want to!
By the time this gets fleshed out, I should have bandwidth available to make improvements to the Code Exchange - to incorporate the ideas that you have listed, and to try to make it play well with the Wiki and the (upcoming) browser based Kojo Player.

In the Big Picture, it's definitely getting this into a browser that would be my suggested priority. The screenshots are nice in the sharing center, but until you get to play with the turtle or run an activity, it's impossible to get a "feel" for just what Kojo is like! (Seeing code snippets is definitely no substitute for experiencing the thing. I think you can go back as far as Design By Numbers in the 1990s to see the importance of getting this kind of educational environment running in a browser!)

Like you said, Drawing Canvas and Storyteller would be essential. And since this is so turtle-based, you're right that some coding is needed - at the very least a command line for people to play with. (Actually having seen some of the work of Dave Briccetti with young kids on youtube, I may be neglecting how the "learning to code" aspect of Kojo is just as vital a purpose as the "building cool math activities" - he's obviously coming at a very different angle to me, and with different experiences, but that's a reflection on Kojo's versatility! I guess there is an argument that if you're learning to code, you may as well download the software so you can get access to the fully power of it, whereas just for running rather than designing an activity it's better if you don't have to.) The Storytelling window is definitely Kojo's USP by the way - because of the way it allows an educator to structure a learning activity, it directly resolves a number of issues that had been bugging me about how to present the material using other software. If this was your own bright idea, please feel vindicated!

In the longer run, some sort of central server with the features you summarised would make it a more attractive offering to schools and other institutions. But at least if someone can embed it in a browser, then if their school has a VLE (normal now in UK) they can achieve a kind of work-around: (1) they can embed it on a page built in the VLE and use the VLE itself to monitor usage, (2) they can use the VLE to build a quiz to assess understanding of the activity (though there are obvious reasons why it can make more sense to build assessment into the storyboard). I wonder whether it's worth building up traction with schools first to see how important it is to them, and what their technical requirements are. Also, various legal requirements: storage of student data is quite a sensitive area. Asking someone who's already using it with classes might help too.

Two other things: from a "content ecosystem" point of view, the sharing center could do with a revamp. Ability to delete, even edit, uploads? Fill in a description (e.g. with suggestions for classroom-use ideas) at time of upload, rather than having to go to the webpage and then write it as a comment? Tags for more topical categories (e.g. trigonometry, kinematics, statistics, economics, etc)? Ability of logged-on users to add a sample to their collection of favorites? All of these ideas are used by other projects similar in intent to Kojo, and I can vouch for the usefulness of all of them!

Secondly, from a "learning to program" point of view (not my usual professional area, but this really leapt out at me when looking at the youtube videos out there) one of the big things you are hoping kids will pick up is control flow. When watching videos of someone else's code executing, I found myself automatically keeping track of the progress through the program and how it related to what the turtles were doing on the screen. But if I imagined myself in a kid's position, not yet proficient in control flow, I would have found it very tricky. The analogy is someone who can't read music, staring at a manuscript while listening to the piece being played, but not being able to join the two together! (And analogous to control flow, it's failure to follow a repeat sign or da capo that results in getting the most lost.) Just as much modern music software highlights the notes being played so that a non-expert reader can keep track of what's going on, can Kojo have an "annotated run script" button? One where, instead of being able to edit the script during execution, Kojo automatically highlights the lines being run, scrolling if necessary? Annotates in some way the line that the script will return to at the end of the current procedure? That would make it far easier for a beginner to trace how control flow works, and render videos of script execution far more comprehensible. Keeping track of loops or case selection would also be easier if you could mouse-hover over a variable in the script, and see what its current state is - I appreciate this is quite deep in wishlist land but I do think it would be helpful for novices. As most students will start programming with the turtle at normal speed, Kojo is unusually well-adapted for teaching control flow - the script actually runs at a speed where following it line-by-line, at least while the turtle is drawing, is humanly possible! Seems a shame not to exploit that to the full!

Apologies for the volume of ideas in my recent posts but I hope you enjoy someone giving you a fresh perspective - I'm trying to limit myself to saying what things would, in my professional opinion and experience, make Kojo a better teaching resource or more attractive to educators and their managers. If I were to add the things that I'd like to be able to do in Kojo purely speaking as an amateur programmer and enjoyer of all things patterned and pretty, then you really would overwhelmed!!

For what's largely been a one-man hobby-project, Kojo is hugely impressive. I mean, what's basically Geogebra plus LOGO plus GSP plus Processing plus Scala plus its own USPs … this is a seriously flexible platform and can be taken in lots of directions.

Thanks for the kind words.

Perhaps the versatility is a drawback in some ways - makes the job of documenting what it's capable of, really quite garganguan!

That's true. But it will happen. It will take time, but it will get done.

I'm only familiar with the UK e-learning market, but I hope this sort of experience means my advice to you is helpful.

Your advice, feedback, and hopefully active participation - will be very helpful!

I'm going to get back to you later about my thoughts for what might be necessary to promote a content ecosystem

I look forward to that.

An interactive demonstration is definitely better than an animation or video, which is in turn likely to add value compared to just using the textbook. Kojo's "mathworld" looks perfect for that.

It's great to hear that. A lot of the work I did on the Storyteller and Mathworld was based on this idea.

My takeaways from the rest of your post (mixed with my own thoughts):

• An applet version of Kojo with the Drawing Canvas, Storyteller, and (later on) Mathworld, with most likely some limited support for coding.
• Applet embeddable on any site.
• Analytics on student usage (via a central server).
• Quizzes inside the Storyteller, with a Wizard to help with authoring. Marks uploaded to a central server.
• Geogebra 4/5 integrated with Kojo/Mathworld.

Did I miss anything? Did I misread anything?

Expect to see these items show up soon, and right at the top, on the Kojo backlog (the top of the backlog is where I pick the next item to work on).

Regards,
- Lalit

In one of my previous jobs I had a lot of involvement with my institution's Information Learning Technology (ILT) provision, including getting paid to train staff in integrating IT into lessons, being given the job of finding and assessing suitable software, helping departments build up their VLEs and so on. I'm only familiar with the UK e-learning market, but I hope this sort of experience means my advice to you is helpful. I'm going to get back to you later about my thoughts for what might be necessary to promote a content ecosystem, as that will involve some thought collection. In the meantime I'm going to ask some specific queries in a more appropriate part of the forum! But I'd like to share why, in my experience, schools and colleges (used in the British sense of the word) have a preference for things that can be done online.

I know I said it before but it bears repeating because it's such a massive barrier: mere educators don't get the user-rights to install software, or upgrade existing software. I didn't even when I had very deep involvement with my institution's ILT development. When I was trialling software to assess for my department, I was given the CD to take home and play with! On the other hand, we're always scouring the web for resources, and even old and grizzled teachers get excited when they find something new and shiny to click on. (Bluntly: we hope someone else has taught the lesson we're going to, and has put up some cool ideas for us. We don't have time to prepare innovative stuff for every lesson. If we can't find something, we're likely to just have another boring old lesson teaching straight out of the textbook. Sure, these days the publishers often give us flashy electronic textbooks that we can project onto the interactive whiteboard…but they're still very limited, not much more functional than a PowerPoint and no more interesting! Occasionally there's an animated demonstration or something, but one we can't tinker with or customise.)

Secondly, we want students to be able to access resources from home. Giving every kid a CD (or back in the day, a floppy!) for their home computer tended to be a recipe for disaster. There were always too many cases of "it wouldn't install", "it keeps on crashing", leave alone issues about keeping versions updated. When firms started offering web-based software solutions, they cornered the market pretty thoroughly! The other good thing about access to resources in your browser, is that we can either embed in our VLEs directly (preferable) or easily provide a link to a resource stored elsewhere (which we don't like doing if that location isn't under our control - it's easy to invest massive effort in a VLE and find it eviscerated by linkrot just a couple of years later).

Thirdly, it often allows us to monitor whether students are making effective use of resources. We can easily see who has been using the VLE for revision, for instance. My ex-colleague who invested many hundreds of hours of hours building that mammoth trigonometry resource, was pleased that some of the students had made use of it, but the data showed pretty minimal usage by others. There's also the joys of automatically-marked homework! Quizzes we build for ourselves on our VLEs, or assignments on sites the school is paying a subscription to.

If someone is teaching a (specifically) programming assignment, then in that case it seems reasonable to expect students to download their own copy of Kojo so they can do their scala coding. But for other storyboards, animations and so on, I don't think it's so important: the StoryTeller, DrawingCanvas and MathWorld (if needed) are the bits that I'd definitely want to see displayed in browser. As for what technologies are used in the UK e-learning market: up to GCSE level (16 year olds), the market is dominated by mymaths.co.uk who use Flash. But at A-level (16-18) there is a very strong offering from mathsnet who make extensive use of applets. Similarly many university-level teaching resources involve simulations written in Java and played in applets (I think this is because a lot of professors in quantitative subjects know how to program in Java but not actionscript!). An in-browser applet like GeoGebra's would be perfectly adequate, and I don't think you need to worry about "reputation" of applets - schools, colleges and local educational authorities are in some cases paying substantial sums of money to access other people's applets!!

Something you might want to consider for the longer run: the value of this educational software to educators would be higher if we could e.g. set quizzes in it. The storyboard aspect of Kojo looks perfect for this - if there were a kind of wizard so that teachers didn't need to get deep into coding to set a quiz, that might help. And you'd need some way of keeping tabs on kids' results - mymaths make this incredibly easy (which goes a long way to explaining their high adoption rate among UK schools), you just upload a copy of your classlist (most teachers have a copy in Excel), and it generates a list of student log-ons (they have to log-on using the school ID plus their own student ID within the school) and passwords, and then throughout the year you can keep tabs on when they've tried an online activity, what mark they got if there was a quiz, and so on. This is obviously really convenient for teachers! (You might even consider whether this aspect of Kojo could be a monetisable service that could fund core development of the software. But I think you're right to consider other funding options - there are a lot of educational grant-awarders out there, and the fact you're based in India and using this software in schools in deprived areas, ought to win you a lot of ticks!)

Incidentally: the Picture stuff is great and contained several featres I'd also been looking for - ideal for e.g. Islamic tile tessellations, an application often used to help present symmetry and transformations lower down high school, or elementary group theory at the high end. GeoGebra is also a powerful tool, though, and shouldn't be underestimated. I was very impressed not just by its GSP-like aspects, but found its graphing capabilities were very strong - competitive with expensive, commerical educational software, and good enough to cover most needs of a high school or technical college. (The biggest hole that I can see is in statistical rather than algebraic graphing, where their commercial rivals like Autograph are well ahead.) GeoGebra has a small but active and growing (especially now it's being used in teacher-training courses) community of educators using it, some of whom are finding the inability to write more sophisticated code a frustration. Kojo has brings something new to the party here, and in fact this is how I discovered Kojo in the first place. I'd suggest GeoGebra4 is a substantial improvement on GeoGebra3, particularly in terms of geometrical objects - I don't know whether a new GeoGebra version is looming, though, in which case it might be worth waiting for that to come out! Your comment about using applied math to produce something whose aim is not explicitly mathematical, is interesting - it really ties in with "learning through doing". On the other hand, for things like interactive demonstrations/tutorials, the maths generally needs to be explicit. From a teaching point of view, it's great when students get to put maths into practice, and develop their skills and understanding by tackling a problem that isn't just "math for maths' sake". But on the other hand, there's a lot of pressure to deliver syllabus content, and to create resources that aid understanding. An interactive demonstration is definitely better than an animation or video, which is in turn likely to add value compared to just using the textbook. Kojo's "mathworld" looks perfect for that.

I note there's a new ebook out so am hopeful there is some activity going on here still.

Yes, there is definitely activity going on in the Kojo Project. If you look at the Kojo blog, you will see that a lot has happened over the past few months. And, as you noticed, there's a new (beta) ebook in the works - which I keep updating fairly regularly.

But here's a fact of life. Kojo has one regular contributor (me). Other folks come in, contribute, and move on (till they come back again, possibly much later). So Kojo's progress is steady (I plan to keep working on Kojo for a very long time!) but not necessarily fast.

The utterly fatal problem is lack of documentation.

Well, at this point we have:
- A Kojo intro book
- An Adventures with Kojo - Level 1 - book
- A Learning to Program with Kojo beta book

I think that's a decent start. But as you point out, what we have is not good enough yet for a school system.

So let's fix that.

The needs of instructors need to be considered separately: only a tiny proportion of us are going to be "power users" or potential developers, but neither are our needs quite the same as the students we hope to teach with Kojo.

How about if we start with you? Let's get the material in place that will make you comfortable with Kojo. Once we get to that point, we will have enough material to share with other teachers.

Obviously there's no training courses available for me,

Where are you located? We can explore the possibility of me coming over for a training session.

There's a second problem with Kojo.
I don't think it's possible to embed in a web browser.

This one has been on my radar for a while, but I have not put any active time into this yet. Let me ask you a few questions about this:

• Does Java Webstart sound like a reasonable solution? With this, you would have a link on the Web, and when you click on the link, Kojo would start as a Desktop Application right off the web, without a separate download/install step. (Java Webstart would, of course, download Kojo the first time it is run, but the user does not need to mess with any of this).
• Does a Java applet seem like a reasonable solution? (Applets these days are much more robust and poweful that they were 10 years ago, when they got a really bad name. But they don't have any buzz these days).
• Would it be acceptable to have a trimmed down Kojo-Player on the web (as a webstart app or applet). The Player would run programs just fine, but would provide limited support for development (no code-completion etc)

If you could give me some advice on how to get up to a "good" standard at using Kojo,

This ties in with getting material ready - with you as the primary audience. More on the specifics of this after I hear back from you.

I would love to help you build a "Kojo content community" which is clearly one of the things you need most at the moment.

That sounds awesome. I look forward to this.

First, please please don't charge for the ebooks.

That's just been something to pay for the hosting of the Kojo website (and some of my bills!). I'm very open to rethinking this as things progress. Maybe I need to go for a sponsorship model (with very modest sponsorship goals) to support my work on Kojo.

Secondly, as I said above, please during your development consider separately the needs of your "professional users"

If you mean this in terms of software features and not just the learning material around Kojo, I'd love to have specific feedback and feature requests from you.

The "mathworld" is wonderful, but seems less powerful than "raw" Geogreba.
Is this because Kojo doesn't use Geogebra 4?

Yes, that's right. Kojo uses GeoGebra 3.x, and I have not had time to think about the upgrade to GeoGebra 4. In fact, of late, my focus with Kojo has been on enabling the use of Pictures for creating drawings, animations, and games, with applied Math coming into the mix for powering all these activities. The GeoGebra part of Kojo (with its support for explicitly playing with Math) has been stagnating a little bit.

Regards,
- Lalit

The utterly fatal problem is lack of documentation. I am sorry, but it is just not possible to use this software in an educational setting unless (a) the students have access to a decent-enough help file that they can solve the "little problems" for themselves (if 2 kids in the class are stuck and have their hands up, I can cope, if 10 do the wheels fall off the session pretty quickly!), (b) the teacher has some way (and it has to be a relatively easy way - educators are stressed and frantically busy, so we tend to take the path of least resistance) of getting seriously competent with the software, so that we can design lessons (or borrow and adapt other people's) and structure the students' learning experience. The experience when using educational software from established for-profit publishers is usually quite pain-free: they often have excellent (extensive and well-written) documentation available in the software, as well as instructor's manuals for teachers to help with lesson-planning. And on top of that is a whole training and support ecosystem: the tech guys can call for product support; the publishers send out training teams into schools, colleges and local authorities; there is "cascading" of training, during professional development sessions, in which educators who have received more training or had greater experience of the software, share ideas and lessons plans, and coach their colleagues (sometimes from a cluster of local schools) in new techniques. When I'm faced with the choice of "shall I do this in Kojo, or something else?" this is the why the path of least resistance is usually "not Kojo", even though Kojo may be the superior as a software tool!

Some kids can experiment with the various options that come up with "autocomplete", and play around until it works. For very able kids that's fine…especially if they have curiosity and motivation. But there are lots of kids who just don't work that way. From a teaching point of view, a first priority is that an activity has to be accessible to the weakest student in the class. Also, it's important as a teacher to feel really confident that you know the software as thoroughly as you need to. For some of your commercial rivals that's easy for me: I got extensive software training when doing teacher-training, then top-up courses from publishers and local authorities, I've seen lots of brilliant ways to use the software innovatively, and on those occasions when in the middle of the lesson I get stuck on something (usually because one of the kids has found a brilliantly innovative way to make a mistake!) then I can usually get unstuck within 60 seconds, with some help from the documentation if needed.

With Kojo I'm not sure how I can get to that point. Obviously there's no training courses available for me, but that's been ok with some other software. (Hopefully in future there will be online videos and so on, and maybe even a "real life" training ecosystem.) But in those other cases I did have a good set of documentation to work through. There are some strategies I can take - e.g. teach very constrained, rigid, formulaic lessons where I stick to the bits of Kojo that I have the most confidence in (due to extensive playing around!) but doing that negates the flexibility of Kojo which is its real strength! There is a lack of material for educators (instructor's manual, lesson plans, or even just more pedagogically-inclined examples) but I could actually live with that - though I know for a fact, other educators will need such content to be available before they adopt. I love teaching, I love using computers to enhance understanding, I love sharing new ideas, I love open-source - for all those reasons I'd be VERY enthusiastic about adding to the Kojo content ecosystem, sharing code and teaching tips, and so on.

But it's difficult for me to get started: I need to find a way to get up to "better than an experimental beginner" status and I can't see how to. The needs of instructors need to be considered separately: only a tiny proportion of us are going to be "power users" or potential developers, but neither are our needs quite the same as the students we hope to teach with Kojo. I think until that's addressed, early-adopting tech-savvy educators whose interest lies in the educational applications of Kojo rather than developing it, are not going to be able to kick-start the sort of content eco-system that can then draw in other educators (who might like a cool shiny new thing, but only after they've seen some concrete examples of what it can do).

There's a second problem with Kojo. This one's not fatal, but will undermine take-up and makes it substantially less useful to me and other educators than some rival software. I don't think it's possible to embed in a web browser. In terms of features, rather than documentation, I think this is the biggest sticking point! It makes it much harder to integrate Kojo into Virtual Learning Environments, or get students to use it remotely (e.g. for homework, or if they want to use a Kojo animation or interactive learning activity to revise). Also, in terms of generating "buzz" about Kojo this is a pretty serious limitation - someone has to install the software to see just how good Kojo is! And at colleges/schools, please remember that most educators CAN NOT install software on the network (user-rights guarded zealously by IT departments) so it'd be impossible to show off to colleagues a "cool new interactive demonstration" you have designed. I have no idea how difficult it would be to get Kojo working in a browser, but if there is some possibility, it would be far more useful than any new feature inside the software itself! Because no matter how good your fancy new feature is, if I can't show it to colleagues then I can't convert them, and if I can't share it with my students via the VLE, then its educational use is limited.

If you could give me some advice on how to get up to a "good" standard at using Kojo, I would love to help you build a "Kojo content community" which is clearly one of the things you need most at the moment. If you could give me some direction about the possibility of Kojo being integrated in a browser, that would substantially change the type of educational material I'd focus on using Kojo to produce. (If it can't be used in a browser, then it's pointless to build animations or interactive demonstrations like the "sine graph from the unit circle" tutorial. I'd be better to do it in e.g. Geogebra for VLE integration and so students can see it at home when they're revising. An ex-colleague of mine actually produced essentially the same tutorial as part of a larger trigonometry e-learning package, and won an award for it. He used a MS Excel plugin because that's what our VLE supported. Kojo excites me because I can see it would be able to produce more powerful and useful interactive demonstrations than both Geogebra and that Excel plugin… but if it can't work in browsers, my educational use of Kojo would focus more on the programming aspect.)

Two pieces of advice and one more query. First, please please don't charge for the ebooks. It's counterproductive - if you want to monetize Kojo, you're far better to focus on growing a professional user community to service. Lack of documentation, particularly for beginners, is the biggest logjam to growth. In contrast, delivering professional training courses, and e-learning consultancy services to education authorities, can be big money! Secondly, as I said above, please during your development consider separately the needs of your "professional users" and your "student users". Unless Kojo draws in the educators, you won't get students experiencing its benefits! Bear in mind that not all educators are tech-savvy or feel confident in their use of computers, even if they hope to use them in their professional practice. Imagine how scary it must be for a teacher, who feels that students know more about computers than him, to use a new piece of software in class that he is not experienced with! Such professional users face a daunting learning curve, but it's not the same one as students do! Finally a query. The "mathworld" is wonderful, but seems less powerful than "raw" Geogreba. I can't seem to find some functionality: polygons and semicircles are useful but seem to be missing. Is this because Kojo doesn't use Geogebra 4? Or are those functions just hidden somewhere?

Best regards and many thanks for making such a wonderful piece of software available for free!