Thursday, April 05, 2018

Thoughts on where CS education tools come from

TouchDevelop is being retired. It’s not the first tool with educational use to be retired by Microsoft. Or from other companies either. While I have used TouchDevelop with my students I was getting ready to move to MakeCode anyway so it is not a fatal move for me by any stretch. I know other teachers who have been using it more extensively who are going to feel some pain though.

I tend to divide CS educational tool sources into four categories.

Professional tools that are useful in education. In this category we see things like Eclipse, Visual Studio, and other development tools that are being used to create real products. These tools are not going anywhere anytime soon. Companies, both those that use them and those who create them, have a vested interest in supporting and improving them.

University developed tools. This includes things like Alice and Scratch but also BlueJ and Snap! and, well, far too many to mention. These are also likely to stay around for quite a while. People tend to build careers around them and use them to attract graduate students who use them, enhance them, and generally want them to be around.

Education companies. Companies like Tynker and Birdbrain are included in this group. As long as these companies can make enough money to keep going their tools will be around.

Industry Research Projects This is often the most state of the art and cool new things. On the other hand they are the least to be around for a long time. AppInventor was originally one such. Fortunately for educators who adopted it, when Google was done with it they were able to hand it off to a major university where it continues. TouchDevelop is another. Hear though, when the researchers moved on to new ideas there was no similar handoff. Both the Touch Develop web app and cloud backend are open source under the MIT license:

There is no indication that anyone is going to take it on though. I always assumed this day would come to be honest. Industry research groups have changing priorities and in some ways that is a good thing. I could argue it is a very good thing in terms of advancing the state of the art in computer science. It’s sad for us in education who aren’t always as quick to adjust as industry is.

As one tool fades another grows. MakeCode is pretty cool and there is a lot there learned from TouchDevelop, the Micro:Bit and educational uses of them both in there. Microsoft is developing a sprite-based game engine for MakeCode and a new course on game programming, in the spirit of CCGA (an interesting curriculum based on TouchDevelop.). One hopes it will be available in some form for teachers to learn it this summer.

I’m not sure where the materials developed by code.org fit into my classifications though. They are a non-profit with industry funding but they are neither industry or academia.  They have some great stuff though. I am using their CS Principles program for example. Chances are I’ll retire before they “go out of business” so I can probably depend on them plenty long enough.

Nothing lasts for ever though and change is the most constant thing in computer science. So we have to learn to adjust and change our tools and our curriculum, probably, a lot more often than educators in other disciplines. At least we’re not going to get bored.

3 comments:

Mike Zamansky said...

Funny - I was thinking about writing something up about tooling today as well. We'll see if I have the energy. Specifically, I was planning on writing about the complexity of modern dev toolsets and how we are, or aren't preparing students to eventually get there.

I think you also missed one category - composite OS tools - editoris like Emacs or Vim, the command line as a toolset, tmux etc.


Garth said...

I have been teaching the same math for 30+ years. My CS curriculum lasts maybe 2 years without a revision or at least me thinking about a revision.

What bothers me about the death of Touch Develop and CCGA is the suddenness of the demise. Microsoft invested a lot of time and money with building the curriculum and in-service training to get teachers to buy into CCGA, then to kill it so abruptly just seems odd. It makes me think there is something going on in Microsoft land that we little people do not know about.

I still want Project Spark back. I want the kids to build their own version of Fortnite.

Alfred C Thompson II said...

Garth, like you the abruptness of the death of TouchDevelop is bothersome. I agree with comments that you made elsewhere that had someone with real classroom experience been involved in the decision there would have been more warning. Companies don't think as long term as education has to think.