2025 Holidays Wishlist with the BBC micro:bit!

PC's didn't even exist when I started. SBCs didn't exist either, but the first SBCs only had 1KB of RAM. Eventually tape recorder interfaces were developed. I just bought some FLASH at about $1 per terabyte.

The early PCs I worked with had 1MB RAM and a 20MB disk.

How times change. What's the minimum spec going to be 30 years from now I wonder.

If you do get a copy, then remember to reset your mind back to the mid 1980s before reading.

The book covers the history o f robots, how they were depicted in the media both as good/bad, their use in the real world (in the 80s), the technical aspects, and then the last half is projects that you can do yourself without significant cost as you may already have an electronic toy or constructor set which has most of the parts. There is also a bit of freedom to explore further there in that for example the draught playing robot has an XY plotter mechanism and a conveyor belt which the author points out could also be used to simulate a warehouse type system.

The author had a real inclusivity challenge in that last section as not only did people own different home computers but also they likely owned different constructor sets e.g. Lego Technic vs Fischertechnik, or electronic toys. Trying to come up with projects to allow everyone to participate using what they had would not have been easy but I thought the author did well for an introductory book.

Perhaps keep in mind that this book predates online shopping so getting your hands on specific parts wasn't always easy. Also perhaps keep in mind what robotics kits were around at that time and their cost. Sure, If you had the money you could order up a Heathkit Hero 1 from Maplin Electronics but I never managed to get one for Xmas... The Fischertechnik computing and robotics constructor kits were also available but once again not that cheap. There were robot arms aimed at the home computer market like the Colne Armdroid but not easy to modify and not cheap, and had a tendency to stretch its cables a lot making position repeatability rather hit and miss. I recall a demo of one being used to try and repeatedly put a key in a car door lock and at the start of the evening it did ok but by the end of the evening the car door was about ready for a respray. (As an aside, apparently machine learning algorithms today are still struggling with keys and door locks).

Modern books of a similar style might be Tom Igoe's 'Making Things Talk'  or even Lorraine Underwood's 'Saving the World with Code' where she has inclusivity by including projects that can use either the micro:bit, Raspberry Pi, or Circuit Playground Express. I guess that there is always a bit of a risk with inclusivity like this in that two-thirds of the book may end up not being of interest to the majority who only have the single platform to work with. On the other hand, if you write for a single platform, then you risk losing two thirds of your audience.

In college, I built a Pentium MMX 200MHz, with 64MB RAM and a 20GB disk, solely on spare parts. Used it for college works such as MatLab modeling, pSpice simulations, and Word for reports... As long as I didn't run these applications at the same time!

My first computer was an 80386 which I used for my first experiments using C, Pascal, Visual Basic. Then I changed PCs several times and in the early 2000s I used my PC, a National Instruments acquisition card, Labview and Matlab to acquire a sound signal fron a piezo sensor and classify it using a neural network.

The learning phase of my poor neural network took many hours using Matlab, today I do the same thing in my school, during a single lesson, simply using an Arduino Nano 33 BLE Sense or an Arduino Nicla and Edge Impulse board for the network implementation neural. I have never used a BBC micro:bit but I have seen how many applications and examples we can find on the Internet and I would like to explore its features with my students