So it would seem that you are not actually having the same problem as me
I'll experiment some more
[edit1] My Chrome was V102 - upgraded to 103 but still same behaviour
[edit2] Just tried using Vivaldi from one of my Pi and that works so it's looking like just an issue with my Win11 machine at moment - don't you just love computers
[edit3] But my Pi software is fairly old (> 1 year ) so it it might be working for that reason
[edit4] Just tried Chrome on my Win10 laptop - that doesn't work either
The issue isn't your Web Serial blocks - it's the use of JavaScript in the menus of the other blocks
I think I can get around that but I'm fighting battles on many fronts at the moment and I'd like to get local LAN webserver access working again as I use it in other projects
NOTE: After loading the Microblocks script into the WEMOS D1 mini, remember to close Microblocks before running the Snap! script otherwise you get a clash between the two programs over the com port
Obviously, each of the scripts could be made to look nicer
Very interesting. The WEMOS microcontroller seems difficult to source, particularly in volume for classroom purposes. Other than eBay, have you encountered any other retailers stocking them with any regularity? You also mentioned the NodeMCU. Have you tested that yet? I may order one and give that a shot as well.
I followed the UK Amazon link that you provided. These items are not available to customers in the U.S. Not sure why they can be obtained in one location but not the other.
Excellent! Now that the Microblocks script has been developed for the Wemos, is it feasible for Snap! on a PC to talk to MicroBlocks on the microcontroller?
The discussion about wireless is just an artifact. I noticed that Raspberry Pi had released a Pico W (for wireless) for $6 and inquired about the possibility of controlling it from Snap! Simon recommended the Wemos D1 mini as a better solution.
However, Bluetooth or even a direct connection via USB would be fine. In another project, Rich Nguyen is looking into the possibility of using long-range LoRa chips (LoRa - Wikipedia) for remote sensing, but we're currently not doing anything with motor design and fabrication projects that requires remote control. Since the schools primarily use Chromebooks, any solution does need to work with that platform.
It's great that Simon was able to make a MicroBlocks library for the Wemos D1 motor shield. It's too bad that there doesn't appear to be a US distributor for that motor shield. Their official distributor seems to be the Chinese company Ali Express. Unfortunately, it is difficult for many US schools buy stuff from either Ali Express or from EBay.
The NodeMCU shield is more readily available than the Wemos motor shield in the US (e.g. from Amazon). Unfortunately, it's expensive at $10-12. On the up side, since it is a shield, it creates in a robust physical "package"; no loose wires that could get pulled out.
Good to know! That is a very low-cost option. I see that it is available for $1.99 from a California company called Banana Robotics. This board could be combined with the Raspberry Pi Pico for a total of about $6 -- plus the cost of the jumper wires.
Independent of the hardware, there are a number of equivalent ways for Snap! to control motors: Snap! sending commands to the Python REPL, Snap! sending commands to MicroBlocks, or Snap! sending commands to an Arduino program. The transport could be web serial or WiFi. All of these solutions work with Chromebooks.
Setting up the firmware on the board may be easier for some of these options than others (e.g. MicroBlocks), but that may not be an important consideration since the firmware can be pre-installed by the teacher and students can just plug in the board and use it from Snap!
That said, I'd still argue that MicroBlocks is the nicest language to run on the board itself because it is a blocks language. Students who are curious -- or teachers who want to extend the firmware capabilities -- can do so more easily in MicroBlocks than in Python or Arduino C++.
I'm sorry I can't organize a SnapCon BoF around physical computing. I'm teaching a PD course for K-12 teachers the directly conflicts with SnapCon! However, if someone else organizes one and the timing works out (e.g. if itis on Saturday) then I will try to join.
I agree about shields. We have gotten much better results with groups of children when a shield reduces the number of wires that may be pulled out.
We don't (yet) have extensive experience with MicroBlocks, but I can say that use of native Arduino sketches at the middle school level has resulted in a high percentage of students in local schools dropping out of the STEM track due to frustration. It stands to reason that if they're already successful with Snap!, the extension to MicroBlocks should be a more seamless transition.