In no particular order:

Selworthy is in at the bottom of the ‘writing a difficult feature’ hole. Lots of chipping away, testing, QAing, re-testing, and re-approaching the problem. By Week 243 I’d had a bit of a breakthrough – some things were pretty much there, and other problems were, at least, understood. I also spent some time – possibly more than I’d like, given my enthusiasm for operations – wrangling getting the whole stack spun up on Centos – we tend to use Ubuntu, but there were reasons. Seeing the whole stack – from Javascript polish right down to compiling forks of ffmpeg – gives you a bit of vertigo.

Thonk have ordered a second run of products from Foxfield, which is great news. I took an inventory of what parts I had, ordered some key bits of the BOM from China, some more from a few British suppliers, and started putting together what was needed for a Mouser order. I’m hoping to package this lot up by the beginning of September. I also wrote the first installment of Foxnotes – similar to these weeknotes, but more infrequent.

I’m running a course on circuit board design in October, over four evenings. I did some promotion of this, and have spent a decent chunk of time beginning to write the course itself – trying to get a feel for what it contains, what the balance between talking and doing is, how to structure learning across the four weeks, and triple-checking the demo circuit I’ll be going over. I also ordered various small prototypes to help me out with demonstrations.

The course is going well, I think – I’m beginning to see the shape of it, and to see places we we can extemporise or build on things we’ve learned. But I also need to remember how much space to leave for unknown unknowns – discussion, and questioning, and inevitable inertia of using software.

A good few weeks – busy, and outside Selworthy, running around a little. But when I look back at it here, there’s definitely forward momentum.