Very cool! I would be interested to learn about how you solved the hidden line rejection problem. My approach uses a z-buffer to draw from near to far, and computes a "shadow" representing the obscured area as I go. There are cases that don't get handled well - notably lines with ambiguous depth.

Love your work! Does this start as a vector image or are you generating vector lines from a rasterized image at some point?

Cool idea! I wonder a different grid system would make the word searching part more intuitive. It would make indexing in your code a lot harder, but a triangular or hexagonal grid may help by changing the number of available neighbors?

Wonderful! The little shadows really add a ton

Thanks!

Do you have any insight into when its best to move from exploration to implementation, or bounce back and forth? My sense is that this awareness is gained through practical experience in different problem spaces, and over time you learn to "feel" it. This makes it a tricky process to teach though!

To me, the advice to fully explore the problem-space is most useful for novices who need to be deliberate about not committing to a solution too soon.
On the other hand, experts can often do this exploration work implicitly by drawing from relevant experience to inform viable directions forward.

I think the ability to effectively switch between different design spaces is something that is developed over time, with expertise. Novices and learners tend to become quickly attached to early solutions because they offer solid ground in uncertain and ambiguous phases of the design process.

Oh that's awesome! Unfortunately, the video compression is really getting in the way...

This one is tuned a bit better - very happy with this result!

(you didn't ask but) the pitch of an ocarina is based largely on the "effective size" of the sound hole. Uncovering a finger hole increases the overall "effective size", and changes the pitch. So, if you change the size of the finger holes you can tune the notes! More holes should increase the range

This is so inspiring! Thanks also for all of the wonderful detail in your blog - it is so nice to be able to learn from your process www.amygoodchild.com/blog/cursive...

You can also bend the bit of wire sticking through the hole toward your potentiometer pit to create a little bridge that the solder will stick to

You'd still be bridging with a pool of solder, but just to the neighboring hole that then has a wire attached.

Solder the pots to the board, connect the traces together with wire - best of both worlds.

Perf board is always a bit of a hassle, but you do need the traces/pads to solder to regardless. I prefer using wire jumpers instead of trying to bridge traces with solder because of the problem youre facing - the solder is designed to not stick between the copper bits so you do need to pool a bit.

Oh I see now - those larger case pins might not need to be connected to anything for the pot to function if you already have a ground pin and your application isn't super noise sensitive. If they are a hassle, you might be able to bend them out of the way and ignore them altogether!