I’m not sure how, but at some point I came across this Instructables article on building models from maps. The article shows you how to use Terrain2STL and Kiri:Moto to get a portion of a map, generate a elevations file in STL format (originally designed for stereolithography, it’s a format supported by lots of 3D programs and tools), and convert that file into topographical slices.

So I revisited the House Mountain area, and went through the process. I chose to exaggerate the vertical considerably to make it more identifiable. The tools yielded me an SVG graphic of all the layers. I did further conversion, and cut the scene out of chipboard using CrashSpace’s Epilog laser.

I lost many of the finer peak tops into the interstices of the laser cutter. Even the ones I did manage to keep were difficult to glue. I’d use a magnifier and tweezers if I were to do it again.

I’m tempted to 3D print the STL file on a filament printer. The output would certainly be smoother and more detailed.

In ye olde days, I designed stuff in POV-Ray to render whatever fantastical scenes I was imagining. I’d spend hours figuring out textures and constructive solid geometry to create images. It was a slow process. Files were extremely slow to render. On my trusty Intel 80386-based PC running DOS, a scene of any complexity would take all night to render at 640×480 pixels.

Now, 30-some-odd years later, I still play with a constructive solid geometry modeler — in this case, OpenSCAD. The idea is that I could output the models to a format like STL, and then 3D print them into physical being. I haven’t actually done very much printing of models, but it’s an interesting possibility nonetheless.

Below are some images from a work in progress. I was inspired by seeing the Breugels painting above in a YouTube video. The tower is not only a great metaphor, but an interesting image and architecture.

My architectural thoughts go more Gothic (more flying buttresses), and parametric. By parametric, I mean that I figure the design can be based on a set of variables, for example, the ratio of height to width of a wall segment. For each value of the variables, the code can generate the appropriate geometry.

My ability to create this way is limited by two things: my trigonometry is not particularly strong, and my ability to keep a stable 3D point of reference in my head is even worse. So I start with sketches and pages of cosines and arctangents, and then end up doing a lot by trial-and-error. Because thinking in this mathematical space is hard, I end up getting frustrated and putting the project aside for days or months before picking it up again. Not to mention, even with today’s super-fast computers, as the complexity increases, the time to render an image increases!

So, my tower of Babel is not complete. There’s been some progress. I played with it a little today. Maybe one day I’ll finish it. Perhaps I’ll even print a model.

I’ve hooked up this blog to the fediverse. The improved visibility will doubtless increase readership from none to the lofty realms of zero.

Posts will federate as posts/toots/notes from @admin. Presumably, comments/responses will propagate back somehow too. I could read the manual, but I’m just going to sit back and see what happens.

I have tens of thousands of photos I’ve taken over the years.

I think some of these photos are pretty good, but most are languishing unseen on random hard drives.

To share them, I’ve been a member of Flickr, I’ve posted on the late lamented Twitter, and I post some onto Mastodon. I’ve also created numerous gallery applications/server scripts/web sites (e.g., Statgal), but they’ve generally been clumsy or take too much work to maintain. So I’ve been working on a dumb PHP/JavaScript slideshow thing that will scan directories, cache the details, etc.

Here’s introducing PhotoSpinner. It’s a quick’n’dirty script to provide photos. It’s very simple and allows me to publish categories of pictures without a lot of effort. Source code’s at Codeberg.

My math is abandoning me. I started out with the design below, which can be used to tile a plane. My thought was that I would rotate the “blade-like” hexagonal elements, since their shapes mesh almost gear-like when rotated in opposite directions. Behind them, the pinwheels would also rotate in alternating directions.

It’s not hard to convert a hexagonal tiling into rectangle tiles that will evenly cover a plane. In this case, drawing a rectangle from the centers of the the blade-like elements at 2, 4, 8, and 10 o’clock will work. Ah, but not necessarily if the elements are rotating! The rotations won’t match up. I’d need to expand the pattern and cut a larger tile from the expanded pattern.

Of course, it also quickly became obvious that I need to consider it a triangular tiling, not the hexagonal tiling that was in my head. With a triangular tiling, I can’t have the blade-like elements rotating in opposite directions because you run into this problem:

Naturally, any odd-numbered circuit will have the same issue. But say we treat the center blade-like element from the initial design differently, and replace it with a different shape, say. The six elements around it could rotate in alternating directions.

You’ll no doubt see where this is going. With this design, I could tile the plane. But what if I wanted a variation of this where the center elements, different though they were, still rotated? I end up back with a variation on my triangle problem.

So after that, I briefly considered going back and altering the original design, but instead decided I didn’t want to animate it after all.

:shrug: