Great page to consult Ogham and any other writing system

omniglot.com/writing/ogham.htm

Ogham alphabet

The Ogham alphabet was used in Ireland and Britain between about the 4th and 7th centuries AD to write Irish, Welsh, Latin and Pictish.

^{www.omniglot.com}

Sensitive content Reveal/hide

@glc

Then it’s a good thing I am speaking in contemporary English and not historical terms. 😂

@glc

Not when the contemporary English is fully established malicious history-distorting propaganda, it isn't.

Learned about that one in a college geology class. Summary: don't be one of the 150K people living on the lahars from past eruptions.

Sounds like one county has evacuation drills and limitations on development in the danger zone, but another doesn't. And insufficient funding for any of it.

@me_valentijn
Yep, it's a mess.

The world is full of people putting houses where they oughtta damn well know they really, really shouldn't.

That really stuck with me after years of hiking with my parents, and them pointing out the geological and forestry details so many times, "Yep, this is a flood plain." etc.

@dodo1095zd@mastodon.social

Blocked and reported.

(His response was ~ “first release the hostages…”)

Asshole.

It strikes me that you can make very high quality surfaces using only a low fidelity CNC router and a high precision flatness measuring technique.

What you do first is modify the CNC router so that it is able to spring on the Z axis, so that Z position in fact corresponds to downforce.

Then you need to derive a surface height map, and put that into an algorithm which can output a polishing tool path. The correct amount of downforce and time for removing a specific number of microns must be determined, for the tool, grit, and material.

But with that information known, you can put the map of high spots into a tool path plotting algorithm and it should give you a tool path that is believed will knock down the high spots.

Assuming the algorithm does actually remove the high spots and doesn't just make the problem worse, you can zero in on the goal by switching to more and more gentle polishing tools.

When you're down to nanometer scale diamond grit on a fuzzy pad making a mirror finish, you've basically gone as far as you can go.

You probably need to build a small random-orbital tool for the router because simple rotation does not do the right thing.

The real limitation on laser angle measurement is dynamic range. If you put a 10x scope in front of a camera with a 5mm wide optical receiver, that's the equivalent of now having an 0.5mm wide optical receiver. But you only need 1 meter of distance for better than 1/10th of 1 arc-second graduations, which is very good.

Arc-seconds are small, about 5 microns per meter. If you had a mile long wrench and you pushed the far end of it by about 3/8ths of an inch, that will rotate the bolt 1 arc-second.

To aid in alignment, there can be a cover over the aperture with a hole the size and shape of the aperture itself, so that if the laser is missing the target, it is obvious whether it's missing left, right, high, or low.

Also 1/2 mm is a really small aperture, so you would think the laser also needs to be a 1/2 mm beam, but that's not how lenses work. The lens takes every ray of light with the same *angle* and directs it to the same point on the receiver, so the laser only needs to emit *parallel* light rays, it doesn't need to emit a point. Industrial lasers with adjustable focus are like 5$ online.

Putting a cover over the lens will still work because the lens tells the camera sensor exactly what *angle* the light comes from even if it's only catching 10% of it. In addition, it can help keep stray light out of the camera, and avoid the laser light from washing it out.

Minor errors in focus are a blessing in disguise because they spread the light over more pixels which gives you more confidence when you solve for the where is the center of the light.

In order to establish that a surface is flat, you need 2 optical receivers at a right angle from each other. In theory you could setup one and then move it, but it's not worth it to not be able to come back to your measurements because you dislocated the receiver.

To realistically set something up for measurement, the receivers need to be able to be adjusted up and down, but this does not require precision in any way, and it's just meant to bring them close enough that the laser hits the aperture, then they can be locked in place.

The measurement process would begin by leveling up the workpiece by eye, then verifying that the X receiver can pick up the laser at the 4 corners of the work - or at least 3 of them (in case of a very high corner). Then, without moving the part, you adjust the Y receiver until it can pick up the 4, or at least 3 corners that the X can pick up. Then you just start working over the part, collecting points and angular deviation which can be resolved into a height map.

Putting a "normal" camera above the workspace can help because then the computer can know where on the work you measured when you got a specific measurement, no need to manually record the results...