My current suspicion about the laser being wonky is that the arm it is mounted on isn't particularly strong. It isn't meant to hold much weight, it was originally holding a pen.
So it can bend just a little bit and that may be enough to throw off the beam angle after it cuts into the material just a little bit.
This is solvable, but may require a new frame.
I can't figure out what our laser is doing. When you start a cut at the surface of some wood quickly it cuts wonderfully with a nice fine line down about 2mm into the wood with 3 to 4 passes. And then it doesn't really cut past that even if you let it go for 20 or 30 passes.
But it won't really cut deeper.
The weird bit is that this seems to be the case when you change the distance between the lens and the wood by up to around 4mm. I have no idea what is going on.
I suspected that the events people wouldn't have any of the support hardware needed for the demo, like cable adaptors and a keyboard and mouse to do the initial setup.
It appears that I was correct. Luckly I got josh to bring all the hardware needed to run the setup. So we get to be the people who actually know what is going on despite everyone else being experienced events people and we are the new guys.
The current build of from our directly paying job is going to have its first public appearance starting tomorrow and I am not going to be there. Despite trying to never have one person handle a conference again Josh is going alone, but it is just to help setup then he is going home.
So not too bad.
But I am still nervous as fuck.
it looks like what I was missing was that the svg arc paths are made to be consistent with other path elements, not to be understandable.
Which makes sense, but it is still annoying.
Todays plan is to fix the arc handling for SVG parsing with the sinister plotter and add support for DXF files.
2d DXF files are very very simple to convert into strokes for plotters, which is by design, I think that 3d dxf files aren't really any harder to parse, but the sinister plotter is currently only 2d so that will have to wait for latter.
And I think that I am missing something big about how SVGs encode arcs, it seems unnecessarily complicated to me.
Ok, I am giving up on parsing postscript files. It could be really simple but there aren't any parsers that I can find that give me the strokes in a flattened format.
But DXF files are easy. So far we have SVG and DXF files as inputs for the Sinister Plotter.
Now I have to figure out what else to add to it.
And in todays bit of internet weirdness, postscript is not only a turing complete language, it can be used to do stuff like control electronics. This may have to go into my list of future things to play with.
well, it is a turing complete language. If I can find a reasonable list of what each command does parsing out the bits that I need may not be too terribly difficult.
We have cut through 3mm plywood with the frickin laser!!!
So for 'testing' we are going to make a bunch of puzzles and stuff.
the EPS definition language is an almost complete programming language. I am not sure if it is turing complete, but it can display output and do math at least.
I was not expecting this.
This will make getting vector paths from the EPS files either much easier than I expected or ridiculously difficult. We shall see.
I in no way understand the asyncio stuff in python, but it works so I am going to call that a win and let the sinister plotter do its thing.
It looks like this account will get more use. I will probably move most of my tech stuff here. Javert and most other things will stay on @inmysocks@kitty.town I think.
I don't exist!
I may be the same inmysocks you see on mastodon.social.... Maybe.
Whatever pronouns you feel like? I would be amused if you alternated.