What is best practice for reference points on a face?

[esr]

"the maintainers" means among others: you!

Yes, in a very small way, now. I have actually contributed to the documentation, and I expect to do more. It would be pretty foolish to assume that all newbies will intantly do likewise, though. Especially since programmers notoriously hate writing documentation; my degree of willingness to do it is extremely unusual.

(As for code, I hate C++ with a passion. OTOH if I find something worth doing in Python I will dive right in. It's too bad the project core isn't in Golang, I'd be all over that.)

This reveals that you are on the wrong track in a free software project: It is extremely egoistic vs. doing things for the community.

Who said anything about community? I write high-quality documentation for my projects for two perfectly egoistic reasons: (1) to reduce the amount of yells for help I get, and (2) out of pride in my work. I would be flat out ashamed if any of my projects had to feature a warning like FreeCAD's about its tutorials - I'd find it intolerable, a public admission that I'm a sloppy and careless craftsman.

But that won't happen to me, because I consider it part of my maintainer responsibility to (a) keep the documentation in sync with the code I modify as I modify it, and (b) skim my tutorials before each point release to make sure I'm not shipping any falsehoods.

You can of course choose not to be this careful. You can choose to be sloppy and careless about the code you write, too. I can't make you have higher standards than you choose to, but I will say that I think you should be ashamed of yourself for defending low standards.

The statement in the tutorials holds for each video tutorial on youtube as well, and they are probably never updated at all.

Yes, and the FreeCAD maintainers aren't responsible for them, either, so they don't reflect badly on you when they are stale. If you have a point here I am failing to see it.

Instead of working through the tutorials and learn the very basics you are asking for personal training by the forum regulars instead. Without offering anything! I see this as an insult against the busy forum helpers.

If that's a swipe at me, I will point out that "Without offering anything!" is already false. I've added some documentation to Spreadsheet, and I'm accumulating a list of dangerous curves to be marked out when I figure out how to write an admonition template. When I learn something from a forum conversation I try to figure out how to add it to the documentation wiki - the next item will probably be on how to force selection of the arc less than 180 degrees on setting up an angle constraint.

If it's a general complaint about newbies who don't read the tutorials, why do you expect them to care what you think is insulting? They have work to get done. They're going to take what they judge is the least-effort path to that. If you don't tell them straight up that your tutorials are likely to be faulty, that may be reading the tutorials. If they think it will be easier to bug people in the forums, they'll do that instead. Humans are going to human, and your conviction that they "ought" to read the tutorials is about as relevant to J. Random Newbie as your taste in cheeses.

[ifohancroft]

All I'm doing is making a couple of pockets in a beam so that two copies of it can mate end-to-end via...it doesn't seem to have a standard name, but I think of it as a "handshake joint". It's a variant of a lap joint that is better against pulling stress - I invented it after watching some videos about Japanese carpentry.

It would be cool to see it in the FEM workbench to see how it behaves under different stress conditions.

In case you're curious, here's the assembly with the handshake joint. It's on the beam end opposite the tenon, negative-X direction. Explanation of the parts and loading stresses follows:

The tenons fit in the small holes in that truss-like piece with the curved top (top is positive-Y direction) - there are two of those trusses in the assembly. There need to be four of the beams connecting them, in two pairs with handshake joints, because a 459mm solid beam won't fit in the print volume of the 3-D printer I have available.

The parts are a prototype for a drilling jig - that curved top surface will have 2mm neoprene sheet foam glued to it. It's intended to support and stablize the curved plate assembly of a vintage Model M keyboard while the rivets are being drilled out for bolt-modding. This one is for M101s - if it's successful I may remix it for M122s, which is why it's carefully parameterized.

Explanation of the procedure here: https://wiki.geekhack.org/index.php?tit ... d_Bolt_Mod

The handshake joint should be acceptable because in normal use it will never see any tensile stress. The jig will be compressively loaded in the negative-Y direction with no more than about 3kg of force at a very pessimistic estimate. I almost certainly could have gotten away with a simple lap joint, the acrylonitride glue I'm planning to use at the joint probably has more than a 3kg breaking strain, but I liked the idea of designing a joint that would recruit the strength of the beam material as well. The parts will be PLA or PETG.

The design isn't finished yet. I'm going to add a pair of compound tilt blocks that can be inserted under the jig to tilt it in 6-degree increments up to 18 degrees. I'm using a simple vertical drill press, but it's desirable for the holes to not be far off the local vertical of the plate curve - thus, I tilt the jig. Maximum 3 degrees of off-vertical deviation is acceptable.

I learned basic FreeCAD exactly so I could do this. It is only my second mechanical-design project ever, and the first one was so trivial it was basically "Hello, world!" If I've done anything obviously dumb or newbieish, feel free to instruct me.

Thanks! That's cool!

I have a DEC LK201 to try and bolt mod (it uses a similar structure to a Model M), but it doesn't need drilling. I don't know if it will work though as it has plastic "structures" that hold the keycaps with very thin legs, so if I don't know if they can even hold a screw. It would have to be something smaller than an M1 screw.

I don't see you doing anything newbie-ish. When you have the design finish, specially if you do it as an assembly or even just positioning the parts so they all fit into the 3D space, I think you may find it easier to use the software's vertical axis for a vertical axis (that is the Z axis), so essentially whatever is sitting flat is sitting on the XY plane and is extruded upwards in the Z direction.

Btw now I see why you called it a handshake joint. That's clever.

P.S. It would be cool if you release it as open source when it's done. Whether it's on GitHub or GitLab. Perhaps even post it on Deskthority after releasing it, if you still go there? Btw not sure if you heard about the sell of Deskthority?