HS_NOTCH

[ja2ui0]

@joshreve mentioned in the config generator code that HS_[NOTCH|UNDERCUT] does not generate properly and the hot swap step needs to be modified.

I generated a HS_NOTCH case this afternoon and I want to make sure I'm seeing the same thing before I go down this rabbit hole. To the untrained (possibly including my own) eye, it looks pretty good. Though on closer inspection the only thing that jumps out is that the sides of the hot swap tray come in closer to the key body than the depth of the notch. Is this the part that needs to be modified, so that the notch depth is the important bit and the hot swap tray doesn't interfere with that?

I have not handwired one of these yet, so I can only trust that the orientation of the sockets is fine (particularly around where the thumb cluster is concerned). I'm sure it's a tight / awkward fit with an iron no matter what when you get near the case edges, but let's not get into the particulars of rotating switches at this juncture 😄 - just so long as none of the rotations are absolute garbage from a wiring perspective.

Assuming all of that is in order, I can work on that step file and submit a PR shortly, just let me know if the bit I mentioned is the part of concern or if there's something I'm just completely missing.

[joshreve]

That was the challenge. Originally it had a third wall blocking the mount notch near the hotswap. I actually half fixed them after that comment by removing the third wall. I'm just not sure how strong it is without the added structure. It should theoretically work, but I'm really not sure how it will print. If it prints and works well, I'd love to remove the warning, but at the moment it is more of a "not sure" rather than a "guaranteed to fail". I added rotations in the cluster to orient them if you have issues, but that is a WIP setting.

You may want to print just a thumb cluster as a trial to check the fit in that area and the notch settings as well.

[joshreve]

There is a setting to clear space behind the socket to make sure you can mount a pcb. It looks like either it is set on the cluster only or it is a bug and the seeing isn't applied consistently. I will look. I'm guessing that is why it is present but trimmed back in the cluster.

[joshreve]

Double check this is set
​    ​'plate_pcb_clear'​: ​False​,
I'm guessing it is and I'm clearing in the cluster anyway due to a bug. I'll take a look.

[joshreve]

Checking the line, you are fine to comment that out. It should be within a conditional based on the parameter in my previous response. I will fix it moving forward. In the meantime your change shouldn't cause any issues since it is exactly what I thought it was and is not used far any other purpose.

[Looooopy]

Thanks for that

[joshreve]

Threw the bugfix in. Should be commented out in the master branch now.