How do you get a screen reader to work with PrusaSlicer?

The unlabeled buttons aren't a setting you're missing. Most of the controls in PrusaSlicer are custom-drawn graphics rather than standard widgets, so screen readers get nothing or just 'button'. The tab order being a mess is part of the same problem. I went through this exact thing with a blind friend at our makerspace, and it took some experimenting to find a workflow he coud run on his own. What worked first was stripping the workflow down hard. I set up profiles for his common materials once, with him listening over my shoulder, and then his whole job became open file, pick profile, slice, export. The slice and export shortcuts live in the menus so they do get announced, and the profile dropdown is readable. He memorized maybe six keystrokes and could do a basic print start to finish without me. Anything beyond that, like support tweaks or per-object settings, stayed out of reach though.

The bigger win was the command line version of the slicer. Same engine underneath, but it runs in a terminal, wich screen readers handle way better than a graphics-heavy app. I wrote him a couple of simple batch scripts, one per material, so he pastes in the file path and gets gcode out the other end. He does most of his slicing that way now. It's not as pretty as clicking through tabs, but it's fully independent, which is what your brother actually wants. Also, there's a thread on this forum right now from a fully blind user who wrote an open letter to Prusa about exactly this and built an open source accessibility layer for the slicer. Worth reading with your brother. People in that thread said the company has staff who watch for accessibility issues, and since the next slicer version is in development, feedback landing now has a real chance of getting built in rather than patched after.

This is helpful because my dad is losing his vision too and I had no idea the computer program was the hard part, I figured the machine itself would be the tricky bit. The part about running it through a terminal confuses me a bit though, I thougt that was only for programmers in movies. Is that something that comes built into the computer allready, or do you have to install something special to get it working? Also curious if that batch script idea works on a Mac, since that's what he has at home.

I'm new here and mostly lurking, but this thread hits close to home. I don't have vision issues myself, I have arthritis in my hands, so mouse-heavy programs are my problem instead. I ended up remapping a lot of functions to keyboard shortcuts and it made a huge difference for me, so I can see why keyboard access would matter for your brother too.

One thing I'm curious about, since I don't know much about screen readers. Do the different ones behave differently with programs like this? My coworker swears by one reader while her son uses another, and she says they announce things differently in the same app. Might matter for whether that accessibility layer works for him out of the box or needs some fiddling.

That terminal workflow idea got me thinking. I'm nowhere near that level but I do a lot of my slicing at odd hours while my partner's asleep, so I might steal the script approach myself just to keep the noise down on the mouse clicks lol.

Also curious how the brother gets on with the actual model prep side. Slicing's one thing, but rotating a part or checking if somethig fits on the plate seems way harder to do without seeing it. Do screen reader users have any tricks for that, or is it mostly a case of keeping models in a known orientation before they ever hit the slicer?