Age | Commit message (Collapse) | Author |
|
I fixed the non-manifold geometry in the sensor mount which causes openscad to fail to render TBIW configurations/
|
|
Added separable thumbs and cleaned up all the connections and thumb connections.
Thumb connections previously contained what should be in the finger connections.
Added extra screws for separable thumb config to secure it. Previously had one screw.
|
|
Also, working on modular trackball, though ball bearings make it difficult to decouple as they typically land in the middle of the walls and webs.
|
|
|
|
|
|
|
|
Added note for when run_config is used vs. configuration argument.
|
|
run the run_config.json file if there are no CLI arguments.
Retained ability to run the run_config.json file if there are no CLI arguments.
Will eventually deprecate run_config in favor of the configs directory, but don't want to mess up existing configs until they have time to migrate.
Un-ignored configs. Would like to build a set of user configs through PRs to see variety. Add gallery directory for those with submit configs for people to see examples. All PRs for configs are expected to have identically named files (with some suffix for multiple) to show example images.
|
|
|
|
|
|
intended.
|
|
Add trackball thumb cluster. Interim commit, needs a few mods for naming and handedness.
|
|
|
|
simplifies construction by removing the setting interdependency.
|
|
moved parts to their own directory
|
|
Minor fix to trackball in wall rotations.
|
|
also fixed bug that would not correctly inlcude parts for configs with a `save_dir` set to a subdirectory of the things directory
|
|
generate_configuration.py now only generates a configuration file; it no longer also runs dactyl_manuform.py.
it will save the generated configuration file to the configs directory, and it will be named according to the config_name. passing a --config=<name> argument to the script will set the config_name and save_dir.
dactyl_manuform.py now also accepts a --config=<name> argument. if non is passed, the default values from generate_configuration.py are used. otherwise, the defaults will be overridden by the values from the config specified.
because of these changes, i removed run_config.json. it should have default values anyway, so it seemed reduandant and confusing.
the run command has also been updated to allow for the config changes, and some bugs have also been fixed.
|
|
also removed check for a command argument; just let an empty command fall through to the default case
|
|
updated bash script name
|
|
|
|
|
|
|
|
|
|
ignore macOS system files
|
|
|
|
|
|
|
|
long.
|
|
|
|
|
|
Discovered why people had an issue with the tray. The cadquery had interference and solid did not. After fixing the unintentional difference, I found both had intereference. Added an undercut to locally thin the wall for tray engagement.
|
|
on component cut/union order.
Fixed base plates and walls to be identical in caquery and solid. Fixed if/elif/else structure that misplaced the MINI cluster screw mount.
Added intial pass at add on PCB as option to add with caps for interference checks.
Moved bottom_hull to helper libraries as they are specific to engine.
Fixed bottom_hull functions to correct minor discrepency between the engines.
Fixed solid intersect function.
|
|
|
|
|
|
mostly working (still WIP), trackball cluster has layout but no webs or walls.
|
|
|
|
|
|
Changing to ** to get build subdirectories.
|
|
|
|
Bad line in Makefile copied from batch file.
|
|
people to look at. Also fixed minor issues.
|
|
and solid python. One centered the cylinder, the other generated it in the positive Z direction. Both now fully center the cylinder and the two models generate consistently.
|
|
|
|
|
|
for the adjustable larger key shape, but no need at the moment.
|
|
|
|
|
|
|
|
|