Slice of Pi
Back when I released SlicerHub earlier this year open source slicers for 3D printing required a reasonably beefy machine in order to slice in a timely manner. At the time I thought the best approach to slicing with a low powered device was to offload that to a more powerful server. However, right after I released my source code Cura came out with a wholly re-engineered slicing engine that was blazing fast. So I switched my efforts at integrating my off loaded solution with Octoprint (the popular host controller software for your 3D printer) and spent the last while adding slicing to Octoprint but having it powered by Cura instead.
This development is an interesting wrinkle in Octoprint’s story because Octoprint is actually a fork of the web interface that was started in the Cura main branch that has since been discontinued. Anyways, sitting in all its glory is my cura-integration branch against the development branch of Octoprint. Hopefully in a few days it will get merged. However, if you want to be on the cutting edge here is how. This is not for the faint of heart as I had to patch Cura to enable this. I’m not going to go into deep specifics on how to do each step so if you have questions contact me on G+ or through the Octoprint google group here https://groups.google.com/forum/#!forum/octoprint
- Install Cura on your Raspberry Pi
- git clone my branch to where you want to run Octoprint from.
- Copy this patch to Cura to where you’ve installed Cura (you’ll have to edit cura.py using root)
* this step could be avoided if you just want to build Cura’s development branch as Daid pulled my patch in. Also, this could mess up your Cura installation.
- Start Octoprint as normal
- Go to the settings tab
- Under features there is now a Cura tab
- You should see two settings that need to change
a. the location of Cura
b. the Cura configuration file you want to use when printing
- Change those settings to match
a. Cura should be installed here if you’re using a deb on raspbian /usr/share/cura
- Fire up Octoprint and upload an STL instead of gcode and wait. It is still fairly slow compared to a faster machine (700MHz stock) but it will eventually complete
Here is a video comparing slicing on a normal machine versus a Raspberry Pi through Octoprint