I was able to compile the 1.5.3 branch on Linux, and it ran, but it was unable to connect to my Makerbot Replicator 2x. 1.5.0 is able to connect. Some functions are a little buggy, but for the most part it works, which is great considering the Makerware Desktop (a) runs on Ubuntu 14.04 but not on the latest 16.04, and (b) runs on Windows 7 but crashes on Windows 10. I turned to MC as a much-sought-after open-source method to control the printer.
Any time table on when we can expect 1.6 on Linux?
On Arch based distros make sure you have installed mattercontrol-plugins as well. This has the closed source plugins including the S3G driver for Flashforge printers. Also make sure you have given your user permissions for the serial port. See the wiki article on running on Linux.
What exactly is wrong with the default start gcode and what are you trying to change it to? We've done extensive testing with our own Flashforge and it works fine.
Why are you trying to change the config file instead of editing the start gcode from within MatterControl itself? The settings in /usr/lib/MatterControl/StaticData/PrinterSettings/ are the default settings loaded when you add a new printer. Your own settings are stored in an SQLite database in ~/.local/share/MatterControl/ (in 1.6 these will be moved to .json files)
Slic3r and Cura engines are disabled for dual extrusion printers, since MatterControl currently does not support dual extrusion using those engines. I realize its a little confusing to have them completely hidden in the menu. If you want to try Slic3r anyways, switch your extruder count to 1.
MatterControl doesn't really work with the 5th generation Makerbots. You can export X3G files for them to print from the SD card, but you cannot connect to them directly over USB. The 5th gens use a different communications protocol than most printers, which we don't support (yet).
Nevertheless, the application should not be crashing under any circumstances. Could you run MatterControl from the terminal and post the output?
I ran MatterControl as 'Root' and still have problems. When I re-scale an object, then hit save as, it changes the icon to a question mark. If open the file folder, I see an '.AMF. file, but attempting to view gives me a blank preview screen.
Bottom line: Slic3r fails completely, MCslice works but has problems where solids cross, CuraEngine works but fails to connect some cross bars. Since I can load G code directly I will probably end up creating my own.
What makes this shape so difficult is the number of interior walls. It is essentially 60 rings held together with bars. Each ring has an interior, and all the cross bars connecting the rings create interiors. Where the bars cross each other the slicing engine has to decide what is inside and what is outside - MCslice calls the solid in between crossed bars "outside" because the surfaces cross there. So it does not print there. CuraEngine does it half the time, and the other half it simply ignores the bar altogether. It gives rows of bars and rings disconnected so the full shape is not one piece.
I've learned a lot, and I like the fact I'm pushing this technology to its limits. I know it can do what I want. It's just going to take a while. Especially since keeping the wife happy is a major factor!