Skip to content

Latest commit

 

History

History
62 lines (39 loc) · 3.76 KB

README.md

File metadata and controls

62 lines (39 loc) · 3.76 KB
Raw
      ╻ ╻      ┏━╸   ┏━╸┏━┓╺┳┓┏━╸   ╻ ╻╻╺━┓┏━┓┏━┓╺┳┓      ╻ ╻
   ╺━╸╺╋╸╺━╸   ┃╺┓╺━╸┃  ┃ ┃ ┃┃┣╸    ┃╻┃┃┏━┛┣━┫┣┳┛ ┃┃   ╺━╸╺╋╸╺━╸
      ╹ ╹      ┗━┛   ┗━╸┗━┛╺┻┛┗━╸   ┗┻┛╹┗━╸╹ ╹╹┗╸╺┻┛      ╹ ╹

GWiz is an advanced gcode "man-in-the-middleware" that aims to make it comfortable to tune a machine (initial tuning and config) as well as tuning complicated prints1, including keeping track of all changes to a print in order to replay and gradually improve the prints. In the near future, it should enable synchronised collaborative work on an arbitrary number of machines.

gp

This repository also include a command-line utility gp to pipe gcode from files or stdin to a machine ; unlike most similar utilities that can be found at the time of this writing, gp doesn't use the naive approach of "I send a command and wait for an answer before sending the next command" but instead uses threading for async read/writes. Still, gp doesn't actually work yet (it is likely to just stop the print anytime) so don't use it for production.

TODO

  • when gcode files are provided as arguments, allow insertion of gcode commands with stdin, therefore combining the main features of pronsole and printcore.
  • allow reading from an open file without closing it, like tail does

GWiz

GWiz sort-of competes with pronsole but it aims to be lighter, more flexible, more user-friendly. GWiz requires very few dependencies and uses urwid as a toolkit, which allows a GUI-like feeling in a terminal or remote shell (ssh).

At the moment, resources used are still pretty high ; typically, a CB1 can be too slow to send instructions for production use (heavy stutter when there is a succession of short paths).

This program moves gcode instructions from the developper's mind (or G-Code file, pile 'zero') sequentially to a number of other piles:

  • WAIT pile: instructions that are scheduled to be sent to the machine, but the machine's buffer is full (or we artificially throttle them2)
  • WIP pile : instructions that have been sent to the machine's buffer, no ack or error message is available yet
  • ACK pile : the last pile, instructions have been processed by the machine ; they have either an 'ok' message or an error message attached. This pile also contains most messages sent by the machine and user comments.

Other notable features include:

  • searchable list of gcode commands by description (list set in machine config) ; in the future it will also display command usage (auto-fetched from firmware doc)
  • a silly MIDI-to-M300 converted that is clueless about rythm so you can play your favorite tunes on your device's speaker

more info in the proghelp.py file (or when running the program itself).

How to run GWiz

for example

./GWiz.py -c machine.conf -g part0.gcode -g part1.gcode 2>/tmp/gwiz.log

and simultaneously running

tail -f /tmp/gwiz.log

in another term for debug output should this be required.

Note

G-Code Wizard is in early development stage.

Screenshot

static/GWiz.jpg

GWiz playing the Tetris theme while the user is searching for commands whose description contains the words terms hot and tem.

Footnotes

  1. prints or any other program sent to a machine ↩

  2. reducing the number of commands in the machine buffer has a number of use cases, for example to allow insertion of other instructions or aborting/pausing prints faster: this is most useful during machine tuning. See buffsize command - or to be able to stop a print faster ↩