brand-new g-code post-processor?

[joergen 2]

Here is a feature request, and idea what could help with generating better g-code.

we have 3-5 working slicers on the market right now, each with its own set of features, and lack of features that would make it perfect and highly usable. if we could combine certain aspects and styles of each slicer, we could get perfect g-code, without the need of writing yet another slicer.

The tool I am proposing is a g-code post-processor, which isn't just copy/pasting c-gode in a text editor, or writing a script that does change things (since I am not capable of writing such thing), but a visual g-code representation that allows you to change things based z-height.

Here is what it should do (and feel free to add to it) (all assuming that only relative E g-code is allowed, or at least reset each layer):

1. load existing gcode and show the printed path (print-run/kliment, cura, KS, and NF all have that)

2. show the actual layer number, z-height in mm, print time/speed, temperature, cooling, flow-rate

3. allow to change those parameters, for individual layers, and a range of layers

4. advanced feature: allow changing those parameters for parts of a layer

5. advanced feature: allow changing those parameters for types of extrusion (i.e. all support material)

6. advanced feature: allow combining of 2 or more gcode files

7. allow multi-extruders (#6 could be helpful for this???)

8. add/remove/modify retraction

9. save the new g-code, but also save the modifiers as a profile (and allow to apply the profile to other g-code files)

I think a universal, GUI-based g-code post-processor would be more helpful than yet another slicer.

Since I can't code, is anybody interested tackling this project (since all the components already exist, it should be doable)?

[alaris2 1]

I would agree that a post-processor is a good idea, tho it obviously avoids 2 rather important issues

1) a lot of people are having problems with getting to any g-code at all - the vast majority of example files I tested for example all had at least one fault - inside-out or non-manifold geometry, flipped normals, self-intersection, infinitely thin planes, etc etc.

2) the ideas are excellent, but creating UI to make it intuitive and easy is.. well, not so easy. there's a matter of designing that which I'd suggest is what this thread tackles first before any code gets written