Some questions to LW4:
Why are some objects not visible in LW4? Path is closed, I touched all anchor points!
Is there a way to make the “Laser fill path” equivalent than the Laser raster. Working with path is more comfortable than raster images. The Gcode is not generated from top to bottom or in my case diagonal from top left to bottom right. This may be the problem on the mobile cover engraving.
And at least: When will LW4/OpenCNC released? 
Workaround for polytri bug is now in place.
@Todd_Fleming and do you have an idea how to generate fill jobs from vector files in the same way than raster files?
It’s two separate code bases. lw.raster-to-gcode (an NPM package) handles raster files (another name for bitmap files) and cam-gcode*.js handle vector files. They use wildly different approaches.
I tried Mill Pocked (just for seeing what happens). I think it should work on an entire object and then move to the next.
At the moment I looks like a random order!
If you want it to do one object at a time, then create a separate operation for each object. cam-gcode combines paths into a single compound path then operates on the whole. It uses a greedy optimization algorithm to order the result.
Ah ok. I understand. So I have to do this for engraving. For laser engraving I think result will be better if an area is burned by one step.
It seems like you care about things most people don’t. You want bitmap-style algorithms on paths. You want a specific ordering on pocketing. Most people just want the final result to look good and the gcode to run their machines at a decent speed.
For sure. If the material is just burned (maybe it is hotter then areas where the laser is far away) it reacts in a different way. Not easy to explain what I mean. E.g. Burn a sheet of paper with a lighter: If It has a burned edge it will burn faster than an untreated edge. (Sorry: English is not my mother language).
Try it on your machine. It we get reports back from multiple users that a particular operation, properly configured and used, produces bad results, then we’ll reconsider how it’s implemented. BTW, mill operations aren’t for lasers, they’re for mills.
Here’s how Fill Paths performs on my K40: https://plus.google.com/101442607030198502072/posts/GdEGNfqjZH6
This is Co2. A way faster! I have a diode laser (2W). Mobilecover was engraved with a feed rate about 2000mm/min.
There’s a couple things that might be hurting the burn quality. We can narrow it down by running a test. Convert your SVG to a bitmap and run it at the same feed and power. Set Laser Diameter to the same value you used for Line Distance. Post a picture of the Fill Paths and bitmap version side-by-side in a single photograph so lighting and focus won’t be different.
I will try it tomorrow.
I did the test. And the result says a lot. Raster diagonal looks best. Fill path has the darker parts like yesterday (see first “S” lower left).
https://plus.google.com/photos/111079322062040624414/albums/6386669194695030833/6386669194820937218
This is a closeup of your Raster Diagonal run. There’s a difference in darkness between these two corners. Can you increase your motor acceleration without losing steps? I suspect your acceleration is too low.
https://plus.google.com/photos/101442607030198502072/albums/6386728427986644497/6386728428235100818
