From Hildo
Will be v6 3D package compressed into STPZ (step.gz / step.zip) format?
Will the same files use the Seth Hillbrand clean-up? (https://gitlab.com/sethhillbrand/stepreduce)
From Hildo
Will be v6 3D package compressed into STPZ (step.gz / step.zip) format?
Will the same files use the Seth Hillbrand clean-up? (https://gitlab.com/sethhillbrand/stepreduce)
We are actively discussing the possibility for using both STPZ and StepReduce to reduce the overall size of the 3d model libraries distributed by default in KiCad. Combined, these technologies should help to reduce the 3d model library size to manageable sub-Gigabyte size.
Stepreduce may increase the loading time and the handling complexity of step files in FreeCAD. Probably this structure wouldn’t be an issue in commercial sw, but in FreeCAD may seriously introduce a big delay wen loading a kicad pcb and handling step files.
Stpz on the opposite is already handled without issues in FC.
PS I haven’t see any discussion on the k mailing list.
The discussion has been on the various bug trackers. If you are interested in learning more about how this works and would be applied, please ask. Statements like this are both wrong and do not apply to the work as it is being discussed.
?
I remember I participated to the stepreduce discussion long time ago…
I have to find out where and when (probably on gh)…
I’m sorry… may be I recall wrongly …
I would be happy to test any models if you have a library generated with the stepreduce process and its counter part before the reducing process …
There still not a post-processing for STEP export at Pcbnew:
Pcbnew already suport STPZ by gz (Unix) compression to read models but missing some work to import ZIP compressed.
I tested FreeCad on Ubuntu using step.gz compression and stepreduce algorithm for STEP Pcbnew export model and third one (using models of 100MB). All tests returned a decreasing in time for load the model.
On Windows, is FreeCad recognizing both: step compressed by ZIP and by GZ?
On FreeCAD under Linux, Win and OSX will parse compressed format in the same way using python code (that I introduced in FC almost 3 years ago FC STEP compression format .stpZ support)
About the stepreduce approach, it seems that the processing doesn’t affect the final step structure. So I probably just recalled a different discussion on step library reworking.
@Seth_h sorry for the noise
@chschlue thanks for the library’s link to let me test with the models.