kicad-developers team mailing list archive

[Mário Luzeiro <mrluzeiro@xxxxx>]

Hello Jakub, NickOe, Jean-Pierre, Maurice,
Thanks again for testing the 3d-viewer!

On the loading times:
- I found the cause and improved it. It was because there was much detail on segment counts, so I manage to get a compromise to (what I experienced) load it in similar times as before.
However, I didn't understood why the Nick board (I tested on my PC) was experience so high load times for such an "apparently" simple board. I believe your board may have something that is making it more polygon count in copper layers (maybe zones? .. something stranger...) 
I found in your board (Nick) that for example, the NPTHs in your big connectors, have about 6 holes on the same place. (This does not explain necessarily the long times, but maybe there are other issues like that on the board that is contributing for more computation needs)
- Please note that: At this moment, the loading process is loading ALL the copper layers. The stable kicad, in realist mode (or with board?) will only load the top and bottom. So if you make comparisons, take this in account.
This may change in a near future..

Please update with my branch if you would like to test the latest upgrades:

- I implemented some progress status, for loading and rendering. Work in progress...
- I make the panning movements faster (and remove the green pivot cursor while panning)

Issues Fixed:
- Fix issue what was positioning the board again after reload (or update parameters)
- Fix an issue in NPTH OpenGL
- Improve technical materials on opengl.


Tips:
- If you have a slow CPU but still want to test raytracing as it is now, you may want to decrease the windows size (or disable some eyecandy options).. it will render faster.

Know issues:
- On Windows I experienced that if Raytracing takes too long, it will keep refreshing the windows and the rendering.. will not stop. .. consider this a temporary issue until I find another way of implement it.
- I still experience some times the bug reported by Jakub (the "squarish segments")

Regards,
Mario Luzeiro