Dear LBT team,
Using the latest LBT 1.10.2 on Mac, HB Annual Daylight Metrics works well if I don’t sync the script from 1.9.x to my latest 1.10.2 installation. But since I synced the script to 1.10.2, HB Annual Daylight Metrics “Failed to compute annual daylight metrics”.
I wonder if this could be probably a bug that you could check?
Thanks.
Farhang
Scripts in versions 1.9 and 1.10 attached.
AnnualCBDM_Test_1_10.gh (67.4 KB)
AnnualCBDM_Test_1_9.gh (73.8 KB)
Thanks for reporting, @farhang.tahmasebi .
I can recreate the issue on my end but only on Mac (not on Windows). If I had to make an inference about what is going on, the issue is likely related to the GPU Acceleration that was recently added, which is unavailable for Mac computers since they do not have GPUs that are compatible with cupy.
@mikkel, let us know if you want to look into the issue first. If not, I can try investigating sometime over the weekend.
Hi @farhang.tahmasebi ,
I did some testing on my old Mac and I realized that the component was actually working correctly under the hood and the result files were generated. It was just turning red because the component could not sense that the process finished successfully. It seems to be a bug that is specific to the IronPython that Rhino uses on Mac.
In any event, found that if I add an extra step to wait() for the process to finish, everything seems to be sensed correctly. So I changed all of the affected components here:
… and you can now get the fixed components by using the LB Versioner.
If you can confirm that this fixes your issue, I think I may update the Food4Rhino installer with this set of fixes.
FYI, @farhang.tahmasebi and @mikkel ,
I went ahead and updated the Food4Rhino installer with the fixed components.
In total, there were 8 Radiance result-parsing components that were not working on Mac in the release. So this issue seemed to merit updating the Food4Rhino installer, particularly because this is the main installation pathway for the Mac users.