You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While reviewing #4104, I noticed a lot of wasted time spent in incorporate_indirect_nodes.
Here's 3s spent at runtime in the performance testing project:
This method is frequently unnecessary, especially with the change to the default indirect selection mode in that PR. This method only takes effect when there are nodes selected "indirectly" but not "directly," and we want to incorporate them on the basis of their parents' direct selection.
This one is completely on me: I added this way back in #3235, a change we're mostly turning off by default in v1. I think it should only take a few lines to skip this method when it isn't needed.
The text was updated successfully, but these errors were encountered:
While reviewing #4104, I noticed a lot of wasted time spent in
incorporate_indirect_nodes
.Here's 3s spent at runtime in the performance testing project:
This method is frequently unnecessary, especially with the change to the default indirect selection mode in that PR. This method only takes effect when there are nodes selected "indirectly" but not "directly," and we want to incorporate them on the basis of their parents' direct selection.
This one is completely on me: I added this way back in #3235, a change we're mostly turning off by default in v1. I think it should only take a few lines to skip this method when it isn't needed.
The text was updated successfully, but these errors were encountered: