build
build copied to clipboard
Optimization question: does `_assetGraph.computeOutputs()` need to be fully computed?
I'm looking at _assetGraph.computeOutputs, and I see it's only used in this line:
Does computeOutputs() need to be run for all nodes?
Do you think populating only the accessed nodes would be faster?
Best regards.
It's reversing the outputs graph, so I think you have to process the whole thing in one go, then the result is cached for later lookups.
It's used if serving the results or if copying to a merged output folder ... I'm not sure if there is a use case where the performance matters too much, but I haven't actually tried to measure.