Repository navigation
Precompute cumulative weights for weighted block-model selection - #577
Open
BonsUnleashed wants to merge 1 commit into
Open
BonsUnleashed wants to merge 1 commit into
BonsUnleashed wants to merge 1 commit into
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Proposed Changes
Picking a weighted block model walks its entries and reads each weight again for every face and render-layer query. This computes the running totals once when the model is built, then searches that small array. The random draw and selected variant stay the same.
Checks on the compiled mixin, applied to Minecraft's model class, covered about 1.3 million picks: identical variants, exceptions and random draws, including zero weights and the negative
Math.abs(Integer.MIN_VALUE)case.The equivalent patch ships in Bons and Furious 1.0.26 for Embeddium 0.3.31. In a separate benchmark of this port, a mixed set of model lists took about 30 ns per pick instead of 54 ns. Whole-frame differences were too small to measure in game.
Compiles on
20.1/forge(compileJava). Not tested in game on this branch.