ViaRewind icon indicating copy to clipboard operation
ViaRewind copied to clipboard

Optimize 1.7 compression by using velocity-native compressor and reduce memory allocations inside chunk rewriting

Open Beaness opened this issue 3 months ago • 1 comments

image Should improve chunk rewriting performance

Beaness avatar Dec 17 '25 16:12 Beaness

Latest commit has the additional following changes:

  • Dropped ExtendedBlockStorage to prevent having extra memory byte array allocations when it can be written directly to the output byte array
  • Allocate only one byte array in bulk chunk map where all chunks are serialized to

Beaness avatar Dec 17 '25 18:12 Beaness