Hash key computed when writing directly to compressed packs is wrong
Closed this issue · 0 comments
giovannipizzi commented
The hasher is applied in the wrong place (while writing rather than while reading) therefore it is hashing after compression.
The hasher should run on the read data instead.
- Tests should be added for this case (one example where 100% coverage does not mean bug-free...)
- We should have a check that objects have the correct hash while packing