No announcement yet.

Feature suggestions: calculate better hash algorithms

  • Filter
  • Time
  • Show
Clear All
new posts

  • Feature suggestions: calculate better hash algorithms

    It would be very helpful if Beyond Compare would be able to calcuate hashes that are more practical with very large directory structures, such as MD5 and SHA-1. CRC is a good "default" hash since its very fast, but there are many cases where switching to a better hash would be more useful.

  • #2
    What are you using the hashes for? BC uses them when comparing against zips (which store CRCs in their headers), and when comparing against FTP sites that support the XCRC command. For regular disk-to-disk comparisons, binary comparisons are better.
    Zoë P Scooter Software


    • #3
      I realize that binary comparisons are better, but in very large comparisons (100+ GB), hashes are faster and sufficient if they're at least as good as MD5.

      Also, strong hashes can be used as a reference to files and deduplication when exported to a report (using Excel, etc).


      • #4
        Hashes are only faster if they are pre-calculated. In order for BC to compute a hash, it must read the entire file thus offsetting the benefits of a strong hash comparison.
        BC v4.0.7 build 19761


        • #5
          Thanks, 01qa. I've added your notes to our Customer Wishlist. It would be useful for reporting purposes, but you may want to try our Binary compare for speed. Since the binary compare can return unequal as soon as it detects a difference, it does not have to read the entire file. You may be experiencing slowness if it is over a network (at which point, we have to download the file first before performing a binary compare).
          Aaron P Scooter Software