Announcement

Collapse
No announcement yet.

Comparison algorithm bug

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Comparison algorithm bug

    Good evening,
    I'm testing beyond compare on my NAS and I think I have found a bug.
    Comparing two network folders (let's call these HDD1 and HDD2) with a very large number of files,
    Let's suppose the following situation:

    HDD1: || HDD2:
    ------- || --------
    FolderA || FolderX
    fileA || fileA'
    ... || FolderZ
    .. || fileA
    .. || ..

    Where fileA of HDD1 has the same name of fileA' and fileA of HDD2,
    however file fileA and fileA' are different.

    When comparing with the flatten option with the "same" filter active, I expected listed fileA of FolderA and fileA of FolderZ.
    However it can happen that the sw compares fileA' of FolderX before fileA of FolderZ.
    In such situation file A of FolderA is not listed as equal to fileA of FolderZ
    Left orphan is corrected since fileA of FolderA is not listed.

    I hope I was clear.
    Thank you in advance,
    Vincenzo

  • #2
    Hello,

    Thanks for the feedback. This is a current limitation of the Ignore Folder Structure view. In all 4 view modes (including flatten folders), alignment occurs before file comparison. This means a file pair is picked and aligned first on file name, and then the comparison runs on all aligned pairs. In 3 of the 4 view modes, folders are important and will always align the expected pair. However, in the flattened folders view, files with the same name in different folders can align before the comparison is run. You can manually re-align with the right-click Align With command to push the alignment around in the compare window, but BC4 cannot currently auto-detect and prioritize duplicate files in different folders in the Ignore Folder Structure view.
    Aaron P Scooter Software

    Comment


    • #3
      Thanks for the answer Aaron,
      manually re-aligning is not a solution when dealing with thousand files and
      I am sure this limitation can cause serious mistakes in such situations.
      I would suggest also to add alignment by hash file, such option would allow a correct duplicate finder.
      Let me know what do you think about this suggestions.

      Thanks

      Comment


      • #4
        It's a good idea. An alignment based on a comparison scan or duplicate scanning is something on our wishlist, but not a feature currently supported by BC4. It's a larger project, so it is not one we'll be able to tackle short term.
        Aaron P Scooter Software

        Comment

        Working...
        X