SUU Design - Videonizer ignores duplicates which are not a duplicate

Videonizer ignores duplicates which are not a duplicate

Date: Wednesday, 2012-01-11, 6:04 PM | Message # 1

Group: Users

Messages: 4

Status: Offline

Hi!

I noticed that Videonizer will ignore some files because it thinks they are duplicates, but in fact they are not. This is difficult to trace because there is no hint which files are actually not added to the database and which files they conflict with. I tracked down one example of two small video files where I can add one but not the other. The files are *not* identical and even have different sizes.

There should be an option when adding files, so that you can see which files are ignored because of which already existing (supposed) duplicates. Also, it should be possible to add the files nevertheless.

Further, the strategy to detect duplicates should be optimized. At least the file size should be taken into account. Better: whenever a possible duplicate is detected (i.e. the hash of the first few kilobytes is matching), a byte-by-byte comparison could be triggerd for the whole file. Maybe it is sufficient to compare only three chunks (for performance reasons), e.g. some kilobytes from the beginning, in the middle and at the end of the files.

If you need an example, I can send it to you (2.5MB).