Useful for identifying duplicates, needs work on scaling
11/23/15 - Its been several months and a few updates to this app since my original review - now instead of the below issue where dupes where incorrectly identified I have terabytes worth of data that it is not seeing as duplicates. MD5 checksums verify they are exact matches and the names are exact matches.
What I liked:
- Fairly straightforward interface.
- The "Select Duplicates In" option was great for ensuring the copy I wanted to keep wasnt deleted. This feature is exactly what I needed.
- The graphing and filtering options helped me identify key problem areas.
- Found and helped remove 1.5TB of duplicates across multiple drives.
What I didnt like:
- There was at least one instance where the MD5 checksums of three files flagged as being duplicates of each other did not match. They did have the same byte size, but different names. Ive edited this to change my rating as it does auto-select them as dupes if they are in different folders (originally I thought it was ignoring them through auto selection tools - but they were in the same folder then). This is pretty serious and may have resulted in lost data for me, especially since I was working with such a large volume of files my oppurtunity for multiple same size files was high.
- The marketing suggests it will take literally seconds even with enormous storages - it took several hours to scan my drives (10TB). All though I expected that, I could see others being disappointed as a result of the claims made on the website.
- When your "results" set is large (several hundred thousands files) Gemini becomes very unresponsive. My initial scan consumed 7GB of RAM and 25GB of virtual memory. I ended up having to force quit because it wasnt able to cope well with that much at once. I was able to use the information from that scan to clear out a lot of duplicates manually and then run a few smaller more focused scans. Re-ran the initial scan and the results where down to under a half TB and less then 100k files - consuming only 2.5GB of both virtual and real memory and the program was very snappy in comparison to some of the earlier scans where I would have to click a button and go do something for 20-60 minutes before it would respond.
- Did not utilize more then 1 core/cpu.
- The shredding animation. In addition to being a bit tacky its confusing imagery when the program is just moving files to the trash and not deleting/shredding anything.
- Being asked to rate this app by the program on second launch and having Clean My Mac (Which I have no interest in) being advertised to me after each run.
NamrogForwards about Gemini: The Duplicate Finder
