Self Cleaning Files

I admit that sometimes I get lazy when it comes to cleaning out old files, databases, etc. I sometimes fear that I will remove something important. It is hard to tell ahead of time what will be important in the future.

One technical approach is to have old stuff slowly fade in ranking. It does not disappear (unless maybe disk gets full), but whatever ranking or ordering scheme we use would by default de-value material over time.

An option would be to have a tag that marks stuff as "important" that can be applied later on, and that would stop the devaluation. However, you still have the problem of when you get more important stuff than you can handle (2+screenfuls, etc)

Usage rate and importance are independent concepts. You should measure both, separately.

I bought one of those 200 gig USB drives. I can see at least two interesting ways of using it for self-cleaning.

One would be to echo the file structure of my computer on the save drive, and to copy everything over to it ala backup. When a thing is older than X, remove it from the computer, leave it on the save drive. This would work interestingly but I'd probably want to search the save drive, not the real drive, for sufficiently old stuff.

The second way would be to have the two folder hierarchies aligned by the OS so that on my computer, when connected to the save drive, all the files would appear to be there. The search boxes might then ask "search archived files" to save time.