UNIX keeps three times for each file: last modification, last inode change,
and last access.
Here are some things you can do with the last-access time:
Find files that have been forgotten.
This information comes from commands like
find -atime +180
(If you use the MH email system, you can find mail messages that haven't
been read or scanned in a long time.)
You can save disk space by cleaning up unused files; see article
files to save disk space.
Some users run a shell script named compresser
, which looks for
nonexecutable files that haven't been accessed in 90 days.
The program runs gzip
on these files:
-type f ! -name '*.gz' ! -perm -100 -atime +90 -print | \
xargs gzip -v
A system like this could automatically archive files to tape and delete them.
It could have a personal "skip" list of files and directories to skip.
And so on...
Check a directory to see which files are being read by programs,
This "sanity check" can help you debug programs by confirming which files are
Some UNIX systems, including versions of BSD and SunOS, do not update
the access time of executable files (programs) when they're executed.
To test yours, use ls -lu
on a pure-executable file (not a shell
script) before and after you run it.