Hello everyone,

while using orgmode to organize data that is strongly bound to files
and directories, I came to the point where in addition to linking from
the orgfile to the target file/directory I longed to have it the other
way round, i.e. getting information about the files/directories
from the orgfile while browsing the filesystem. That led to the
following idea:

Orgmodes folding and linking capabilities are easily combined to
make rudimentary but comfortable file system browsing:

* [[file:test/dirA][dirA]]
** [[file:test/dirA/subdirA][subdirA]]
*** [[file:test/dirA/subdirA/some_file][some_file]]
** [[file:test/dirA/another_file][another_file]]
* [[file:test/dirB][dirB]]

Attached is a bash-script that generates this representation from a
given subtree of the filesystem as a proof of concept.

For each file/directory displayed search for any link to it in a
given orgfile and
   - provide links back to these occurrences
  and/or
   - display the tags and associated content (read-only).

The obvious weakness inherited from linking itself is synchronization on directory
structure changes. A possible solution would be to store ids as
hidden files for each directory or even for each file that is referred to.

Does anybody else see the benefits of such a mode?

Andreas