It would be technically cool and potentially very useful to write a little bridge program allowing a running Muq server to appear on Linux as a virtual file system: This would allow the huge suite of existing Linux tools from 'find' to emacs to grep to what have you to be easily applied to the contents of Muq dbs.
For example, this would provide another way of using standard emacs to browse and edit the Muq db.
The existing tools would only understand text presentations of the various Muq datatypes, of course, but this could still be quite useful.
Careful attention to security issues would be important!
Alternate ways of broadening the Muq/Unix interface while maintaining
decent sandbox security include a
muq/pub directory containing
files visible to both Muq and Unix code and a hack allowing MUQ
code to be executed from the Unix commandline and within Unix scripts
and pipelines -- this would just require a named pipe, a Muq daemon, and
a small C wrapper. Or one could use Unix signals and a filesystem
convention in place of the named pipe.
Go to the first, previous, next, last section, table of contents.