Hi Tom,

Thank you for the comments.

Tom Gillespie writes:
 > All of the issues that I'm aware of are related to what
 > happens if tangling fails part way through the process.

That's not something I had considered. I wrote a new version of the
patch (attached) which addresses the insecure behaviour and the
possibility of failure. Please tell me what you think.

I've also
  + silenced the ~write-region~ messages, since I'm now writing to
    temporary files.
  + added the list of tangled files to the message to the user at the
    end of the tangling process.
  + replaced the use of ~when-let~.

Regards,

--
Sébastien Miquel