Hi Tom,

Thank you again for your comments.

Tom Gillespie writes:
> I think that the location of condition-case is ok, but I wonder what
> would happen if something were to fail before entering that? I think
> that only a subset of the files would be tangled, but they would all
> have their correct modes, so I think that that is ok.
On second thought, I'm uneasy about my approach. If tangling fails,
the user might miss the error message since it is quickly replaced by
the tangling info. Ideally we should backup all the tangled files and
restore them all if a single one fails to ensure we're back to a
consistent state.

I'm unsure what would be best practices here. In case of a remote
tangled files, I don't know if temporary files should be remote or
not, and what guarantees do emacs primitives such as ~rename-file~
offer.

Although a robust tangling system that deals with errors and
guarantees that the state ends up consistent would be nice to have,
I'll take the failure considerations off this patch to keep it simple.
It'll make a better starting point for future work at least.

As is currently the case, if tangling fails, an error with be thrown,
the user will certainly notice and should assume that everything is
broken until another tangling succeeds.

I've kept the modes improvements.

Regards,

-- 
Sébastien Miquel