[Gnu-arch-users] Re: 'arch send' format

[Miles Bader]

Amit Shah <address@hidden> writes:
> We could have a shell script with a normal patch that shows changes within
> files. For representing changes to the file structure (like renames,
> deletes, etc.), we could put shell commands like "mv foo.h bar.h" and so
> on, so that patches created by people using arch for people using other
> systems (or even arch, but don't want to "get" the changesets from some
> archive) can be applied normally.

People always suggest this when the topic comes up, and it seems like
a really bad idea to me; it's just too dangerous for people to get in
the habit of running shell scripts they receive.

I don't even like the idea of a representation that only _looks_ like
shell commands (but is applied by a special program) -- people might
still get into the habit of running it anyway, or trying to.

It seems simpler and more comforting to have a nice simple abstract
representation which is easily turned back into a changeset and applied
by `dopatch' (perhaps a single tla command does everything, but anyway).

[
 Maybe it's nicer to use a renaming syntax instead of the
 id-tag/filename pairs that changesets use, so it's easier for a human
 to read, but my argument is that it should be something like:

    "FILE1" => "FILE2"

 not

    mv "FILE1" "FILE2"
]

-Miles
-- 
Is it true that nothing can be known?  If so how do we know this?  -Woody Allen