Re: RFC: cp -if a b: should it prompt? POSIX says `yes', most don't

[Adrian]

Hi. I found the following thread on the "cp -if" problem:

  http://lists.gnu.org/archive/html/bug-coreutils/2004-02/msg00077.html

which also refers to the discussion at:

  https://www.opengroup.org/sophocles/show_archive.tpl?
source=L&listname=austin-group-l&first=1&pagesize=80&searchstring=cp+-
if&zone=G

Recently I upgraded to FedoraCore3, and I saw that my "cp -f" commands,
when launched as root, now ask me for confirmation (they didn't in the
past):

  # cp -f a b
  cp: ¿overwrite «b»? (y/n)

This happens because old redhat patch is not used any more, as Tim Waugh
explains:

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=150558
> This is POSIX territory.  We *used* to patch cp so that -f also had
> the effect that --reply=yes now does, because at the time the latter
> option did not exist.

Jim Meyering also commented that in the first mentioned thread:

http://lists.gnu.org/archive/html/bug-coreutils/2004-02/msg00077.html
> Note that RedHat has been using a patch that makes cp's behavior
> non-POSIX, yet consistent with that of most other systems, for years.


I've seen that the behaviour only occurs with root because it has an
alias:

  # alias cp
  alias cp='cp -i'

So the "cp -f" if equivalent to "cp -i -f". This is what Paul Eggert 
comments:

http://lists.gnu.org/archive/html/bug-coreutils/2004-02/msg00092.html
> The main motivation here, I think, is that if you have an alias like
> this:
> 
> alias cp cp -i
> 
> then "cp -f" should cancel the -i.  This alias is (or was) common for
> root on Red Hat systems.

I think I'll define a new default alias for root, for example:

  # alias cpf='cp -f'

in order to have a short alternative to the long "cp --reply=yes", that 
I commented here:

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=150558#c4
> Hmmm, the new option looks a bit long:
> 
>   cp --reply=yes ...
> 
> Let me make a suggestion. What about patching cp so that -y assumes
> "yes" in all questions?
> 
>   cp -y ...
> 
> The -y option is used in many commands, and I suppose it would be
> useful for cp too.

What's your opinion about adding a -y option to assume "yes" in all
questions?


Thanks