Re: rm patch suggestion

[Oystein Viggen]

* [Joshua Judson Rosen] 

> On Tue, May 07, 2002 at 04:57:56PM +0200, Wolfgang J?hrling wrote:
>
>> At least with the -f option, this behaviour should not occur, because
>> one might actually _want_ to remove everything. Maybe there should also
>> be an additional option (say, -t) for this, so that both
>> 
>> rm -r foo/
>> rm -rft foo/
>
> How about an -x/--one-file-system option like cp has?

Problem is, you never ever want to recurse into directory translators
belonging to other people, because they are likely to be firmlinks to /,
infinite depth fake directory trees, or something worse I haven't
thought of yet.

People will expect -Rf to as safe from subversion on the Hurd as it is
in any other Unix like OS.  Also, -f is mostly just the opposite of -i,
so I see no reason why -f should make rm run in an unsafe way.

Adding a -t/--translator switch for doing something interesting to the
translators could be a possibility, though.

Another possibility would be to make rm remove translators by default,
also removing the need for any -t switch.  The reason I did not do this,
was that this would be quite an invasive change to the rm code, and I
was also unable to find any documentation on file_set_translator, so I
decided to use the easiest solution first.

It would be interesting to hear more opinions on how rm should behave.

Oystein
-- 
When in doubt: Think again.