Re: coreutils bug with "ln x d/"

[Paul Eggert]

address@hidden (Paul Jarc) writes:

> when at least one trailing character of the link name is given
> literally rather than through a variable, there will be no
> surprises.

Yes, if you have control over the suffix, then you won't have any
problem.  But I was worried about the more general case.

>>   case $y in
>>   */) echo >&2 "target ends in /"; exit 1;;
>
> Stripping off the "/" before calling ln would work too.

Not if the target is "/", or "//", or "foo//".

> (The DOS version of --syntactic would have to check for
> either "/" or "\" at the end, yes?)

No, it's more complicated than that.  For example, "C:" counts as a
syntactic directory in DOS.  (And I have no idea what the syntactic
rules for OpenVMS would be.  :-)

I guess my point is that the --syntactic rules are relatively
complicated, even for POSIX (as per the "/" and "foo//" examples), and
certainly for non-POSIX systems; whereas the --no-target-directory
rule is relatively simple.

> --syntactic won't be surprising, and can offer some convenience -
> you can omit the basename from the target if it's the same as the
> source's.

If you want to do that, you can use --target-directory already.