Re: bug with 'test'

[The Wanderer]

Eric Blake wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> You broke threading again.  Usually on mailing lists, you want to hit
> reply-to-all instead of reply.

Actually, no; on every single mailing list I've ever been on (unless you
count those old "string a thousand E-mail addresses together in the Cc:
field" AOL messes) except for this one and the Enlightenment user list,
the Reply-To field has been set and so a simple Ctrl-R has been
sufficient. Regardless of how one may feel about the correctness of a
mailing list setting Reply-To, I don't think that "usually" applies.

> According to The Wanderer on 12/8/2005 11:55 PM:

> > As it happened, I had also tried '[ ] ] --help', and got the same
> > message - but presumably this can be explained in the same way. It
> > somehow seems a little weird that such fundamental programs don't
> > permit options to come in 'any place which isn't nonsensical', but
> > I'll admit that I don't see any way they could do that and still be
> > as useful in the ways in which they are...
>
> POSIX mandates some non-typical behavior on several of the coreutils;
> for example "echo --" must print "--" instead of ignoring it, or
> "printf %s - --help" must print "--help" instead of treating it as an
> out-of-order option.  But where POSIX does not forbid it, the bulk of
> the coreutils have GNU behavior of accepting any number of options,
> in any location prior to the first --.

I somehow suspected as much. Virtually every behaviour which has seemed
odd to me about one of these things has turned out to be a POSIX
requirement. (And, as I indirectly noted, sometimes it makes sense once
I spend enough time thinking about it.)

Hmm. Speaking of echo and its requirements, is there any way to have it
print '-n', '-e' or '-E' in an instance where they are not preceded on
the line by anything which is not a recognized option?

And, as long as I'm asking off-the-thread-topic questions anyway,
something various people (including me) have ben wondering about which I
know I'd never get around to starting a thread to ask: is there any
reason why 'mv' does not have a --preserve option, comparable to that of
'cp'? I would have found it useful on quite a number of occasions, and
have been forced to resort to copying manually and deleting by hand
afterwards. Admittedly it can be automated somewhat by find and/or by
scripting, but the absence still seems puzzling.

--
      The Wanderer

Warning: Simply because I argue an issue does not mean I agree with any
side of it.

Secrecy is the beginning of tyranny.