[ On Tuesday, April 9, 2002 at 15:48:48 (+0700), Robert Elz wrote: ]
> Subject: Re: VERY strange command-line syntax for 'df -t' 
>
>     Date:        Mon,  8 Apr 2002 13:08:33 -0400 (EDT)
>     From:        woods@weird.com (Greg A. Woods)
>     Message-ID:  <20020408170833.AE651AC@proven.weird.com>
> 
>   | What about if other command-line operands imply a default list of
>   | filesystems which is a subset of the complete list?
> 
> There are no such other command line options.

I didn't say "options", I said "operands", and I mean other parameters,
arguments, or whatever you want to call them.  I.e. those things that
are not option flags and are not operands for option flags.

>   | Under my proposal I would expect your example to generate a warning if
>   | no "mfs" filesystems are mounted (and no files are specified on the
>   | command line).
> 
> No, the warning is if you explicitly ask for output about a filesystem that
> is of a type that df isn't going to list.   Not if you ask for a type and 
> there don't happen to be any filesystems of that type.

You're talking about the current implementation.  Repeat after me:
"Under Greg's proposal...."

> The -t arg as it stands now is just fine.  The way the "no" is used is a
> little unusual, but there's no real good reason to change it

I've given several reasons why I think it should be fixed, and I think
they're very well "good enough" reasons.  What I'm asking is what kind
of fix would be most widely acceptable.

> (I wouldn't
> object if "no" were permitted on later filesystem types in the list if
> it were present in the first, but apart from that, nothing should be altered).

Thank you.

-- 
								Greg A. Woods

+1 416 218-0098;  <gwoods@acm.org>;  <g.a.woods@ieee.org>;  <woods@robohack.ca>
Planix, Inc. <woods@planix.com>; VE3TCP; Secrets of the Weird <woods@weird.com>