Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 31 Aug 1999 01:02:10 +0200
From:      Martin Cracauer <cracauer@cons.org>
To:        "Jordan K. Hubbard" <jkh@zippy.cdrom.com>
Cc:        "Rodney W. Grimes" <freebsd@gndrsh.dnsmgr.net>, Mark Murray <mark@grondar.za>, cvs-committers@FreeBSD.ORG, cvs-all@FreeBSD.ORG
Subject:   spending toolchain options (Re: cvs commit: src/bin/mkdir mkdir.1 mkdir.c)
Message-ID:  <19990831010210.A30447@cons.org>
In-Reply-To: <3072.936051369@localhost>; from Jordan K. Hubbard on Mon, Aug 30, 1999 at 03:16:09PM -0700
References:  <199908302112.OAA66970@gndrsh.dnsmgr.net> <3072.936051369@localhost>

next in thread | previous in thread | raw e-mail | index | archive | help
> Let's *discuss* this for an additional 48 hours and then try to come
> to a concensus decision on these changes.  If we decide that -v isn't
> worth all the freakin' contraversy then fine, it goes.

What we need are decisions about some past rules of thumb:

0) How much effort such be done to spare option letters for possible
   later use?

Reasons to spare them:
- Some standard may come along and use an option we've already spent.
- Misspelled commands don't do "other" things. If all UNIX derivates
  would start adding their own options, shell script with wrong
  assumptions woudn't fail with suntaxt errors, but run with wrong
  switches turned on.

1) Old rule of thumb: Don't add options to commandline tools that are
   easily substituted by standard (POSIX) bourne shell constructs.

Example 1:
  date -n
was supposed to suporess the newline. This is easily replaced by
backquotes or $(), maybe within double quotes.

Example 2:
  date -l
was supposed to echo the date in the timezone the machine thinks it is
in, no matter what $YZ is set two. Most useful for working on freefall
over the atlantik, but replaceable by (unset TZ; date)

Will this be turned into a "real" rule?

2) Do we want our users to learn the shell?

Don't dismiss the question easily. In my opinion, we loose badly if
usage of UNIX things drops too much.


Overall, I don't like the -v stuff, and adding -v to mkdir(1) just to
make sure every util has -v is something I'm strongly against.

As Mike Smith pointed out, these are not intended for shell script
usage (then sh -x would be the better way), but for interactive usage.
I don't think this is a useful addition for beginners. Add them as
aliases to .bashrc and every shellscript started with bash will bomb
out.

The code committed also had quality issues, as pointed out in my first
mail and by others.

Martin
-- 
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Martin Cracauer <cracauer@cons.org> http://www.cons.org/cracauer/
BSD User Group Hamburg, Germany     http://www.bsdhh.org/


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990831010210.A30447>