Date: Tue, 2 Mar 2004 21:40:53 -0800 From: "Jordan K. Hubbard" <jkh@queasyweasel.com> To: wollman@freebsd.org Cc: freebsd-standards@freebsd.org Subject: What's up with /usr/src/usr.bin/alias? Message-ID: <561FC4D0-6CD5-11D8-9000-000393BB9222@queasyweasel.com>
next in thread | raw e-mail | index | archive | help
--Apple-Mail-16--946477790 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=US-ASCII; format=flowed As some people already know, the SUSv2 standards require that a number of commands which are typically implemented as shell built-ins also exist in /usr/bin (for reasons which are at best clear to The Open Group). Unfortunately, the Open Group UNIX conformance test suite also tests that these things actually *work*, which can't be said for FreeBSD's current set, as we can see by the following interaction with /bin/sh: # cat ^Z [1] + Suspended cat # /usr/bin/fg fg: No current job # /usr/bin/jobs # jobs [1] + Suspended cat # fg cat ^D # Using the fg and jobs builtins work, using the "command equivalents" do not. The same is true for things like alias. I suspect this is simply due to the implementation, which invokes the commands in a subshell and thus won't actually work in the context of a test harness which is actually verifying that they did what they said they were going to do afterwards. Naturally, implementing something like "cd" as a command vs a builtin is also pretty darn hard so I'm not saying I don't understand why somebody might have punted on these, I'm simply wondering what the justification for trying to do this at all was given the shortcomings of the chosen approach. -- Jordan K. Hubbard Engineering Manager, BSD technology group Apple Computer --Apple-Mail-16--946477790--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?561FC4D0-6CD5-11D8-9000-000393BB9222>