Date: Tue, 26 Feb 2008 14:48:59 +0200 From: Giorgos Keramidas <keramida@ceid.upatras.gr> To: Peter Jeremy <peterjeremy@optushome.com.au> Cc: cvs-src@freebsd.org, src-committers@freebsd.org, cvs-all@freebsd.org Subject: Re: cvs commit: src/usr.bin/du du.1 du.c Message-ID: <20080226124859.GA3618@kobe.laptop> In-Reply-To: <20080226094931.GT83599@server.vk2pj.dyndns.org> References: <200802251906.m1PJ6hwB005451@repoman.freebsd.org> <20080226094931.GT83599@server.vk2pj.dyndns.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2008-02-26 20:49, Peter Jeremy <peterjeremy@optushome.com.au> wrote: >On Mon, Feb 25, 2008 at 07:06:43PM +0000, Giorgos Keramidas wrote: >> Implement from scratch a -l option for du(1), to match the same option >> of the GNU utility. The default behavior of our original `du' is to >> count hardlinked files only once for each invocation of the utility. >> With the new -l option they count towards the final size every time >> they are found. > > The third possible option is to amortize the size of hardlinked files > across all their occurrences by adding size/nlinks each time they are > found. IMHO, this results in the least "surprises" in the result. Each command-line argument is traversed separately, so I'm a bit confused about what the amortization gains us.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20080226124859.GA3618>