Date: Fri, 30 Dec 2005 00:39:06 -0500 From: Barney Wolff <barney@databus.com> To: Martin Cracauer <cracauer@cons.org> Cc: Barney Wolff <barney@databus.com>, freebsd-current@freebsd.org, Sean Bryant <sean@cyberwang.net> Subject: Re: fetch extension - use local filename from content-disposition header Message-ID: <20051230053906.GA75942@pit.databus.com> In-Reply-To: <20051229220403.A16743@cons.org> References: <20051229193328.A13367@cons.org> <20051230021602.GA9026@pit.databus.com> <43B498DF.4050204@cyberwang.net> <43B49B22.7040307@gmail.com> <20051229220403.A16743@cons.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Dec 29, 2005 at 10:04:03PM -0500, Martin Cracauer wrote: > > The security implications are about the same as for the base > functionality. Any filename in the current directory can be wiped out > if you fetch or wget and a URL redirects to another URL which leads to > a filename that matches. If fetch uses a redirected name as its local filename it is seriously broken and must be fixed. The manpage does not mention it. > The default behavior already *is* that the sending server has control > over your local naming. What does the security officer have to say about that, if true? > I will forbit "/" to appear in the suggested filename, though. Remember that the check must be made after any decoding of %xx et al. But no check will save the gullible from creating .shosts in $HOME or overwriting .profile . That's why I believe the whole thing is a bad idea. -- Barney Wolff http://www.databus.com/bwresume.pdf I never met a computer I didn't like.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051230053906.GA75942>