IETF-SSH archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: OpenSSH/scp ->> F-Secure SSH server Problems



On Tue, Mar 13, 2001 at 10:56:57AM +1100, Damien Miller wrote:
> On Mon, 12 Mar 2001, Thor Simon wrote:
> 
> > > The ssh implementation and the sftp implementation are quite
> > > independent. They have a common origin and uses a common language and
> >
> > Yeah, that's my point.  I think it's entirely reasonable to *not*
> > implement sftp in an ssh implementation, given its large size and
> > general crustiness.
> 
> I don't think that you have demonstrated that sftp is either large or
> 'crusty'. Could you give examples of the latter?

I think I'd have to define it a bit better first. :-)  Later tonight
maybe.

> The sftp-server that we ship with OpenSSH is 1028 lines, not counting
> the infrastructure shared with the rest of the implementation. BTW I
> don't think that SSH1 is a valid benchmark for comparison.

Well, Niels suggested 5000-10000 lines.  That's "big" by my definition,
and I do think that it was reasonable to point out that at that size,
it'd be larger than an entire v1 implementation; that implementation
was already too damned big for an embedded system, as far as I was
concerned.

Anyone got a line count for the sftp client and server in the ssh.com
code?  I'd look myself, but for various reasons I don't read through
that codebase, just the specifications.

You can't get much simpler than "rcp", and it works.  It's a shame
that the traditional BSD "rcmd" semantics don't patch stderr,
which forces you to ship a whole separate "scp" command instead
of just using your system rcp (the wire protocol is the same).

Thor



Home | Main Index | Thread Index | Old Index