Re: new ssh exploit?

[Bennett Todd <bet () rahul net>]

2003-09-16T21:16:34 Blue Boar:
> Out of curiosity, what leads you to believe that lshd will be
> better in terms of future bugs vs. OpenSSH?

Good question.

lsh seems to be using coding techniques designed to help prevent
the sort of silent, quiet bugs that don't show up in normal use but
allow attacks. I say seems because I've not yet audited the code.
But it works with length-counted strings, memory management is very
carefully thought out and clearly documented, and the code is very
tightly structured.

> You specifically mentioned OpenSSL libs and SSHv1 support as
> concerns with OpenSSH.  And sure, it seems unlikely that they just
> got the very last bug.

Actually, it's not so much I'm opposed to sshv1 support, and more
like willing to live without it. OpenSSL hasn't had the best
security track record to date, although as best I recall none of
OpenSSH's bugs of the sort that produce remote root attacks were
attributable to OpenSSL bugs.

> You also talk about a number of libraries needed by lshd, and some
> other things that aren't quite fully implemented in it yet.

It's a young implementation yet, that's for sure, and it does lean
on some libs that aren't ubiquitous, notably liboop.

> Is it just a matter of having some diversity?

That's a big motivation, for sure.

-Bennett

Attachment: _bin
Description: