Autore: Ian Goldberg Data: To: Thibaut Varène, 725779 CC: intrigeri, tails-dev, security Oggetto: Re: [Tails-dev] Bug#725779: libotr: OTR clients supporting both
OTRv1 and v2 are subject to protocol downgrade attacks
On Wed, Oct 23, 2013 at 12:35:09AM +0200, Thibaut Varène wrote: > > intrigeri@??? wrote (08 Oct 2013 09:27:56 GMT) :
> >> as you are surely aware of, it's been known [1] since 2006 that
> >> clients supporting both OTRv1 and v2 (such as libotr 3.x) are subject
> >> to protocol downgrade attacks clients. It's also been known for
> >> a while that OTRv1 has serious security issues (that were the main
> >> reason for a v2, actually). In short, support v2 only is the only safe
> >> way to go these days.
> >
> >> [1] http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.165.7945 > >
> >> It took a while to obsolete older v1-only software, and another while
> >> to complete the libotr 4.x transition and get to a sane state in
> >> Debian testing. Now, I think the time has come when we can reasonably
> >> expect v2-only to work for everyone.
> >
> >> I think that the only reasonable course of action from now on is to
> >> patch libotr in stable and oldstable to only support OTR v1.
> >
> > (s/v1/v2/ in the last sentence, obviously.)
> >
> > Ping? If you have no time to take care of that, fair enough, but then
> > I would really appreciate to read your general opinion on the matter,
> > even if it's a simple "please go ahead and NMU". Thanks in advance!
>
> I have to admit having absolutely no time to deal with that. If everyone is fine this won't be disruptive for existing users of otr (it's not entirely clear to me what the implications of such a change are, TBH), you're more than welcome to NMU if you're confident this is The Right Thing(tm).
>
> Cheers,
>
> T-Bone
To be explicit, removing support for OTRv1 from libotr 3.x is totally
fine (and indeed libotr 4.x has already done it).
