As I said, there is no such thing as a "signal protocol" (as far as I know), there is only the libsignal implementation and a bunch of specs.
> It seems you dont have a SignalProtocol implementation, and are > searching in this XEP for answers how to implement SignalProtocol. That's right, but the XEP doesn't mention that it requires libsignal. Maybe it's only a result of not well defined terms. If you know that the term "signal protocol" means libsignal and that there is no spec for that, everything is clear. On 18.09.2017 16:35, Philipp Hörist wrote: > The section says that the key element contains the encrypted payload, > which was encrypted with a SignalSession. > I think its out of scope for the XEP to describe what a SignalSession > encrypted payload (a SignalMessage) contains. > The XEP is about using the SignalProtocol with XMPP, not about how > SignalProtocol works. > > So it depends on you having a SignalProtocol library. > > It seems you dont have a SignalProtocol implementation, and are > searching in this XEP for answers how to implement SignalProtocol. > > > 2017-09-18 16:16 GMT+02:00 Paul Schaub <[email protected] > <mailto:[email protected]>>: > > Hi! > > > Am 18.09.2017 um 15:27 schrieb Klaus Herberth: > > > > Hi Paul, > > > > thanks for reading that lengthy email. > > > > If I understand you correctly, the complete magic happens in the key > > element and there is no description in the XEP or in the linked > > "signal protocol" which describes it. So all implementations use > > libsignal? I think this is terrible for a protocol and I'm a bit > shocked. > > > > Yeah, you are not the only one thinking so. > > > One more think which I noticed was that somewhere in time the > > namespace changed from the well known urn:xmpp:omemo:0 to > > eu.siacs.conversations.axolotl. What was the reason for this > uncommon > > format and is it not weired that there is the term axolotl, but the > > protocol uses signal now? > > The reason for that is, that there are two different OMEMOs right now. > There is "siacs-OMEMO", which is what's implemented everywhere and > uses > the siacs namespace. You could call it the "real world-OMEMO". And > there > is the "perfect-world-OMEMO", of which no implementations exist and > which uses the omemo namespace. For quite a long time the XEP-0384 was > describing the perfect-world-OMEMO, which was unconvenient, since the > XEP is obviously the first place someone who wants to implement OMEMO > would look at. A few weeks ago the council decided to roll back > the XEP > to the state of the siacs implementation. I'm not sure, if there are > still people working on the perfect-world-OMEMO XEP. I recommend > you to > read the quite lengthy mailing list discussions in the archive about > that topic to fully understand what was/is going on. > > > > > Is there any explanation for the mentioned authentication tag? > > > > I'm not an expert. Maybe Daniel knows more about this? > > Greetings vv > > > _______________________________________________ > Standards mailing list > Info: https://mail.jabber.org/mailman/listinfo/standards > <https://mail.jabber.org/mailman/listinfo/standards> > Unsubscribe: [email protected] > <mailto:[email protected]> > _______________________________________________ > > > > > _______________________________________________ > Standards mailing list > Info: https://mail.jabber.org/mailman/listinfo/standards > Unsubscribe: [email protected] > _______________________________________________
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Standards mailing list Info: https://mail.jabber.org/mailman/listinfo/standards Unsubscribe: [email protected] _______________________________________________
