<div dir="ltr">This looks really interesting Jack.<div><br></div><div>A few quick questions:</div><div><br></div><div>1) Why MessagePack vs other binary serialisation formats? (i tend to use protobufs)</div><div><br></div><div>2) When staying in binary, what sort of overhead does the format impose?</div><div><br></div><div>3) If you imagine a mix network for routing of small binary messages, is saltpack an appropriate format to use for protecting the messages in your estimation? Or are there gotchas that its replacement-for-pgp design would create for the case of pure machine-to-machine messaging?</div><div><br></div><div>4) MIME type? Could you maybe forbid/strongly discourage in the spec emails that contain ASCII armoured saltpack messages? I think some clients have struggled in the past with the UI for showing a message that contains partially signed and partially not signed text, as they tend to treat the signedness of a message as a boolean. Formally forbidding mixing of the two can solve that.</div><div><br></div><div>5) The format appears to be at least partly defined through unversioned reference to a particular library (NaCL). In particular it does not specify what a "NaCL public key" actually is (curve25519 presumably). That seems like it should be fixed for a realistic spec.</div><div><br></div><div>6) It'd be nice if there was a way to embed X.509 cert chains (i.e. signed curve25519 certificate) into the headers, to allow the sender to authenticate themselves with a PKI instead of Keybase. Then it could act as a competitor to CMS.</div><div><br></div><div>thanks,</div><div>-mike</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Feb 5, 2016 at 6:30 AM, Jack O'Connor <span dir="ltr"><<a href="mailto:oconnor663@gmail.com" target="_blank">oconnor663@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">> I don't think that's right. If an attacker creates a new ephemeral,<br>
> they won't be able to encrypt the original payload key. All they'd<br>
> accomplish is forging a message that decrypts to gibberish.<br>
<br>
</span>Oops, you're totally right.<br>
<br>
Thanks for letting us steal so much of your time, Trevor. If you're<br>
ever near the Keybase office in SF or NYC, we owe you a beer!<br>
<span class="HOEnZb"><font color="#888888"><br>
- Jack<br>
</font></span><div class="HOEnZb"><div class="h5">_______________________________________________<br>
Messaging mailing list<br>
<a href="mailto:Messaging@moderncrypto.org">Messaging@moderncrypto.org</a><br>
<a href="https://moderncrypto.org/mailman/listinfo/messaging" rel="noreferrer" target="_blank">https://moderncrypto.org/mailman/listinfo/messaging</a><br>
</div></div></blockquote></div><br></div>