[OTR-dev] Length of random exponents in SMP

[Ian Goldberg]

On Thu, Aug 08, 2013 at 01:41:41PM +0200, Philipp Balzarek wrote:
> OTR v3 protocol descriptions states regarding the random exponents in
> SMP under "Receiving a type 2 TLV (SMP message 1)": "The random
> exponents are 1536-bit numbers." 
> But in "User requests to begin SMP" it says that a2, a3, r2 and r3
> should be 128 bits each. 
> Could someone please clarify if this intentional or an oversight? Am I
> missing something?

No, that indeed appears to be a typo in the spec.  All random exponents
should indeed be full length (1536 bits).  I'll fix it in git, thanks!

   - Ian