Re: [Taler] overline Typo or misunderstanding

[Christian Grothoff]

Hi!

You're right, as you can see in
 http://api.taler.net/api-mint.html#post--refresh-reveal
we don't send the b^{(i)} in that step in the specification or the
implementation, as you say in step (A): they're redundant.

So implementation and API specification are correct. Fixed in the paper
now as well.

Thanks for spotting!

Christian



On 10/04/2015 07:03 AM, Fabian Kirsch wrote:
> Hi,
> 
> there is the following assignment in step 7 of refresh:
> 
>     $\overline{B^{(i)}} :&= B_{b^{(i)}}(\overline{C_p^{(i)}})$
> 
> I am not shure if you intended to overline the $b^{(i)}$ as well.
> 
> A) if it is overlined:
>   $\mathfrak{R}$ contains $b^{(i)}$ but that is never used.
> 
> B) if it is not overlined:
>   Step 7 does not find out whether $\overline{b^{(i)}}$ recovered from
> $E^{(i)}$ is valid.
>   The link can be easily broken by $E^{(i)}$ being calculated from a
> false blinding factor.
>   You should add a check $\overline{b^{(i)}}=b^{(i)}$ in Step 7.
> 
> greetings
>   Fabian
> 
> this mail's math is readable at:
> http://quicklatex.com/cache3/8c/ql_6bb084100c39076512b05dd4f268b38c_l3.png
>

signature.asc
Description: OpenPGP digital signature