Author: Amir Taaki Date: To: Jorge Timón CC: libbitcoin@lists.dyne.org Subject: Re: [Libbitcoin] "secp256k1 ready and well tested"
definitely should be used.
I think if it was ported with a pull request, eric would accept it.
On 01/09/2015 07:57 PM, Jorge Timón wrote: > Note that it is not ready and will only be used for signing for now:
> http://www.reddit.com/r/Bitcoin/comments/2rrxq7/on_why_010s_release_notes_say_we_have_reason_to/cnj1wan > Hopefully by 0.11 things will be ready for using libsecp256k1 for
> validation too.
>
> On a related subject...are there any plans to use libconsensus within
> libbitcoin/darkwallet?
> Right now it only has a function for script verification, but the goal
> is to eventually expose all consensus-critical checks so that
> alternative bitcoin implementations using it can minimize their
> consensus risks (getting the "bug by bug consensus compatibility"
> without having to use the rest of bitcoin core's code).
>
>
> On Fri, Jan 9, 2015 at 2:04 PM, Amir Taaki <genjix@???> wrote:
>> http://www.reddit.com/r/Bitcoin/comments/2rrxq7/on_why_010s_release_notes_say_we_have_reason_to/ >>
>> _______________________________________________
>> Libbitcoin mailing list
>> Libbitcoin@???
>> https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/libbitcoin