search

Doichain / dapp

Doichan dApp - a Meteor.js implementation of the "Doichain Atomic Double-Opt-In" Protocol
https://www.doichain.org
MIT License
6 stars 6 forks source link

Using standard signMessage / verifyMessage instead of bitcore functions - privKey procedure #4

Closed silkroadnomad closed 6 years ago

silkroadnomad commented 6 years ago

After sending a new doi-in request via:

curl -X POST -H 'X-User-Id: xxxxx' -H 'X-Auth-Token: yyyyyy' -i 'http://x.x.x.x/api/v1/opt-in?recipient_mail=my-email@mydomain.com&seder_mail=info@doichain.org'

after a new block got successfully mined, we receive a
"Invalid address (aNpxxxxxxxx) [namecoin.firstUpdate.exception"

checking this address from the log via "namecoin-cli validateaddress" also results into a invalid address.

It seems like the algorythm for generating an address from the public key (from DNS txt entry doichain-opt-in-provider=example-domain.org) works differently in mainnet then in regtest.

Ocean15 commented 6 years ago

The version bytes were for the regtest. I changed them to mainnet in 6e729ba71b72431400a2cdca5150c16f734ce162

I will implement a setting to switch between mainnet and regtest soon.

silkroadnomad commented 6 years ago

I'd like to reopen this issue, since I stumbled across this issue yesterday in "Confirm-dApp" again, look how I solved it:

https://github.com/Doichain/dapp/blob/0.0.2-hotfix/server/api/namecoin.js#L31

I guess I have to make some changes also in Send - dApp and Confirmation - dApp. This way we do not have to implement a switch, since this should originaly happen inside the blockchain.

silkroadnomad commented 6 years ago

changed it in places where it seems useful. left it the old way where it causes more work then before. don't know if we still need a RegTest switch for that and in which places.