I think the really cool part about this is you could have the code automatically generate new BM-addresses as often as you like and update the namecoin record at the same time. In this way, you can have your BM-address changing regularly for security, as suggested by BM author, but keep the name, i.e. bitmessage email address secured by namecoin blockchain in human readable form. If you need higher security and change addresses often you will incur more nmc fees and vice versa less security, less fees; you will pay for the security you get, in nmc fees.
Namecoin and bitmessage will work well together since they both solve different elements of resource allocation.
Fortunately BM- addresses don't need to be changed for any security related reason. The reason is to help streams stay a reasonable size as Bitmessage takes on new users- the new addresses will automatically be created in more available streams. We won't ever have to tell people "Hey everybody, it would be good if you replaced some of your addresses with new addresses to free up the over-full streams", rather it will be just fine to say "Feel free to create new addresses and abandon your old ones as much as you like." People will appreciate being encouraged to "waste" addresses.
Maybe. Although it doesn't gain you much and makes things more complicated, e.g. extracting the priv. key from wallet ... and then particularly changing encryption keys regularly. Also I'm not intimate with how bitmessage works but I know Artheros is using separate keys for signing and encryption. Namecoin/bitcoin private keys should probably be kept for signing and not used for encryption as much as I've read about best practices ... icbw.
Indeed. Although no one has yet thought of a vulnerability related to using the same keys for both encryption and signing, all the cryptographers would tell you it is unwise to assume no one ever will.
Concerning Namecoin, I would love to integrate it with Bitmessage; non-human-friendly names are a big complaint. But I would want it to be able to actually scale- which means getting a name would have to cost a non-trivial amount of money. And my next concern is that if 30 million people each generate an average of 4 named addresses, the Namecoin blockchain would be 4 GB of raw data- without counting the overhead.
