From joroark at vt.edu Thu Dec 17 20:13:29 2015 From: joroark at vt.edu (Douglas Roark) Date: Thu, 17 Dec 2015 12:13:29 -0800 Subject: [Lightning-dev] Idea: Using DANE to store node data In-Reply-To: References: <87wpsgh28z.fsf@rustcorp.com.au> <87d1u7xoxl.fsf@rustcorp.com.au> Message-ID: <56731769.1080700@vt.edu> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hello. I wanted to throw an idea out into the open and see if anybody had any thoughts. Earlier this year, I worked on an effort to attach payment data to DNSSEC via DANE[1]. This included working on an RFC draft[2] that explained the concept. The idea was to use the DNSSEC system as a database of sorts that could get trusted payment info out to users. If you wanted to pay Top Quality Widgets for their goods, you could look them up on the DNS system and be reasonably certain that you were actually paying them. The setup was designed to allow for all manner of payment methods, with ACH and Bitcoin being the first two to be supported. (Note that the effort eventually fell by the wayside before a much-improved second RFC draft could get out the door; the first draft barely touches Bitcoin at all. If anybody's interested, I can provide a basic outline of what was going to be in it, at least on the Bitcoin side.) I'm wondering what others think of this idea, and how it could apply to Lightning. The main hangup out the gate would be that the payment info has to be attached to a domain (e.g., topqualitywidgets.com) that supports the scheme. This would, in all likelihood, mean custom domains would be needed where support for the scheme was explicitly included. This could be tricky. Despite that hangup, I think this is an interesting scheme that may be worth exploring, perhaps in a modified form. We had the scheme set up where, on the Bitcoin side, the keys were really flexible and could support things like per-user payment addresses (i.e., static addresses weren't mandatory). I thought it was pretty slick, but hey, I'm biased. :) One related-but-separate thing to keep in mind is that, if there is any sort of ID at host scheme deployed in Lightning, it really ought to support internationalization. [3] suggests taking the ID and hashing it with SHA-256. Any thoughts? I really like the idea of having some sort of quasi-permanent method for storing data without having to rely on a complicated scheme to find other nodes in the network. It should be optional, of course, and should work alongside any mandatory schemes to find other users on the network. Thanks. [1] https://en.wikipedia.org/wiki/DNS-based_Authentication_of_Named_Entities [2] https://tools.ietf.org/html/draft-wiley-paymentassoc-00 [3] https://tools.ietf.org/html/draft-ietf-dane-smime-09 - -- - --- Douglas Roark Cryptocurrency, network security, travel, and art. https://onename.com/droark joroark at vt.edu PGP key ID: 26623924 -----BEGIN PGP SIGNATURE----- Comment: GPGTools - https://gpgtools.org iQIcBAEBCgAGBQJWcxdpAAoJEEOBHRomYjkkdDAP/1eCph6RSQJoUAWo4AONWQEp nF9C55nRhNdtoMZI52cTN1gU75PS3HNnEfmE0oejZdlt31E+A3V7ibyzOQaHAPzm ZyVD5/Ke29k+IxPhuk++AMLDJsCbdW/U284NMf3d6yWLZGYCDxaNVEIgs96jHuLN c1Ot827Wcl19D18SZjYSXeCXrYkXEZs5MQuGjw1sPOo/z8UdWHBAY9J8y8+bhS48 4/ZHQMq+XVawrl+p65vIKwhrQvaq0PgtWjHSyG/5lTQEKQMphDppPXQRyyMdbzX8 JO7Agrjp4MVcwccvsW6oCID+o49l3qeLY+A6sBqcgjCRIgcP3pEbsBb8zCWTNMH/ /lTkywMisEGctbdO3IKHXGXlmOLMb96jiwY1UJBr7yws0IxPM5F/Lw7PufojrcPQ uPe/kQ1RSMQaDDQnxGRiY796g9OImNImhZu0WcyRoR91Mu4stJkOABHdwu2voZA8 DrJRGJ+BbP0xV8dSGn1dqpnlAMIqKhWfUc67GPoLdgsPd021VbzlctXfLlQe8cWG qvLwsgnzvpzPn4MmCPrGyGg+cCOgH0XWBUeFPnPUZIFP5bosTGvXUPrsaJw+qkd+ fiYxQlm9QuhNtdmhIMAapbfsSoiRgxhPb8PsKb6tkj/blk4QI8m5p2GRVY+2BBOr LPHm5WXTCn1XgdCXOZM8 =3chy -----END PGP SIGNATURE-----