[Cryptography] Komitments

[Phillip Hallam-Baker]

OK, so I have added the capability to the UDF scheme used in the
MatheMatical Mesh:

https://tools.ietf.org/html/draft-hallambaker-udf-11

Just wanted to make sure I hadn't overlooked something basic while writing
this.

I considered a choice of using an HMAC or concatenating the key into the
second Hash operation. I prefer the second approach on balance. While this
is a keyed hash and HMAC is the primitive for that purpose, the
construction of HMAC isn't really that far from the concatenation approach
and this is a lot simpler to implement and debug. I can probably be
persuaded otherwise.

KeyedFingerprint = <Version-ID> + H (<Content-ID> + ?:? + H(<Data>) +
':' + <Key>)


The Mesh is finally at a stage where (parts) of it are useful as a
standalone capability. I am working on the unit tests etc. to nail things
down.

This first release is pretty boring as all it provides are some digest,
fingerprint and commitment capabilities, the last two in a scheme only I
use at the moment. The mesh should become much more interesting once people
can use the Mesh service as an end-to-end secure password manager. The
release plan is as follows:

1) Standalone tools not requiring key management
2) Profiles, key management, configuration on a single device (no service),
Credential Catalog
3) Personal Mesh service, users can only exchange messages between their
own devices.
4) Contact Catalog, ability to send/receive messages with other users.

Over the past year I have cut the number of lines of code in half while
extending functionality. The major simplification comes about from the use
of a cryptographic container based on a Merkle Tree. Yes, it does look a
little blockchainy but I am pretending not to notice.

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.metzdowd.com/pipermail/cryptography/attachments/20181222/d6eb76c8/attachment.html>