[noise] Proposal: certificate and private key format

[Rhys Weatherley]

On Tue, May 3, 2016 at 6:46 AM, Tony Arcieri <bascule at gmail.com> wrote:

>
> Granted this is generally less of a problem with Protobufs than other
> formats, but it seems like one with solving with some standard method than
> an ad hoc format-specific invention...
>

It's not too "ad hoc".  The protobufs encoding specification recommends
minimal integer encoding and field ordering by tag number anyway, so the
"canonical" format is the same as the "recommended" format for protobufs.
If you use a regular protobuf library in the recommended way to write then
canonical should "just happen".  The standard encoding does allow for other
field orderings so that multiple protobufs can be concatenated to create
one larger parsable protobuf.  We don't need that except for certificate
chains.

And yes, I do realise the irony of saying "MUST be canonical" one day, and
then "canonical maybe doesn't matter" the next. :-)

Cheers,

Rhys.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://moderncrypto.org/mail-archive/noise/attachments/20160503/e4c2025a/attachment.html>