Hi Uri,

On 6/7/16 7:32 PM, Blumenthal, Uri - 0553 - MITLL wrote:
>
>       * We now include a signature mechanism for the MUD files.  It
>         was always the plan to do this.  There were two choices:
>         CMS/PKCS#7 or JWS.  Again for tooling's sake, so that people
>         don't need to roll their own, especially for anything security
>         related, we've gone with CMS and a detached signature at
>         that.  Thanks to John Bashinsky and others for their advice on
>         this.  This area in particular could stand close scrutiny.
>
> Wouldn’t CMS still require serialization/canonicalization?
Yes/No, depending on what you mean.  Currently we're treating the MUD
file as binary precisely to avoid canonicalization.  The transport for
all of this is HTTPS and UTF-8.
>
> Tooling-wise, OpenSSL is indeed prevalent (and seems to do CMS quite
> well) – but  JWS tools are around, so you wouldn’t need to roll your
> own if you decided to go that way.

I'm not saying there are no tools around, but I personally found slim
pickings in the form of various libraries.  If people think we should go
in a different direction...  The key here though is not whether I would
want to roll my own but whether others would be able to do the same. 
This is a pretty large uplift for some, and one goal is to make that as
painless as possible.

>
> Do I need the ability to tell whether a MUD file was not signed or its
> signature was deleted? 

The spec requires that all files be signed and that the signatures be
verified prior to processing.

And thanks for again for the comments.

Eliot

Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
OPSAWG mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/opsawg

Reply via email to