Re: [Acme] ACME signature mechanics

[Hannes Tschofenig <[e-mail]>]

Hi Phillip,

the relevant OMA specification can be found here:
http://openmobilealliance.org/about-oma/work-program/m2m-enablers/
(Click on the 'OMA Lightweight M2M (LWM2M) protocol' for the version for
IoT devices and 'OMA Device Management (DM)' for the stuff that is used
on mobile phones.)

Here is a tutorial:
http://community.arm.com/servlet/JiveServlet/previewBody/8693-102-3-15745/ARM%20OMA%20Lightweight%20M2M%20Tutorial.pdf

When you look through the tutorial then you will see that IoT devices
need more than just credentials (which is why the specification talks a
lot about device management).

Here is open source code:
https://github.com/jvermillard/leshan
https://github.com/01org/liblwm2m

The next interop event will take place in January 2015:
http://openmobilealliance.org/oma-lwm2m-testfest-registration-now-open/

> IoT is not going to be a special case of the Internet

The LWM2M specification re-uses the work from the CORE working group
(including CoAP, and CoAP resource server), DTLS, JSON and many IETF
other specifications.

Still, the needs for provisioning a certificate to a Web server are,
however, different from provisioning a light bulb.

Judging from the abstract of the ACME specification their document is
focused on the Web and nothing else. That's fine (delta the duplication
of already existing work in that area).

Ciao
Hannes

On 12/18/2014 03:11 PM, Phillip Hallam-Baker wrote:
> 
> 
> On Thu, Dec 18, 2014 at 8:23 AM, Hannes Tschofenig
> <[e-mail] <mailto:[e-mail]>> wrote:
> 
>     Hi Phillip,
> 
>     we already have a mechanism for issuing certificates to embedded
>     devices, namely OMA Lightweight M2M. It is already used today.
>     That specification is a version of the OMA device management protocol
>     (which is also widely used) but uses different protocols that are more
>     suitable for the embedded side, such as CoAP and JSON.
> 
>     Hence, I doubt that this work is something the IoT community is
>     asking for.
> 
> 
> Is there a pointer to the spec that is publicly accessible?
> 
> In the short term that might be the case. But in the longer term IoT is
> not going to be a special case of the Internet, it is going to be the
> Internet.
> 
> I think you are making a case for looking at the OMA protocol and
> deciding if we can use it.
>