Re: [Lwip] Discussion about IoT Device Classes

Hi Carsten,

thanks for the initiative. I support it!

Coarsly, from the software perspective, there are 3 classes of devices:

- "motes" with fixed, simplistic functionality in software, on which
resources are so constrained that an operating system and (secure) software
updates do not make sense.

- "low-end IoT devices" with more resources and more functionalities in
software, which run an OS but cannot run generic operating systems such as
Linux or equivalents/derivatives, and hence run IoT-specific operating
systems such as RIOT, Contiki etc.

- "high-end IoT devices" which have enough resources so that they can run
generic operating systems such as Linux or equivalents/derivatives.

Each category presents specific challenges, but the "low-end IoT device"
category is the one where the most fundamental progress is expected, and
achievable.
By that I mean that we can hope to transform low-end IoT devices into
"standard" Internet citizens if we do things right.
On that level, there is no hope for motes and, on the other hand, high-end
IoT devices are already Internet citizens.

>From that perspective, I'm not sure defining a "Class 7" would be useful.
I'm not even sure if defining a "Class 0" is so useful either in the end --
if we have no hope that such devices will become "standard" Internet
citizens.

Best,

Emmanuel

On Wed, Feb 1, 2017 at 4:01 AM, Christian Groves <cngroves.std@gmail.com>
wrote:

> I did a quick read of the draft and to me its not clear what the goal of
> clause 5.2 "Class of Internet Integration" is.
>
> The table talks about internet technologies, the text makes reference to
> communications patterns (e.g. device-to-cloud) whereas the section is on
> integration. It also lists I9 which seems to suggest there will be
> "degrees" on classes of integration between I1 and I9.
>
> So is the aim to only have 3 types? e.g.
>
> Device Internal IP usage - IP Interoperability not an issue.
>
> Device to Provider Server - IP Interoperability within a service provider.
>
> Device to Any - IP interoperability required between multiple service
> providers.
>
> Or to have something more specific to IP listing what parts of the IP
> suite are supported?
>
> Regards, Christian
>
>
>
>
> On 26/01/2017 6:57 PM, Carsten Bormann wrote:
>
>> On 26 Jan 2017, at 00:38, Carsten Bormann <cabo@tzi.org> wrote:
>>
>>> Sure.  We started that discussion a few IETFs ago and have a bis draft
>>> out at
>>> draft-bormann-lwig-7228bis.
>>>
>> … and the editors’ draft is now at:
>>
>> https://github.com/lwig-wg/terminology
>> and
>> https://lwig-wg.github.io/terminology/
>>
>> Issues and pull requests are welcome.
>> (Please see https://github.com/lwig-wg/terminology/blob/master/CONTRIBUT
>> ING.md ).
>>
>> (The proposed bis document is an individual submission at this point; we
>> still put it up under the “lwig-wg” organization as there appears to be
>> some interest.)
>>
>> Grüße, Carsten
>>
>> _______________________________________________
>> Lwip mailing list
>> Lwip@ietf.org
>> https://www.ietf.org/mailman/listinfo/lwip
>>
>
> _______________________________________________
> Lwip mailing list
> Lwip@ietf.org
> https://www.ietf.org/mailman/listinfo/lwip
>