Re: [scim] device model and scope

[Eliot Lear <[e-mail]>]

Hi Peter,

This was something suggested last summer.  I don't have a particular use 
case.

Eliot

On 28.03.2024 09:59, Liuchunchi(Peter) wrote:
>
> Hello Elliot,
>
> Sorry for necrobumping such an old thread. But this non-IOT use 
> case/device model was discussed during Brisbane scim wg meeting and I 
> am not sure if I understood correctly.
>
> Can you suggest a few more specific instances of such non-IOT devices?
>
> Best,
>
> Peter
>
> *From:* Eliot Lear <[e-mail]>
> *Sent:* Sunday, August 20, 2023 5:43 PM
> *To:* SCIM WG <[e-mail]>
> *Cc:* [e-mail]
> *Subject:* [scim] device model and scope
>
> During the f2f meeting (in which my face was on a screen ;-) I think 
> it was Danny who suggested that the device model needs to be expanded 
> to handle the non-IOT use case.  That's a nice meaty topic to take on, 
> so let's start going there.
>
> I had suggested one of two paths: either incorporate some changes to 
> cover that use case, or separate out two models. The basis for the 
> choice, I think, is commonality.  If there is substantial overlap, 
> then either incorporating additional elements, or creating a standard 
> extension, might be a way to go.
>
> "IOT" is one of those nebulous things, and can sometimes get in the 
> way.  Right now, I could /imagine/ some additional elements for 
> “non-IOT”, which I'll take to mean computers that humans use that have 
> no or limited direct cyber-physical capabilities.  Some of those might 
> include an owner, some policy elements, VPN parameters, etc.  One 
> could imagine the SCIM model reaching all the way into MDM 
> parameters.  But that's just me imagining things without a good 
> understanding of what Danny had in mind.
>
> What I propose is that we get into a discussion about this, and see 
> how far we can get before Prague.  I think the next step, though, is 
> to understand what the new objects/elements would be.
>
> Thoughts?
>
> Eliot
>