Re: [netconf] type for a PSK's "id" node?

[Henk Birkholz <[e-mail]>]

Hi all,

this might be pretty much "just in time", but let me elaborate a bit here:

After re-exploring the source of PSK Identity (hints), I arrived at PSK 
Identity Encoding again:

> https://tools.ietf.org/html/rfc4279#section-5.1

If that is the PSK "id" that you are looking for, that always involves a 
conversion to UTF8 encoded bytestrings. If that is interoperable with 
the type string here, that is fine, I'd say. Interoperability already is 
a rather painful compromise here in (D)TLS. Please take care to adhere 
to the encoding guidance, therefore. I am not sure, if you want to cover 
identity hints, too:

> https://tools.ietf.org/html/rfc4279#section-5.2


Viele Grüße,

Henk

On 21.07.20 16:44, Kent Watsen wrote:
> Hi Henk,
> 
> I’m trying to close a couple issues on list before the meeting...
> 
> Below you’ll note the "is this the right type?” comment.  Currently the 
> “id” node is type “string”, what type is used by TLS?
> 
> 
> case*psk*{
> if-feature*psk-auth*;
> container*psk*{
> description
> "Specifies the server identity using a PSK (pre-shared
>                or pairwise-symmetric key).";
> uses*ks:local-or-keystore-symmetric-key-grouping*{
> augment"local-or-keystore/local/local-definition"{
> if-feature"ks:local-definitions-supported";
> description
> "Adds an 'id' value when the PSK is used by TLS.";
> leaf*id*{
> typestring; // FIXME: is this the right type?
> description
> "The key id used in the TLS protocol for PSKs.";
>                  }
>                }
>              }
>            }
>          }
> 
> 
> K.
> 
>