Re: [core] draft-ietf-core-senml - Last Call comments

[Jim Schaad <[e-mail]>]

Carsten,

Sorry for the slow reply on this, I needed to figure out what my problem was and why I was not getting a good interpretation.

My real issue boils down to the table that you are asking IANA to create as a registry.  I believe that this table is missing a number of columns and some of the data in it may be considered to be wrong.  The fact that I asked for the column 'XML Type' to be changed to 'Type' is probably not helping here.

I think that the columns for the table should probably be:

Name, Label, Type, CBOR Key, XML/JSON Type, EXI ID, Note

The Type column holds the abstract type, which as long as it maps directly to a CBOR Type means that column is not needed.

The XML/JSON type for "Data Value" should probably hold "Base64 string" which, while the expected item, should be made explicit.
 

Jim


> -----Original Message-----
> From: Carsten Bormann [mailto:[e-mail]]
> Sent: Friday, October 27, 2017 6:55 AM
> To: Jim Schaad <[e-mail]>
> Cc: [e-mail]; [e-mail]
> Subject: Re: draft-ietf-core-senml - Last Call comments
> 
> On Sep 23, 2017, at 07:24, Jim Schaad <[e-mail]> wrote:
> >
> > * Section 6 - I am sad, but I understand the reasons for the decision,
> > that Octets are not encoded as binary values.
> 
> Hi Jim,
> 
> the intention is that the bytes (“octets” in French) of the data value are
> encoded as a CBOR byte string;
> 
>    o  Characters in the String Value are encoded using a definite length
>       text string (type 3).  Octets in the Data Value are encoded using
>       a definite length byte string (type 2).
> 
> Is that not binary enough?
> 
> Grüße, Carsten