Re: [Netconf] Alexey Melnikov's Yes on draft-ietf-netconf-yang-patch-12: (with COMMENT)

[Andy Bierman <[e-mail]>]

Hi,

Here are some examples of registrations:

cellml+xml

The charset parameter of application/cellml+xml is handled in the
same fashion as for application/xml, as specified in Section 3.2
of [RFC3023]. However, per conformance rule 4 of
[CELLML-UMBRELLA], valid CellML Umbrella documents MUST be in the
UTF-8 character set. If the charset parameter is present, it MUST
take the value "utf-8". CellML processing software SHOULD check
the charset field, and if it is present but not equal to "utf-8",
MAY attempt to recover from the error by processing the document
in the specified character set.


3gpdash-qoe-report+xml

Encoding considerations: 8bit
The utf-8 charset is always used for this type.


several use this text:


Encoding considerations:  Uses XML, which can employ 8-bit
   characters, depending on the character encoding used.  See RFC
   3023 [RFC3023], Section 3.2.


YANG data and YANG patch use UTF-8.
The UTF-16 charset is not used.
Can you suggest a registry entry that we should follow
to properly convey this requirement?


Andy


On Sat, Nov 12, 2016 at 6:03 AM, Alexey Melnikov <[e-mail]>
wrote:

> Hi Andy,
>
> On 12 Nov 2016, at 19:13, Andy Bierman <[e-mail]> wrote:
>
>
> On Fri, Nov 11, 2016 at 5:57 PM, Alexey Melnikov <[e-mail]>
> wrote:
>
>> Andy,
>>
>> On 12 Nov 2016, at 02:14, Andy Bierman <[e-mail]> wrote:
>>
>> The media type registrations in RESTCONF say 8-bit.
>>>
>>>
>>> Then it might be broken too.
>>>
>>>
>> It was approved already, so take it up with th IESG
>>
>>
>> Different people on IESG are experts in different things. What you
>> effectively said is that no bug can be accepted after the fact. It doesn't
>> matter in this case, but I do find your attitude disturbing.
>>
> Best Regards,
>>
>
>
> I do not see why RESTCONF and YANG Patch are broken because the
> media types say 8-bit.  Please explain what breaks.
>
>
> I think you misunderstood: I didn't say that RESTCONF and YANG Patch are
> broken, I just said that the IANA registration template is slightly
> incorrect.
>