[dispatch] Re: IETF WG state changed for draft-ietf-dispatch-mime-protobuf

[John C Klensin <[e-mail]>]

FWIW, I'd be less unhappy if the proposed language referenced PRECIS
as well as draft-bray-unichars.  If the statement is going to be
"...can be quite complex with problems discussed, for example,
in...", then the examples should include some discussion of the
really complex cases rather than only the relatively more
straightforward cases covered by the Unichars documents.

Also, if one is being careful, at least one of the two examples in
that paragraph that mention "UTF-8" is probably wrong:  these issues
are associated with Unicode generally and not with UTF-8.  The only
important condition on UTF-8 is that it be well-formed.

    john




--On Tuesday, July 15, 2025 19:24 -0700 "Murray S. Kucherawy"
<[e-mail]> wrote:

> Tim, are you good with this alternative proposal?
> 
> -MSK
> 
> On Tue, Jul 8, 2025 at 9:57 AM Rob Sloan <[e-mail]> wrote:
> 
>> > Would be happy to suggest language, but this isn't rocket
>> > science
>> 
>> I have concurrently added this patch
>> <https://github.com/wkumari/draft-murray-dispatch-mime-protobuf/co
>> mmit/c666f8e1c6810cbc54b0e93658e1ae6ec3dc5886> which is a bit more
>> general: please let me know whether that looks good to you and
>> feel free to comment directly if not.
>> 
>> On Tue, Jul 8, 2025 at 12:32 PM Tim Bray <[e-mail]>
>> wrote:
>> 
>>> On Jul 8, 2025 at 10:26:23 AM, Rob Sloan <rmsj=
>>> [e-mail]> wrote:
>>> 
>>>> Hi Tim,
>>>> 
>>>> > applies no restrictions to the contents of strings
>>>> 
>>>> I am happy to add a section to encourage the safe handling of
>>>> Unicode strings within a Protobuf encoding and cross-reference
>>>> to this document. Your point that special or unassigned code
>>>> points may cause application misbehavior is certainly well taken.
>>>> 
>>> 
>>> Would be happy to suggest language, but this isn't rocket
>>> science: "Be aware that some Unicode code points can cause
>>> problems, which are discussed in [UNICHARS] and [PRECIS]."
>>> 
>>> Note that parse-time UTF-8 format enforcement on string fields is
>>>> optional
>>>> <https://protobuf.dev/editions/features/#utf8_validation> for
>>>> backward compatibility with proto2.
>>>> 
>>> 
>>> Didn't know that. Blecch.  Might be worth mentioning in the doc?
>>> 
>>> -T
>>> 
>> 



--On Tuesday, July 15, 2025 19:24 -0700 "Murray S. Kucherawy"
<[e-mail]> wrote:

> Tim, are you good with this alternative proposal?
> 
> -MSK
> 
> On Tue, Jul 8, 2025 at 9:57 AM Rob Sloan <[e-mail]> wrote:
> 
>> > Would be happy to suggest language, but this isn't rocket
>> > science
>> 
>> I have concurrently added this patch
>> <https://github.com/wkumari/draft-murray-dispatch-mime-protobuf/co
>> mmit/c666f8e1c6810cbc54b0e93658e1ae6ec3dc5886> which is a bit more
>> general: please let me know whether that looks good to you and
>> feel free to comment directly if not.
>> 
>> On Tue, Jul 8, 2025 at 12:32 PM Tim Bray <[e-mail]>
>> wrote:
>> 
>>> On Jul 8, 2025 at 10:26:23 AM, Rob Sloan <rmsj=
>>> [e-mail]> wrote:
>>> 
>>>> Hi Tim,
>>>> 
>>>> > applies no restrictions to the contents of strings
>>>> 
>>>> I am happy to add a section to encourage the safe handling of
>>>> Unicode strings within a Protobuf encoding and cross-reference
>>>> to this document. Your point that special or unassigned code
>>>> points may cause application misbehavior is certainly well taken.
>>>> 
>>> 
>>> Would be happy to suggest language, but this isn't rocket
>>> science: "Be aware that some Unicode code points can cause
>>> problems, which are discussed in [UNICHARS] and [PRECIS]."
>>> 
>>> Note that parse-time UTF-8 format enforcement on string fields is
>>>> optional
>>>> <https://protobuf.dev/editions/features/#utf8_validation> for
>>>> backward compatibility with proto2.
>>>> 
>>> 
>>> Didn't know that. Blecch.  Might be worth mentioning in the doc?
>>> 
>>> -T
>>> 
>>