Re: [Json] The text in draft-ietf-json-text-sequence

On 2014/07/08 05:57, Paul Hoffman wrote:
> Greetings again. There is still disagreement in the WG about what the text separator should be for draft-ietf-json-text-sequence. The current draft, draft-ietf-json-text-sequence-04, specifies it as the LF character, and gives rationale for that decision in Section 2.2. Many people in the WG have expressed a preference for a different character, one that would not require any stripping from the normal whitespace that can exist inside the JSON text. There were many suggestions for such a different characte
 r;
>    the one that seemed most favored was the ASCII record separator (RS) character, U+001e.
>
> In order to simplify the decision, I'd like to ask a few simple questions:
>
> - Do you feel that the protocol will be significantly broken if it uses LF as the text separator?

Yes.

> - Do you feel that the protocol will be significantly broken if it uses RS as the text separator?

Yes.

> - Do you feel that using LF or RF as the text separator is essentially the same for the protocol?

Yes.

> Answers of "yes", "no", and "don't know" are most appreciated. Explanation of a "yes" and "no" are probably not helpful at this time because they are likely to cause the thread to go off into discussion of the explanation, not of the questions.

[The above are just a reflection of the fact that I think that this 
format (it's not really a protocol) is unnecessary and therefore broken 
from the start, and therefore it's broken with whatever separator there 
is (except with ',', i.e. simply as a JSON array). The problem of 
diverging opinions is best solved by abandoning this spec.]

Regards,   Martin.

> Wearing my chair hat, I'm not sure how I will respond depending on different patterns in the results. Let's answer the questions and then see what we think it means.
>
> --Paul Hoffman
> _______________________________________________
> json mailing list
> json@ietf.org
> https://www.ietf.org/mailman/listinfo/json
>