Re: [Json] Using a non-whitespace separator (Re: Working Group Last Call on draft-ietf-json-text-sequence)

+1 on the separator MUST be an ascii character, not a UNICODE sequence.

Most of my projects don't actually use or need Unicode because they
are low level and don't deal with anything the user needs to see. I
use ASCII tags to make it easier to support non-Unicode languages like
C. Having to move to multibyte Unicode sequences for this one purpose
would be a pain.

If not RS then zero.

On Wed, Jun 4, 2014 at 3:30 AM, Joe Hildebrand (jhildebr)
<jhildebr@cisco.com> wrote:
> On 6/1/14, 6:03 PM, "Nico Williams" <nico@cryptonector.com> wrote:
>
>>5) #4 with ASCII RS as the separator
>
> This is the best one on the list.  It's what RS is for.  If you also want
> to allow optional inter-item ignorable whitespace (e.g. \n) for better
> printability, you get everything you want.  RS is only used for parser
> state synchronization.  RS is forbidden in a JSON text by 7159, so it's
> unambiguously a recoverable parse error if it is detected in the middle of
> an item.  Grep works fine with this format, particularly when combined
> with newlines.  Several editors I tried worked fine with it; one even
> rendered the RS as a nice RS-surrounded-by-box glyph.  It's certainly
> going to be more interoperable with tooling than U+FFFE, for example.
>
> I also agree with Tim that there should be precisely one way of doing item
> separation.  Having to do multiple ways is going to make the parsers
> slower and more complicated than they have to be, for no technical benefit.
>
> --
> Joe Hildebrand
>
>
>
>
> _______________________________________________
> json mailing list
> json@ietf.org
> https://www.ietf.org/mailman/listinfo/json