Re: [Json] JSON and int64s - any change in current best practice since I-JSON

[Carsten Bormann <[e-mail]>]

On 2024-01-17, at 06:01, Tim Bray <[e-mail]> wrote:
> 
> […] I’ve never managed to convince myself that the advantages of binary over text are worth the aggravation.

Eleven years ago, I would probably have sympathized with this sentiment.
When I worked with SGML and later XML, I was so happy to replace ASN.1 and a boatload of ad-hoc binary formats with a text-based format.
JSON certainly could (mostly) displace XML only because it also kept the tradition of being text-based.

However, this is no longer the world we live in.

Today, we are using all kinds of tools to look at interchanged data.
Even JSON is usually packed for interchange and needs a pretty-printing tool to be useful as a text-based format.
CBOR diagnostic notation is a way to pretty-print CBOR that is as useful (upwards compatible with JSON, and, by the way, now [1] fixes most of the really jarring aspects of JSON that we inherited).

HTTP started as a text-based format in the 1990s, and we have since made an almost complete transition to HTTP/2 and HTTP/3, which use binary representation.  

Most of our interchanges today are encrypted anyway; this and other crypto content (signatures, keys, endorsements…) lead to tons of base64 (or worse, hex!) content in nominally “text-based” formats.  Again, a tool is needed to look at these.

The age where text-based formats were to be preferred for interchange is pretty much over.

CBOR diagnostic notation shows that you can have the remaining crumbs of that cake and eat the simplicity of the binary cake, too.

Grüße, Carsten

[1]: https://datatracker.ietf.org/doc/draft-ietf-cbor-edn-literals/