Set encoding rules for bytes represented in text and JSON

[rakyll]

What are you trying to achieve? When writing exporters, it's not clear how bytes should be represented in text or JSON output.

For example, trace ID (16 bytes), span ID (8 bytes) and trace flags can be represented as hex encoded strings.

[andrewhsu]

from the issue triage mtg today, discussed this. is this just about the traceid? if so, may already be specified in the spec (@mwear may know?)

if not, can you also suggest appropriate wording change?

[chulkilee]

Is it for OTLP/HTTP JSON? If so, doesn't Protocol Buffer 3 JSON Mapping define the encoding?

JSON value will be the data encoded as a string using standard base64 encoding with paddings. Either standard or URL-safe base64 encoding with/without paddings are accepted.

(We may use the same encoding for standard output)

[anuraaga]

For OTLP, we use hex instead of the standard proto mapping

https://github.com/open-telemetry/opentelemetry-specification/blob/master/specification/protocol/otlp.md#otlphttp-request

the trace_id and span_id byte arrays are represented as case-insensitive hex-encoded strings, they are not base64-encoded like it is defined in the standard JSON Mapping.

[chulkilee]

Oh, I missed that part 🤦

So the spec says hex (or base16) for OTLP/HTTP request with JSON encoded prototobuf message..

Reference: #786 and #911

Then what's this issue for? For text format? I'm wondering there is written spec for standard output - while the matrix has "Standard output" entry..