Add clarification about async vs. sync instruments

[reyang]

Changes

• Editorial change - added clarification and examples about sync vs. async instruments.

Related oteps OTEP146

[jsuereth]

Nit: meant to be invoked inline with application/business processing logic, within the scope of [Context](../context/context.md).

Baggage (as specififed) is available via a context. I think you can call it out below as an additional thing, but the key here is;

• The logic is inline with the mainline processing of the application (not sheddable when under load)
• The logic can have automagical attachment of context if desired (and all the fun contexty things).

[reyang]

Please take another look, I believe this has been addressed.

[jkwatson]

👍🏼

[carlosalberto]

Let's merge this on Monday prior to the 1.4.0 release.

[bogdandrutu]

Why html tags like this? Should use markdown headers if needed.

[carlosalberto]

Ping @reyang

[reyang]

It's the same as other tags in this doc (and other spec doc such as the tracing one):

1. we don't want them to be in the top level ToC (since they interrupt the ToC flow)
2. we want to have an anchor

IIRC @bogdandrutu you've asked the same question before?
e.g. https://github.com/open-telemetry/opentelemetry-specification/blob/main/specification/metrics/api.md#instrument-naming-rule

#1578 (comment)

[carlosalberto]

@bogdandrutu Please fill an issue to follow up this if this is still a problem, etc. Thanks!