[User Feedback] Improve JS example by splitting 'crash report' and 'stand-alone user feedback' use cases

[bruno-garcia]

Core or SDK?

Platform/SDK

Which part? Which one?

JS

Description

As reported by a user here:

• https://github.com/getsentry/sentry-javascript/issues/13001

Our docs currently leads user to capture a feedback by calling first captureMessage to get the event_id. This is left over from our product capabilities before adding the ability to capture feedback outside an event.

Suggested Solution

I don't believe there's any need to show captureMessage. Since the use cases are 1: On Error, and 2: Stand alone. The message example was a 'work around' used in the past to allow users to capture a feedback stand-alone/outside an error.

We should have 2 snippets, one for each of the following use cases:

1. Collect a user feedback at any point in time

Show the user how to call captureFeedback passing values.

3. Collect a user feedback when an error happens.

Similar to how the crash report modal works. Originally documented this using Sentry.lastEventId and a beforeSend snippet that called captureUserFeedback.

[aliu39]

Link to docs: https://docs.sentry.io/platforms/javascript/user-feedback/#user-feedback-api

[bruno-garcia]

Note that we claim that:

Sentry needs the eventId to be able to associate the user feedback to the corresponding event.

This is not longer true since we introduced captureFeedback (it was true for captureUserFeedback) but the docs have the example below without id as "you can add more context` so at no time we say that the id is optional now.

[getsantry[bot]]

Routing to @getsentry/product-owners-user-feedback for triage ⏲️

[aliu39]

part of https://github.com/getsentry/sentry-docs/issues/11635