Consider moving unreleased changes to a separate "Unreleased.md" file

[jamescrosswell]

Perhaps we can think of a different way of adding entries to the CHANGELOG.md, as this is most commonly the merge conflict ... maybe we don't add it directly, but place it "somewhere" from where it's then pushed to the CHANGELOG.md in the correct place when (squash) merging 🤔 ... something for us to think about.

Originally posted by @Flash0ver in #4724 (comment)

[linear]

NET-538

[jamescrosswell]

I think fixing this would require changing this:

sentry-dotnet/.github/workflows/release.yml

Lines 33 to 40 in b3d9b2f

	- name: Prepare release ${{ github.event.inputs.version }}
	uses: getsentry/action-prepare-release@3cea80dc3938c0baf5ec4ce752ecb311f8780cdc # v1.6.4
	env:
	GITHUB_TOKEN: ${{ steps.token.outputs.token }}
	with:
	version: ${{ github.event.inputs.version }}
	force: ${{ github.event.inputs.force }}
	merge_target: ${{ github.event.inputs.merge_target }}

The publish process (which is not handled in this repo) has the logic to patch the changelog.

We could probably change that in a way that would continue to support the existing behaviour but would also/alternatively allow us to put the changes in a separate unreleased.md file or something... so it could attempt to merge changes in from either/both locations.

[Flash0ver]

I like that, as it reminds me of RS2008.