Skip to content

docs: file the v0.2.0 changes under v0.2.0 - #16

Merged
Fiona2016 merged 1 commit into
publishfrom
docs/changelog-v0.2.0-bridge-fixes
Aug 7, 2026
Merged

docs: file the v0.2.0 changes under v0.2.0#16
Fiona2016 merged 1 commit into
publishfrom
docs/changelog-v0.2.0-bridge-fixes

Conversation

@Fiona2016

Copy link
Copy Markdown
Collaborator

Two things, both aimed at the v0.2.0 release notes.

Two fixes had no entry at all

Both are the kind a reader needs, and both describe behaviour that predates this version — so the notes say so:

  • A renderer hangs when it starts before the SDK is ready. The bridge preload asks over a synchronous channel, and Electron never releases a synchronous request nothing answers. The window stays blank and unresponsive for good. It reaches 0.1.0 too, through the preload dd-trace ships, and two of the four ways in need no ordering mistake at all: init() rejecting its configuration (a mistyped clientToken hung every window), or never being called.
  • A session that timed out could not be renewed. Renewal ran on click actions forwarded over the bridge, so it depended on the Browser SDK still collecting — and from browser-rum 0.0.7 it stops while the host reports no session. Also already broken in 0.1.0 for anyone running with trackUserInteractions off, where the signal never fired at all.

The rest were filed under a released version

They sat in ## [0.1.0] — the section that opens "First FlashCat release". Left there they would claim 0.1.0 shipped fixes it does not have, and prepare-release.ts writes a fresh section per version rather than renaming that one, so nothing would have moved them at release time.

## [0.2.0] now exists and holds them, plus the minimum @flashcatcloud/browser-rum version under Notes.

No prose was rewritten in the entries that moved.

Checked

#13 still merges into publish cleanly with this applied — verified by a test merge, not assumed.

Two fixes had no entry at all, and both are the kind a reader needs: a
renderer that hangs before the SDK is ready, which 0.1.0 has too, and a
session that could never be renewed after its first timeout.

The rest were filed under `[0.1.0]`, which is the released section — it opens
with "First FlashCat release". Left there they would claim 0.1.0 shipped
fixes it does not have, and the release script writes a fresh section per
version rather than renaming that one, so nothing would have moved them
later.

Both new entries say where the behaviour predates this version. The renderer
deadlock reaches 0.1.0 through the preload dd-trace ships, and renewal was
already impossible for anyone running with `trackUserInteractions` off.
@Fiona2016
Fiona2016 merged commit 0267056 into publish Aug 7, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant