A public changelog and a live status page
The page you are reading is new. Every release now lands on a public timeline with the detail behind it, and the platform health we watch internally is on a status page anyone can open, with no account needed for either.
Highlights
- A public changelog timeline at /changelog, no sign in required
- Each release opens into the individual changes that shipped with it
- Search and filter by release channel across stable, beta and hotfix
- A live status page covering the database, Redis and the send queues
- A health indicator in the header that links straight to the status page
In this release4
- NewSystem
Release notes are now a public timeline
Until today the only way to find out what changed was to notice it. Releases are now published to a timeline you can read, link to and search, and you do not need an account to open it.
What you get:
- A rail of releases newest first, grouped by year, with the date alongside each one on wide screens.
- Every release opens into the individual changes that shipped with it, each tagged with its type, the area it touched and how much impact it carries.
- Search across version, title, summary and highlights, plus a filter for the stable, beta and hotfix channels.
- Older releases load on demand rather than all at once, so the page stays fast as the history grows.
- The whole thing is responsive, so the timeline reads the same on a phone as it does on a desktop.
Shipped as commit 9c23843 in wbsender-apps and d242617 in wbsender-api.
BeforeChanges shipped silently. You found out when something looked different, and there was nothing to link a teammate to.
AfterEvery release has a public page with its full detail, and a URL you can send to anyone.
- NewSystem
A live status page for the platform
The health checks we watch internally are now on a page you can open yourself at any time.
What it shows:
- Database and Redis state, reported as operational, connecting or down rather than a single green or red light.
- Depth of the campaign and message queues, so you can see whether a slow send is your account or the platform.
- API latency and process uptime, sampled every minute and kept for 90 days.
- A health pill in the changelog header that reflects the same state and links straight through.
The status endpoint is public and rate limited, so it works for signed out visitors and for an external uptime monitor if you point one at it.
- ImprovedSystem
Before and after comparisons inside a release
Some changes are easier to show than to describe, so an entry can carry a comparison.
Three kinds are supported:
- Metric, for a number that moved, rendered with the old value, the new value and the percentage change. The CPU and memory figures in the 1 September release use this.
- Text, for a behaviour that changed, shown as a before and after pair.
- Image, as a slider you drag across the two frames, which also works with touch and with the arrow keys.
- ImprovedSystem
Health checks and release tooling behind the scenes
Groundwork that came with this release and is worth recording.
What changed:
- The existing health endpoint and the new status page now share one probe instead of each running their own database, Redis and queue checks.
- A health snapshot is written every minute, which is what the 90 day history is built from.
- Releases can be generated from git history or written by hand, and there is a prune script that trims the archive to a cutoff date with a full backup taken first.