Freeplay after 12 months - honest review of features and bugs

Hey folks! datadog_dave here 👋. I've been using Freeplay as my main side-project observability stack for a full year now. Wanted to share my honest thoughts on where it shines and where it still stumbles.

**What I genuinely love:**
* The **unified query layer** is fantastic for small teams. Writing a single query that joins logs, traces, and metrics saved me so much time building dashboards.
* **Cost predictability** – their pricing model is simple and I've never had a surprise bill, which is more than I can say for some other tools.
* The **playbook integration** is slick. Linking a query directly to a runbook for a specific error pattern is a workflow I now miss in my day job.

**The rough edges (bugs/quirks):**
The alerting engine sometimes has a mind of its own. I had a critical alert rule silently stop firing for two days last month. Support said it was a "state propagation issue" and fixed it quickly, but still 😬. Also, the log parsing UI is buggy – if you try to define a custom parser with a complex regex, the preview often freezes. I've had to drop to the CLI to get it done.

Here’s a snippet of the config I use for their synthetic checks, which works well:
```yaml
checks:
- name: "API Health Check"
endpoint: "https://api.myapp.com/health"
method: GET
assertions:
- type: "status_code"
value: 200
- type: "response_time"
value: 1000
operator: "<"
```

**Verdict after 12 months:**
It's a solid choice if you're a small team or an indie hacker who wants a unified view without managing five different agents. The bugs are annoying but not deal-breakers, and their support is responsive. For my next project, I'll probably stick with it, but I really hope they stabilize the alerting system.

Would love to hear if others have had similar experiences, especially with their distributed tracing – it's gotten better, but I still sometimes miss the depth of Datadog's flame graphs.