Notes
Things that break between a repository and a URL
Write-ups on the unglamorous half of shipping a website: which region served it, what the cache did, whether a crawler could read it. Also the closest thing this site has to editorial content.
-
Static HTML still wins the boring races
Why a directory of files remains the most predictable thing you can put behind a CDN, and what you give up when you stop doing it.
-
Using htmx without hurting your SEO
Partial swaps are fine. Replacing your URLs, your titles and your server-rendered content is not. Where the line sits.
-
Reading the headers your edge actually sends
Cache status, region codes and request identifiers - what each one tells you when a deployment looks wrong from the outside.