Before you begin
You have a small local page to inspect. The simulated cases need no hosting account.
What you are working towards
A repeatable publishing checklist and a method for diagnosing a failed resource.
Follow the stations in order, or revisit a topic when you need it. Use your browser’s Back command to return here. This path does not track or save your progress.
Understand what the server sends
Distinguish source, generated output and the files served to visitors.
Resolve the exact address
Predict an asset URL from a nested page and from a deployment under a subdirectory.
Check every letter
Compare folder names, filename and extension exactly.
Trace a change through delivery
Locate the edit in source, generated output and the received response.
Check secure dependencies
Distinguish mixed content from a missing file or a policy restriction.
Verify missing-route behaviour
Test a genuinely missing URL as well as a valid page.
Follow the evidence to a repair
Complete the case and state what you would verify after publishing.
Publishing extension: make a useful sitemap
After the site works, help crawlers discover its intended public pages. Write a small sitemap.xml, or generate one with sitemap.download and review it. Finish with three sitemap checks. The generator is another project by Mario; its proxy and crawl limits are explained in the guide.
Your final task
Prepare a release checklist for a small site you own. If you publish it, check the exact deployed address, a nested page, an asset and a missing route. Otherwise, apply the same reasoning to the supplied trails.
Check your understanding
- The upload contains the intended current output, without an extra outer directory.
- Requested paths match deployed names, including case.
- The received stylesheet contains the intended change.
- A missing route gives the expected response and a useful recovery page.
- You can distinguish a delivery problem from a styling problem.
Using AI along the way
You may ask an AI tool to explain a line or suggest an alternative. Then predict the result, inspect the output and test it yourself. Do not treat a confident answer as proof. This project was itself developed with substantial AI assistance: read the transparent account and Mario’s web journey.
Practise predicting what code does · Practise reviewing AI code