Today I spotted a second case where an HTML validation test doesn't complete properly - redirect loops.
A redirect loop occurs when a URL that issues a
redirect to another URL (perhaps back to the previous URL) in such a way
that an unending loop of redirects is created.
This can never be intended for any meaningful website; a redirect loop is a problem that should be addressed.
Our HTTP client library happily follows redirects and has a redirect limit.
What we don't yet have in the testing system is a means of spotting that the HTTP client has reached its redirect limit and marking the test as failed as a result.
Production bug #002 spotted.