traverse_the_history_1.html and traverse_the_history_write_after_load_1.html can never complete: start_test_wait() is never called #63098

Both tests can only finish from start_test_wait(), a function each file defines, but nothing in the automated test ever calls it:

• traverse_the_history_1.html and traverse_the_history_write_after_load_1.html open history_entry.html, which never calls opener.start_test_wait().
• The only caller is the manual-test copy, non-automated/history_entry.html.

So both files time out in every engine regardless of behaviour. wpt.fyi on 2026-09-30 (Chrome 154, Firefox 157, Safari TP 253) shows TIMEOUT 0/1 for both in all three. The siblings traverse_the_history_2.html through _5.html poll pages with setInterval instead of waiting for start_test_wait(); they complete, and pass in Chrome and Firefox.

The files date from dfe897f (2013). #11296 (part of #11269) moved traverse_the_history_1.html into a popup and fixed its URL, but the path to t.done() stayed unreachable.

traverse_the_history_write_after_load_1.html has a second problem. Its expected pages of [2, 3, 1] assumes that document.write() after the load event, an implicit document.open(), creates a new session history entry. Since whatwg/html#3946 (part of the document.open() overhaul, whatwg/html#3818), document.open() never adds a history entry. So even with a working done path, the expectation contradicts the current spec. traverse_the_history_write_after_load_2.html already carries the comment "the pass condition here is based on the idea that the spec is wrong and browsers are right".

Suggested fix:

• Make both tests poll pages the way traverse_the_history_2.html does, or have history_entry.html call opener.start_test_wait() when it exists.
• Update traverse_the_history_write_after_load_1.html's expectation to the post-whatwg/html#3946 model, or remove the test if _write_after_load_2.html already covers it.

Found while triaging harness timeouts for a new engine; we can't pass these as written.