Context Navigation

Changes between Version 1 and Version 2 of Ticket #59450, comment 3

v1	v2
1	1	The `1e4` iterations are not part of the benchmark but the warmup. I found that the website shows inconsistent numbers, probably due to the way `hrtime()` interacts with the server or how it lazy loads PCRE -- I cannot tell. If you flip the warmup tests, the second will always be slower during -- everything thereafter is indicative of real-world performance. I should exclude those from the output in the future to alleviate confusion about this quirk.
2	2
3		The point I'm making about not writing to variables is that the entire codebase is trying to write variables before reading them. This often happens needlessly, also in `apply_filters()`, albeit obscure: https://wordpress.slack.com/archives/C02RQBWTW/p1695251506451989. This is the aftermath of supporting PHP 5.2 for far too long, accumulating code debt that should be undone.
	3	The point I'm making about not writing to variables is that the entire codebase is trying to write variables before reading them. This often happens needlessly, also in `apply_filters()`, albeit obscure: https://wordpress.slack.com/archives/C02RQBWTW/p1695251506451989. This is the aftermath of supporting PHP 5.2 for far too long, accumulating code debt that should be undone if we wish to make WordPress load in under 0.01 seconds; I'm certain that number is attainable.