On <!DOCTYPE html> versus <!doctype html>: it’s actually not so simple. If you’re using gzip, you might manage to save as much as a few bytes (though typically

On <!DOCTYPE html> versus <!doctype html>: it’s actually not so simple. If you’re using gzip, you might manage to save as much as a few bytes (though typically only 1–2), but if you’re using Brotli, it’ll actually compress something like 2–4 bytes worse, because Brotli includes a static dictionary based on a popularity contest of largely-not-optimised content. This is a thing I dislike about Brotli: it encourages you to follow the herd, rather than doing what would otherwise be more optimal. (Sure, the encouragement is just a byte here, a byte there, but it’s the principle of the thing.) Here are sizes based on running `‹brotli or gzip› --stdout <<<'‹doctype›‹charset›' | wc -c`: ┌─────────────────┬────────────────────────┬──────────────────────┐ │ BROTLI │ <meta charset="utf-8"> │ <meta charset=utf-8> │ ├─────────────────┼────────────────────────┼──────────────────────┤ │ <!DOCTYPE html> │ 19 │ 27 │ │ <!doctype html> │ 23 │ 29 │ └─────────────────┴────────────────────────┴──────────────────────┘ ┌─────────────────┬────────────────────────┬──────────────────────┐ │ GZIP │ <meta charset="utf-8"> │ <meta charset=utf-8> │ ├─────────────────┼────────────────────────┼──────────────────────┤ │ <!DOCTYPE html> │ 58 │ 56 │ │ <!doctype html> │ 58 │ 56 │ └─────────────────┴────────────────────────┴──────────────────────┘ In practice, when you add more content, predominantly lowercase, doctype tends to save a byte or two compared to DOCTYPE. So gzip gravitates to the bottom left corner—that removing unnecessary characters and matching case where case-insensitive is good. But Brotli was made by popularity contest, and so the inferior representation actually ends up compressing better than the superior, because most people did things the inferior way!