tree 0ef9ee7a611dc94df60fc070c42d63a636ca0c61
parent 86a6f71391e48581aec60e6b9e25f9a80dc86a75
author Martin Thomson <martin.thomson@gmail.com> 1649031349 +1000
committer GitHub <noreply@github.com> 1649031349 +1000
gpgsig -----BEGIN PGP SIGNATURE-----
 
 wsBcBAABCAAQBQJiSji1CRBK7hj4Ov3rIwAAIvAIALI4MR1CLiqBHkt0WPKd7WDi
 IL9lYxFjoI9Fag/lYsrvekI21IkBKoWUB/HSt9BaM6iE4IJJoC1p7wX/yibwV6dR
 YS4tHIfg7PNfBFiKj0kRijFgFxaSlSDKk0cLbZVqpdg1SQtc2uxvl0125QY5513I
 Hom3c3/kfmkT5hZZP2oOdPAM56/RFTNfelWCTaQ+Y3AGdtJq+phd/vLQIILPVVX1
 WC4uUKkhDp7Clo2+6z71fD4uzKd4+9Cl8Xr35/mb2wkvdGZ9tkoTXILwKaqyFiHI
 IW09SG1OyFGnnbuCqOZ1v4f0mhISDIJivQPS9mKQgn43NGTOVoS3eW2feclTz9U=
 =lcST
 -----END PGP SIGNATURE-----
 

Update test.html

Fix test 64 for running under https and redirect to HTTPS for this test.

There's an alternative here that might be simpler: change test 64 to look for window.location.protocol using indexOf() === 0, but I think that this version will produce a better outcome for all involved.

I haven't investigated it, but there is an outside chance that HSTS on the server is cleaner.  Given the amount of stuff that server serves, I doubt that that change would be trivial though.