url hashes count as cache misses

[whmountains]

Like the title says. example.com works, but example.com#chapter2 counts as a cache miss and the page fails to load.

Is this a bug, or the intended behavior. If it's intended, can we add a configuration option to strip hashes? It could be done inside the stripIgnoredUrlParameters function.

[whmountains]

Update: I tested in Chrome and it works fine with hashes, so the behavior seems to be Firefox only.

I did some debugging and I think I've found the root of the problem. event.request.url does not include the hash segment in Chrome but in Firefox it does, hence the cache misses.

This may be a bug in Firefox, Chrome, or just an ambiguity in the spec, but in any case it would be an easy fix to work around it in the service worker code.

[jeffposnick]

Thanks for pointing this out—it's due to w3c/ServiceWorker#854.

That new behavior shipped in Firefox 52, and will ship in the upcoming Chrome 59 beta as well. I'll take care of adjusting the sw-precache logic to account for it.

[whmountains]

Thank you!

Will you publish a bugfix version immediately, or will I have to install directly from git until the next release comes out?

[jeffposnick]

I don't anticipate much of a delay in releasing this as 5.1.1 once the PR is merged.