Align spec more closely with cache freshening

[gterzian]

FIX #950

• At least two implementers are interested (and none opposed):
  • …
  • …
• Tests are written and can be reviewed and commented upon at:
  • …
• Implementation bugs are filed:
  • Chrome: …
  • Firefox: …
  • Safari: …

(See WHATWG Working Mode: Changes for more details.)

---

Preview | Diff

[gterzian]

@mnot I'm wondering if the test added in servo/servo#26999 would be testing for the correct behavior?

I'm actually not sure if the infra for http-cache test would be happy with that second fetch in case of a failing validation, might require some changes there too, but I'll first try to actually implement the logic to see what happens.

[gterzian]

@annevk I think this is probably worth some implementation bugs, just to check that this behavior is indeed what others are doing?

I'm not sure about "implementer interest", I think this simply aligns fetch with the http-caching spec, and I wouldn't be surprised if this already matches what implementations are doing.

[annevk]

1. It seems this allows for reducing the effective scope of storedResponse. We should probably make the text reflect that.
2. And yeah, I think we should have some tests here to be sure. I also would not expect implementations to do a lot of this properly. E.g., combining of headers is one of the areas with known bugs (and perhaps some lack of clarity in the protocol as to which headers can be updated and such; see Updating stored headers httpwg/http-core#165).