Add the navigation API

[domenic]

This imports much of the specification from https://github.com/WICG/navigation-api/blob/91c2e7f959418d71e26e5b9a6723fed2944001d9/spec.bs. It includes some fixes and renamings, as well as a lot of rearranging to flow better, but the feature set is the same.

In addition to the new "navigation API" section under "APIs related to navigation and session history", this makes the following changes to integrate the navigation API:

• Introduces the NavigationHistoryBehavior enum, as a superset of the "history handling" specification type. This includes an "auto" value to reflect that existing entry points will sometimes default to "replace" instead of "push". The navigation API allows overriding that auto behavior in some circumstances by explicitly supplying "push". (But in some cases, i.e., the initial about:blank or javascript: URLs, this is still not allowed.)

• Introduces the concept of "user navigation involvement". This will be helpful for solving issues such as Sec-Fetch-User and Sec-Fetch-Site for user-initiated UI navigations w3c/webappsec-fetch-metadata#71.

• Introduces the concept of "history-action user activation", which is a separate type of user activation meant to be used specifically to prevent back button tracking. It has no associated timeout (unlike transient activation), but can be flipped to false (unlike sticky activation). It is used by the navigation API to prevent calling navigateEvent.preventDefault() on two traversals in a row without intervening user activation. It is likely also to be useful for issues such as Specify back button history-entry skipping? #7832.

• The activation behavior for <a> and <area> elements is made more rigorous and consolidated into one place, so that we can hook into it appropriately to fire navigate events.

• Some surgery was needed on "apply the history step". The result is that it now has slightly more parameters, but also several wrapper algorithms which take care of setting up those parameters correctly for the cases of: navigable creation/destruction, push/replace, reload, and traverse. It also returns a value indicating if the application of the history step was canceled, and if so, how; this is used to give informative errors as return values of navigation.traverseTo().

• The "check if unloading is user-canceled" algorithm has become "check if unloading is canceled", as it now handles firing a possibly-cancelable navigate event at the top-level traversable.

Other changes scattered throughout are mostly integrating appropriate calls to the navigation API as necessary, especially to fire navigate events.

---

• At least two implementers are interested (and none opposed):
  • Chrome
  • Mozilla
• Tests are written and can be reviewed and commented upon at:
  • https://wpt.fyi/results/navigation-api?label=experimental&label=master&aligned (to be moved into html/browsers/ after this PR is merged)
• Implementation bugs are filed:
  • Chromium: https://bugs.chromium.org/p/chromium/issues/detail?id=1183545
  • Gecko: https://bugzilla.mozilla.org/show_bug.cgi?id=1777171
  • WebKit: https://bugs.webkit.org/show_bug.cgi?id=258384
  • Deno (only for timers, structured clone, base64 utils, channel messaging, module resolution, web workers, and web storage): N/A
  • Node.js (only for timers, structured clone, base64 utils, channel messaging, and module resolution): N/A
• MDN issue is filed: add docs for the Navigation API mdn/content#21960

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

---

/acknowledgements.html ( diff )
/browsing-the-web.html ( diff )
/document-lifecycle.html ( diff )
/document-sequences.html ( diff )
/form-control-infrastructure.html ( diff )
/form-elements.html ( diff )
/forms.html ( diff )
/iframe-embed-object.html ( diff )
/image-maps.html ( diff )
/index.html ( diff )
/indices.html ( diff )
/infrastructure.html ( diff )
/input.html ( diff )
/interaction.html ( diff )
/links.html ( diff )
/nav-history-apis.html ( diff )
/parsing.html ( diff )
/popover.html ( diff )
/semantics.html ( diff )
/server-sent-events.html ( diff )
/text-level-semantics.html ( diff )
/webappapis.html ( diff )

[tbondwilkinson]

Just want to leave a comment that the timing of dispose isn't important for most use cases I believe, and I think if it has the potential to cause confusion, dispose could be limited in scope to just same-document.

[tbondwilkinson]

Just nits really. This is a massive rewrite, really impressive seeing it all laid out!

[tbondwilkinson]

The other section titles in this spec page are written like "The History interface". Any reason this title deviates?

[domenic]

This section contains more than just the Navigation interface, as it's the section header for all of the navigation API. IN that way it's more like the section titles elsewhere of "Dynamic markup insertion", "User prompts", "Server-sent events", "Cross-document messaging", etc. The addition of "API" is a bit unusual, but it's necessary because of the generic term we've chosen: throughout the spec we need to distinguish "navigation" and "the navigation API", whereas we don't need to distinguish "WebSockets" and "the WebSockets API".

[tbondwilkinson]

I think this note could be just a little clearer.

"If upcoming non-traverse API method tracker is still apiMethodTracker after the the navigate algorithm, it means that the navigate algorithm did not successfully promote the upcoming non-traverse API method tracker to the ongoing method tracker."

I also wonder why the navigate algorithm does not return an error code which would make the algorithm clearer, I think?

[domenic]

Thanks; I tweaked the note similar to what you suggested.

As for why not modifying the navigate algorithm, it seems cleaner to keep the navigation API-specific stuff a bit separated, especially since it's important only for this one call site. Navigate is called by many different parts of the spec, and having a return value that 90% of call sites ignore seems a bit unfortunate.

[tbondwilkinson]

It's a bit odd that the navigationType is null for updateState(). Are we sure we don't want to throw in an extra navigation type for that?

[domenic]

I think since it's very much not a navigation, it's best to have it be null, to signal the exceptional nature of what's going on.

[smaug----]

This is surprising. Why would removing a node clear such flag.

[domenic]

So how focusReset: "after-transition" is supposed to work is as follows:

1. If nothing relevant changes in the document, focus gets reset after the transition
2. If the user (or developer, but the user is more important) changes focus during the transition, we don't reset the focus: that would be too disruptive.
3. But, if the user or developer changes focus during the transition, and then the developer removes the currently-focused DOM node from the document, we do reset the focus.

Without this step you're commenting on, (3) will be missing.

Note here "reset the focus" means the slightly-special behavior in https://whatpr.org/html/8502/nav-history-apis.html#potentially-reset-the-focus , of using the autofocus behavior and also resetting the sequential focus navigation starting point. So it's more than just the focus fixup steps.

What do you think?

[smaug----]

It is unclear why this is needed. (but I'm still reading all this, so perhaps something somewhere clarifies the issue this is trying to solve.)

[domenic]

This is added to prevent back-trapping. See "traversals have special restrictions on canceling the navigation" in https://github.com/wicg/navigation-api#restrictions-on-firing-canceling-and-responding . It is also envisioned to be used by https://github.com/WICG/close-watcher .

[smaug----]

This looks surprising. One couldn't use the API for top level opaque documents for example.

[domenic]

Yes. We haven't run into people using top-level opaque documents who complained, but it's something we could loosen up in the future if you are in contact with some developers who would like to use this API, and can't.

[smaug----]

Isn't this surpising? if the key is already there, reuse that even if the map might have gotten other entries after the key was added.

[domenic]

I don't understand your second sentence, or what's surprising about this behavior. Could you say more?

To be clear, this is about cases like:

const key = navigation.entries()[navigation.currentEntry.index - 1];

const result1 = navigation.traverseTo(key);

// ... probably in some other part of the app ...
const result2 = navigation.traverseTo(key);

and says that result2 should just reflect what's going on with result1 (i.e. have the same promises). I'm not sure what other behavior you'd envision here...

[zcorpan]

@smaug---- from Gecko and @annevk (at Gecko at the time; now at Apple) were involved in the design of this API, and we have some positive sentiments in mozilla/standards-positions#543. I'd love to have them confirm their support.

Mozilla is positive.

[domenic]

Thanks so much Simon!

Given that we've gotten review on this from multiple sources already, and all the checkboxes in the OP are filled out, I'll plan to merge this Monday (Japan time). Of course, if more reviews come in, I'm happy to delay.