#573: DOM Review Draft — Published 15 June 2020
Discussions
2021-01-Kronos
Alice: Opening each of the changes they listed in order.
... [Define ParentNode's replaceChildren()](https://github.com/whatwg/dom/commit/3a11492f09212c944a1b495c47ce3c850136b296
Tess: This just seems like a win.
Alice: Agreed.
... Avoid enqueuing a tree mutation record when there is no mutation
Tess: This just sounds like a bugfix.
Alice: What's the assert for?
Tess: In case there are other callers to the "queue a tree mutation record" algorithm.
Alice: Makes sense.
... Actually looking at the bug, it's making the spec match implementations.
... Use self.origin instead of document.origin
Tess: It's redundant with self.origin.
Alice: What is self? Oh, it's synonymous with window... https://developer.mozilla.org/en-US/docs/Web/API/Window/self
Tess: Looking at the discussion on this bug, it seems good... removing something that's now only in one engine, and is redundant with something that's implemented everywhere.
... Replace child text content change steps
Alice: What is hr-time?
Tess: High-Resolution Time. It was removed because of Spectre... now being added back.
... They replaced "child text content changed steps" with "children changed steps". It's a spec organisation question. Bikeshed would catch any errors where something is referring to the old steps.
... Address infinite loop in TreeWalker's nextNode()
Tess: Sounds like a great bugfix!
... Change add/remove event listener behavior for service workers
Alice: ... ok
... Define the onabort event handler IDL attribute
Alice: Great.
Tess: Good bugfix.
... Add the DOM XPath interfaces
Tess: This is where browsers have a weird subset of an old version of XPath, and nobody wants to specify it.
... So now DOM is going to define it, and this is the beginning of that work.
Alice: Looks good.
... Add StaticRange constructor
Alice: Apparently this one was our fault. "Somehow StaticRange got added to browsers without the TAG noting the lack of a constructor. #589 adds StaticRange to the DOM Standard."
Tess: It's good that they're fixing it.
... The algorithm steps for the constructor don't do much validation, which is good because the code that handles them handles invalid ranges already.
... Catch errors while upgrading customized built-in elements
Tess: It's just adding a try/catch, more or less.
... Avoid enqueuing "disconnectedCallback" on detached custom elements
Tess: Another bugfix.
... Add delegatesFocus to attachShadow and flag to ShadowRoot
Alice: Doesn't define what it is, but it's defined in HTML.
Tess: I wish the DOM spec did a better job of explaining those external dependencies.
... this looks fine.
... Define the onslotchange event handler IDL attribute
Alice: Just adding the idl attribute.
... Run adopt as part of insert
Tess: Issue boils down to a concern about DocumentFragments, and that's what gets fixed in this PR.
OpenedNov 17, 2020
HIQaH! QaH! TAG!
I'm requesting a TAG review of DOM Review Draft — Published 15 June 2020.
The WHATWG published a new Review Draft of the DOM Standard on 15 June 2020. The HTML WG plans to take this Review Draft to W3C Recommendation in 2021, thus we are inviting another wide and horizontal review of the commits of the DOM Standard between 19 June 2019 (b5eac4d) and 15 June 2020 (ad06a4c).
Further details:
You should also know that...
We put together a list of significant changes for each horizontal review group. Please note that the classification is only an attempt to help quickly go through the changes, if possible, please look at all of the commits: https://github.com/w3c/htmlwg/issues/14#issuecomment-728034670
WebComponents
We'd prefer the TAG provide feedback as (please delete all but the desired option):
🐛 open issues in our GitHub repo for each point of feedback