Commits on Sep 19, 2018
  1. Revamp how fieldset and legend rendering is defined

    zcorpan committed Sep 19, 2018
    Properly define the rendering of the fieldset and legend elements.
    The layout model used is most similar to Gecko, which uses an anonymous box to hold the fieldset's contents.
    Fixes #3955, fixes #3930, fixes #3929, fixes #3928, fixes #3927, fixes #3915, fixes #3913, fixes #3660, fixes #3331, fixes #2756, fixes #4013.
Commits on Sep 6, 2018
  1. Tweak how accesskey on legend works

    zcorpan committed Sep 6, 2018
    Fixes #3950.
    Tests: web-platform-tests/wpt#12800
    * Don't special-case focusable legend, like Edge/Chrome/Safari
    * Fix grammar
    * Fix wording
    * Use lists to make it less ambiguous
    * and
    * Address domenic's comments
    * Add an example
    * Source formatting nits
Commits on Aug 24, 2018
Commits on Aug 21, 2018
Commits on Aug 17, 2018
  1. Fix attribute descriptions in the index for fieldset

    zcorpan authored and annevk committed Aug 17, 2018
    Fixes #3335.
Commits on Aug 16, 2018
  1. Change fieldset's default padding to match the majority

    zcorpan authored and annevk committed Aug 16, 2018
    WebKit/Chromium/EdgeHTML have 0.75em left/right padding and
    0.625em bottom padding. Gecko has it the other way around.
    Tests: web-platform-tests/wpt#12501.
    Fixes #3921.
Commits on Jun 11, 2018
Commits on Jun 5, 2018
  1. Change <td hidden> to render as display: none

    zcorpan committed Jun 5, 2018
    This aligns with implementations and avoids potentially breaking
    existing web content that expect display: none.
    Tests: web-platform-tests/wpt#11339
    Fixes #3728.
Commits on Feb 12, 2018
Commits on Nov 17, 2017
  1. Update how DOMMatrix is created from a dictionary

    zcorpan committed Nov 4, 2017
    This follows w3c/fxtf-drafts#218 and should
    have no normative difference.
Commits on Nov 16, 2017
Commits on Oct 20, 2017
  1. Meta: disambiguate focus() and blur() for Window and HTMLElement

    zcorpan authored and domenic committed Oct 20, 2017
    Add data-dfn-for="" attributes as appropriate. This should help other
    specs (in this case CSSOM View) cross-referencing these terms.
    Also fixes a few cases where data-for="" should have been
Commits on Sep 26, 2017
Commits on Aug 31, 2017
  1. Editorial: Remove unused reference [AUXCLICK]

    zcorpan committed Aug 31, 2017
    This should have been removed in cdc5273.
  2. Meta: make link-fixup.js only look at the path for 404s

    zcorpan committed Aug 31, 2017
    The Table of Contents link to index.html was redirected to
    indices.html#index, and similarly all multipage pages without a
    hash was being redirected (sometimes correctly, sometimes not).x
Commits on Aug 29, 2017
  1. Use DOMMatrix2DInit dictionary in 2d canvas setTransform()/addPath()

    zcorpan committed Aug 29, 2017
    Fixes #2826. The DOMMatrix2DInit dictionary is introduced in
    Tests: web-platform-tests/wpt#7007
Commits on Aug 22, 2017
  1. Make U+0000 a parse error in bogus comment and bogus DOCTYPE states

    zcorpan committed Aug 17, 2017
    Also make sure to reconsume instead of switching to the bogus
    DOCTYPE state so that a U+0000 that causes the parser to enter the
    bogus DOCTYPE state is itself caught as a parse error.
    Fixes #2684.
    Tests: html5lib/html5lib-tests#103
Commits on Aug 21, 2017
  1. Remove <applet>

    zcorpan authored and annevk committed Aug 21, 2017
    This removes support for the applet element from:
    * document.all[name]
    * document[name]
    * window[name]
    * document.applets now returns an empty array
    * Removed handling for applet elements in <iframe sandbox>
    * Removed rendering rules for applet elements
    * Removed the element itself and HTMLAppletElement
    * Applet elements now use HTMLUnknownElement
    The HTML parser is intentionally not changed.
    Tests: web-platform-tests/wpt#6684.
    Fixes #454.
Commits on Aug 19, 2017
  1. Meta: fix link-fixup to work when there's no fragment

    zcorpan authored and annevk committed Aug 19, 2017
    Also do nothing if the path doesn't start with /multipage/ or /dev/.
    Helps with #2934.
Commits on Jun 14, 2017
Commits on Jun 13, 2017
  1. Meta: Persist dfn panel across multipage navigations

    zcorpan committed Jun 13, 2017
    The intent here is that the dfn panel on multipage should act much
    like it does on single-page. Following links in the panel should
    let the panel stay open across page navigations. Closing the panel
    or clicking on a new dfn, and then traversing through history,
    should reflect the latest state.
    Also fix subtle bugs that sometimes caused more than one panel to
    be created, e.g., if a link in the panel is control-clicked.
Commits on Jun 8, 2017
  1. Meta: various fixes to html-dfn.js

    zcorpan authored and domenic committed Jun 8, 2017
    * Clicking anywhere but on a link caused cross-references to load.
      Now fixed to only load if a `dfn` is clicked.
    * Calling dfnLoad from dfnShow with a timeout is unnecessary with
      the new setup.
    * Clicking the permalink didn't close the panel.
    * Clicking a cross-spec definition (in the Dependencies section)
      didn't open the panel. It still doesn't show any references.
    * The click handlers were invoked in the wrong order; the onclick
      on the links in the panel should be invoked first so that
      event.stopPropagation() has an effect.
    * The permalink was gone for the "No references" case.
    * Made the permalink (and spec link, for cross-spec definitions)
      appear immediately without waiting for fetching.
    * Show an error message in the panel if the fetch fails.
    * Run the script async instead of defer to make the panel available
      as soon as possible.
Commits on Jun 7, 2017
Commits on Jun 2, 2017
  1. Make <hr> have 'overflow: hidden' by default

    zcorpan committed Jun 2, 2017
    This causes the element to create a block formatting context, which
    means it will avoid floats and grow to contain floats. This matches
    EdgeHTML. WebKit/Chromium/Gecko avoid floats but do not create a
    block formatting context, which is weird.
    Fixes #2715.
    Tests: web-platform-tests/wpt#6115
Commits on Jun 1, 2017
  1. Editorial: minor fixes to the parse error table (#2728)

    zcorpan authored and domenic committed Jun 1, 2017
    Element names in the domTree should be lowercase. Also add comma
    and space after "e.g.".
Commits on May 12, 2017
  1. Add Ivan Nikulin to acks

    zcorpan committed May 12, 2017
    For #2669 and earlier contributions.
Commits on Apr 20, 2017
  1. Meta: remove an unnecessary id with %

    zcorpan authored and annevk committed Apr 20, 2017
    This id was introduced in b29fb58.
    It repeats the id of the heading above but with `(` and `)`
    percent-escaped. This is not done for any other ids that also contain
    those characters.
Commits on Apr 19, 2017
Commits on Apr 17, 2017
  1. Make <body topmargin> win over <iframe marginheight> (#2489)

    zcorpan authored and domenic committed Apr 17, 2017
    This matches Gecko and Edge. Chromium and WebKit copy the
    marginheight attribute from <iframe> to the child document's
    <body>, which then takes precedence over topmargin, but copying
    the attribute is non-conforming.
    Fixes #2484.
Commits on Apr 6, 2017
  1. Specify the tokenizer for's features argument

    zcorpan committed Apr 6, 2017
    This was specified in CSSOM View but the "noopener" feature did
    not use the same tokenizer as the legacy features. Fixes #2474.
    Also specify the aliases screenx, screeny, innerwidth, innerheight
    for left, top, width, and height, respectively. Part of #2464.
    Closes w3c/csswg-drafts#1128.
    The tokenizer specified here closely follows Edge.
    Chromium and WebKit are also very similar to Edge.
    Difference from Edge: U+0000 does not end the string.
    Difference from Chromium: U+0000 is not a separator.
    Difference from WebKit/Chromium/Edge: U+000C is a separator.
    For the input `width toolbar=450, height=450`,
    Edge tokenizes like `width, toolbar=450, height=450` while
    WebKit/Chromium like `width=450, height=450`.
    The Edge behavior seems better.
  2. Clarify what should actually happen to negative timeouts

    zcorpan authored and domenic committed Apr 6, 2017
    Although d5f415e tried to define
    the behavior for negative timeouts, it didn't match implementations
    and was reverted in 3716990.
    Instead a negative timeout should just be set to 0.
    Tests: web-platform-tests/wpt#5380
Commits on Mar 24, 2017
  1. Remove default style for <dialog> for small screens

    zcorpan authored and domenic committed Mar 23, 2017
    Having a breakpoint in the user agent style sheet is unprecedented,
    unexpected, and more work to maintain and test. It has also not
    been implemented in Chromium.
    Fixes #624.
  2. Make <dialog> have explicit display: block by default

    zcorpan authored and domenic committed Mar 24, 2017
    This results in saner behavior should the author stylesheet
    override 'position' or set the offset properties to `auto`.
    Fixes #2405.