CSS Basic User Interface Module Level 4

1. Introduction

This module describes CSS properties which enable authors to style user interface related properties and values.

Section 2.1 of CSS1 [CSS1] and Chapter 18 of CSS2 [CSS21] introduced several user interface related properties and values. User Interface for CSS3 (16 February 2000) introduced several new user interface related features.

[CSS-UI-3] was later introduced to incorporates, extends, and supersedes these. This specification continues this work, and in turn replaces [CSS-UI-3].

1.1. Purpose

The purpose of this specification is to achieve the following objectives:

• Extend the user interface features in [CSS21] and [CSS-UI-3]
• Provide additional CSS mechanisms to augment or replace other dynamic presentation related features in HTML.
• Introduce directional navigation properties to assist in the construction of user interfaces which make use of a directional navigation model.

2. Module Interactions

This document defines new features not present in earlier specifications. In addition, it replaces and supersedes [CSS-UI-3], which itself replaced and superseded the following:

• Section 18.1, section 18.4, and Information on the stacking of outlines defined in Appendix E of Cascading Style Sheets, level 2, revision 1 [CSS21]
• User Interface for CSS3 (16 February 2000) [CSS-UI-3]

3. Box Model addition

3.1. Changing the Box Model: the box-sizing property

content-box

This is the behavior of width and height as specified by CSS2.1. The specified width and height (and respective min/max properties) apply to the width and height respectively of the content box of the element. The padding and border of the element are laid out and drawn outside the specified width and height.

border-box

Length and percentages values for width and height (and respective min/max properties) on this element determine the border box of the element. That is, any padding or border specified on the element is laid out and drawn inside this specified width and height. The content width and height are calculated by subtracting the border and padding widths of the respective sides from the specified width and height properties. As the content width and height cannot be negative ([CSS21], section 10.2), this computation is floored at 0. Used values, as exposed for instance through getComputedStyle(), also refer to the border box.

Note: This is the behavior of width and height as commonly implemented by legacy HTML user agents for replaced elements and input elements.

Note: In contrast to the length and percentage values, the auto value of the width and height properties (as well as other keyword values introduced by later specifications, unless otherwise specified) is not influenced by the box-sizing property, and always sets the size of the content box.

The following terms, whose definitions vary based on the computed value of box-sizing are introduced:

The Visual formatting model details of [CSS21] are written assuming box-sizing: content-box. The following disambiguations are made to clarify the behavior for all values of box-sizing:

1. In 10.3.3, the second width in the following phrase is to be interpreted as content width: If width is not auto and border-left-width + padding-left + width + [...]
2. In 10.3.7, width is to be interpreted as content width in the following equation: left + margin-left + border-left-width + padding-left + width + [...]
3. In 10.4, width, height, min-width, max-width, min-height and max-height are respectively to be interpreted as content width, content height, min inner width, max inner width, min inner height and max inner height in the following phrases:
   1. The tentative used width is calculated [...]
   2. If the tentative used width is greater than max-width, the rules above are applied again, but this time using the computed value of max-width as the computed value for width.
   3. If the resulting width is smaller than min-width, the rules above are applied again, but this time using the value of min-width as the computed value for width.
   4. Select from the table the resolved height and width values for the appropriate constraint violation. Take the max-width and max-height as max(min, max) so that min ≤ max holds true. In this table w and h stand for the results of the width and height computations [...]
   5. All instances of these words in the table
   6. Then apply the rules under "Calculating widths and margins" above, as if width were computed as this value.
4. In 10.6.4, height is to be interpreted as content height in the following equation: top + margin-top + border-top-width + padding-top + height + [...]
5. In 10.7, width, height, min-height and max-height are respectively to be interpreted as content width, content height, min inner height and max inner height in the following phrases:
   1. The tentative used height is calculated [...]
   2. If this tentative height is greater than max-height, the rules above are applied again, but this time using the value of max-height as the computed value for height.
   3. If the resulting height is smaller than min-height, the rules above are applied again, but this time using the value of min-height as the computed value for height.
   4. [...] use the algorithm under Minimum and maximum widths above to find the used width and height. Then apply the rules under "Computing heights and margins" above, using the resulting width and height as if they were the computed values.

div.container {
  width:38em;
  border:1em solid black;
}

div.split {
  box-sizing:border-box;
  width:50%;
  border:1em silver ridge;
  float:left;
}

<div class="container">
<div class="split">This div occupies the left half.</div>
<div class="split">This div occupies the right half.</div>
</div>

4. Outline properties

At times, style sheet authors may want to create outlines around visual objects such as buttons, active form fields, image maps, etc., to make them stand out. Outlines differ from borders in the following ways:

1. Outlines do not take up space.
2. Outlines may be non-rectangular.
3. UAs often render outlines on elements in the :focus state.

The outline properties control the style of these dynamic outlines.

The stacking of the rendering of these outlines is explicitly left up to implementations to provide a better user experience per platform. This supersedes the stacking of outlines as defined in Appendix E of CSS 2.1 [CSS21].

Keyboard users, in particular people with disabilities who may not be able to interact with the page in any other fashion, depend on the outline being visible on elements in the :focus state, thus authors must not make the outline invisible on such elements without making sure an alternative highlighting mechanism is provided.

The rendering of applying transforms to outlines is left explicitly undefined in CSS3-UI.

4.1. Outlines Shorthand: the outline property

4.2. Outline Thickness: the outline-width property

4.3. Outline Patterns: the outline-style property

4.4. Outline Colors: the outline-color property

The outline created with the outline properties is drawn "over" a box, i.e., the outline is always on top, and doesn’t influence the position or size of the box, or of any other boxes. Therefore, displaying or suppressing outlines does not cause reflow.

Outlines may be non-rectangular. For example, if the element is broken across several lines, the outline should be an outline or minimum set of outlines that encloses all the element’s boxes.

Each part of the outline should be fully connected rather than open on some sides (as borders on inline elements are when lines are broken).

The parts of the outline are not required to be rectangular. To the extent that the outline follows the border edge, it should follow the border-radius curve.

The position of the outline may be affected by descendant boxes.

User agents should use an algorithm for determining the outline that encloses a region appropriate for conveying the concept of focus to the user.

Note: This specification does not define the exact position or shape of the outline, but it is typically drawn immediately outside the border box.

The outline-width property accepts the same values as border-width (CSS Backgrounds 3 §4.3 Line Thickness: the border-width properties).

The outline-style property accepts the same values as border-style (CSS Backgrounds 3 §4.2 Line Patterns: the border-style properties), except that hidden is not a legal outline style. In addition, in CSS3, outline-style accepts the value auto. The auto value permits the user agent to render a custom outline style, typically a style which is either a user interface default for the platform, or perhaps a style that is richer than can be described in detail in CSS, e.g. a rounded edge outline with semi-translucent outer pixels that appears to glow. As such, this specification does not define how the outline-color is incorporated or used (if at all) when rendering auto style outlines. User agents may treat auto as solid.

The outline-color property accepts all colors, as well as the keyword invert. invert is expected to perform a color inversion on the pixels on the screen. This is a common trick to ensure the focus border is visible, regardless of color background.

Conformant UAs may ignore the invert value on platforms that do not support color inversion of the pixels on the screen.

If the UA does not support the invert value then it must reject that value at parse-time, and the initial value of the outline-color property is the currentColor keyword.

The outline property is a shorthand property, and sets all three of outline-style, outline-width, and outline-color.

Note: The outline is the same on all sides. In contrast to borders, there are no outline-top or outline-left etc. properties.

This specification does not define how multiple overlapping outlines are drawn, or how outlines are drawn for boxes that are partially obscured behind other elements.

button { outline: thick solid }

Graphical user interfaces may use outlines around elements to tell the user which element on the page has the focus. These outlines are in addition to any borders, and switching outlines on and off should not cause the document to reflow. The focus is the subject of user interaction in a document (e.g. for entering text or selecting a button).

:focus  { outline: thick solid black }
:active { outline: thick solid red }

Note: Since the outline does not affect formatting (i.e., no space is left for it in the box model), it may well overlap other elements on the page.

4.5. Offsetting the Outline: the outline-offset property

By default, the outline is drawn starting just outside the border edge. However, it is possible to offset the outline and draw it beyond the border edge.

If the computed value of outline-offset is anything other than 0, then the outline is outset from the border edge by that amount.

:focus,:active  { outline-offset: 2px }

Negative values must cause the outline to shrink into the border box. Both the height and the width of outside of the shape drawn by the outline should not become smaller than twice the computed value of the outline-width property, to make sure that an outline can be rendered even with large negative values. User Agents should apply this constraint independently in each dimension. If the outline is drawn as multiple disconnected shapes, this constraint applies to each shape separately.

5. Resizing & Overflow

CSS2.1 provides a mechanism for controlling the appearance of a scrolling mechanism (e.g. scrollbars) on block container elements. This specification adds to that a mechanism for controlling user resizability of elements as well as the ability to specify text overflow behavior.

5.1. Resizing Boxes: the resize property

The resize property allows the author to specify whether or not an element is resizable by the user, and if so, along which axis/axes.

none

The UA does not present a resizing mechanism on the element, and the user is given no direct manipulation mechanism to resize the element.

both

The UA presents a bidirectional resizing mechanism to allow the user to adjust both the height and the width of the element.

horizontal

The UA presents a unidirectional horizontal resizing mechanism to allow the user to adjust only the width of the element.

vertical

The UA presents a unidirectional vertical resizing mechanism to allow the user to adjust only the height of the element.

Currently it is possible to control the appearance of the scrolling mechanism (if any) on an element using the overflow property (e.g. overflow: scroll vs. overflow: hidden etc.). The purpose of the resize property is to allow control over the appearance and function of the resizing mechanism (e.g. a resize box or widget) on the element.

Note: The resizing mechanism is NOT the same as the scrolling mechanism, nor is it related to any UA mechanism for zooming. The scrolling mechanism allows the user to determine which portion of the contents of an element is shown. The resizing mechanism allows the user to determine the size of the element.

The resize property applies to elements whose computed overflow value is something other than visible. UAs may also apply it, regardless of the value of the overflow property, to:

• Replaced elements representing images or videos, such as img, video, picture, svg, object, or canvas.
• The iframe element.

The effect of the resize property on generated content is undefined. Implementations should not apply the resize property to generated content.

Note: the resize property may apply to generated content in the future if there is implementation of Interface CSSPseudoElement.

When an element is resized by the user, the user agent sets the width and height properties to px unit length values of the size indicated by the user, in the element’s style attribute DOM, replacing existing property declaration(s), if any, without !important, if any.

If an element is resized in only one dimension, only the corresponding property is set, not both.

The precise direction of resizing (i.e. altering the top left of the element or altering the bottom right) may depend on a number of CSS layout factors including whether the element is absolutely positioned, whether it is positioned using the right and bottom properties, whether the language of the element is right-to-left etc. The UA should consider the direction of resizing (as determined by CSS layout), as well as platform conventions and constraints when deciding how to convey the resizing mechanism to the user.

The user agent must allow the user to resize the element with no other constraints than what is imposed by min-width, max-width, min-height, and max-height.

Note: There may be situations where user attempts to resize an element appear to be overriden or ignored, e.g. because of !important cascading declarations that supersede that element’s style attribute width and height properties in the DOM.

Changes to the computed value of an element’s resize property do not reset changes to the style attribute made due to user resizing of that element.

iframe,object[type^="text/"],
object[type$="+xml"],object[type="application/xml"]
{
  overflow:auto;
  resize:both;
}

5.2. Overflow Ellipsis: the text-overflow property

This property specifies rendering when inline content overflows its line box edge in the inline progression direction of its block container element ("the block") that has overflow other than visible.

Text can overflow for example when it is prevented from wrapping (e.g. due to white-space: nowrap or a single word is too long to fit). Values have the following meanings:

clip

Clip inline content that overflows its block container element. Characters may be only partially rendered.

ellipsis

Render an ellipsis character (U+2026) to represent clipped inline content. Implementations may substitute a more language, script, or writing-mode appropriate ellipsis character, or three dots "..." if the ellipsis character is unavailable.

<string>

Render the given string to represent clipped inline content. The given string is treated as an independent paragraph for bidi purposes.

fade( <length> | <percentage> )

Clip inline content that overflows its line box. Characters may be only partially rendered. In addition, the UA must apply a fade out effect near the edge of the line box, reaching complete transparency at the edge.

Do we need to define the way the fade out is calculated so that the fading is identical across browsers? It should probably be something like mask-image: linear-gradient(to right, rgba(0,0,0,1), rgba(0,0,0,0)), except applied to the relevant portion of the line only.

The argument determines the distance over which the fade effect is applied. The <percentage> is resolved against the width of the line box. Values lower than 0 are clipped to 0. Values greater than the width of the line box are clipped to the width of the line box.

If the line box is too short to display the fade effect at the desired length, should we drop the effect, or shrink the distance it is applied over until it fits, or clip the end of the fade?

How should we deal with things overflowing out of the line box, or overlapping onto it? Should fade apply to the logical content of the line, or to the physical area of the line box, or the intersection of both?

fade

Same as fade(), but the distance over which the fading effect is applied is determined by the UA. 1em is suggested as a reasonable value.

The term "character" is used in this property definition for better readability and means "grapheme cluster" [UAX29] for implementation purposes.

If there is one value, it applies only to the end line box edge. If there are two values, the first value applies to the line-left edge, and the second value applies to the line-right edge. The terms end, line-left and line-right are defined in [CSS-WRITING-MODES-3].

Note: the use of line-left and line-right rather than start and end when there are two values is intentional, to facilitate the use of directional characters such as arrows.

For the ellipsis and string values, implementations must hide characters and atomic inline-level elements at the applicable edge(s) of the line as necessary to fit the ellipsis/string, and place the ellipsis/string immediately adjacent to the applicable edge(s) of the remaining inline content. The first character or atomic inline-level element on a line must be clipped rather than ellipsed.

ellipsing details

• Ellipsing only affects rendering and must not affect layout nor dispatching of pointer events: The UA should dispatch any pointer event on the ellipsis to the elided element, as if text-overflow had been none.
• The ellipsis is styled and baseline-aligned according to the block.
• Ellipsing occurs after relative positioning and other graphical transformations.
• If there is insufficient space for the ellipsis, then clip the rendering of the ellipsis itself (on the same side that neutral characters on the line would have otherwise been clipped with the text-overflow:clip value).

user interaction with ellipsis

• When the user is interacting with content (e.g. editing, selecting, scrolling), the user agent may treat ellipsis, string values, fade or fade() as text-overflow:clip.
• Selecting the ellipsis should select the ellipsed text. If all of the ellipsed text is selected, UAs should show selection of the ellipsis. Behavior of partially-selected ellipsed text is up to the UA.

div {
  font-family: Helvetica, sans-serif;
  line-height: 1.1;
  width: 3.1em;
  border: solid .1em black;
  padding: 0.2em; margin:1em 0;
}

<div>
CSS IS AWESOME, YES
</div>

<div style="text-overflow:clip; overflow:hidden">
CSS IS AWESOME, YES
</div>

<div style="text-overflow:ellipsis; overflow:hidden">
CSS IS AWESOME, YES
</div>

<div style="text-overflow:ellipsis; overflow:hidden">
NESTED
 <p>PARAGRAPH</p>
WON’T ELLIPSE.
</div>

<div style="text-overflow:fade; overflow:hidden">
CSS IS AWESOME, YES
</div>

Note: the side of the line that the ellipsis is placed depends on the direction of the block. E.g. an overflow hidden right-to-left (direction: rtl) block clips inline content on the left side, thus would place a text-overflow ellipsis on the left to represent that clipped content.

insert RTL example diagram here to illustrate note.

ellipsis interaction with scrolling interfaces

This section applies to elements with text-overflow other than text-overflow:clip (non-clip text-overflow) and overflow:scroll.

When an element with non-clip text-overflow has overflow of scroll in the inline progression dimension of the text, and the browser provides a mechanism for scrolling (e.g. a scrollbar on the element, or a touch interface to swipe-scroll, etc.), there are additional implementation details that provide a better user experience:

When an element is scrolled (e.g. by the user, DOM manipulation), more of the element’s content is shown. The value of text-overflow should not affect whether more of the element’s content is shown or not. If a non-clip text-overflow is set, then as more content is scrolled into view, implementations should show whatever additional content fits, only truncating content which would otherwise be clipped (or is necessary to make room for the ellipsis/string), until the element is scrolled far enough to display the edge of the content at which point that content should be displayed rather than an ellipsis/string.

div.crawlbar {
  text-overflow: ellipsis;
  height: 2em;
  overflow: scroll;
  white-space: nowrap;
  width: 15em;
  border:1em solid black;
}

<div class="crawlbar">
CSS is awesome, especially when you can scroll
to see extra text instead of just
having it overlap other text by default.
</div>

As some content is scrolled into view, it is likely that other content may scroll out of view on the other side. If that content’s block container element is the same that’s doing the scrolling, and the computed value of text-overflow has two values, with the value applying to the start edge being a non-clip value, then implementations must render an ellipsis/string in place of the clipped content, with the same details as described in the value definition above, except that the ellipsis/string is drawn in the start (rather than end) of the block’s direction (per the direction property).

While the content is being scrolled, implementations may adjust their rendering of ellipses/strings (e.g. align to the box edges rather than line edges).

If there is insufficient space for both start and end ellipses/strings, then only the end ellipsis/string should be rendered.

6. Pointing Devices and Keyboards

6.1. Pointer interaction

6.1.1. Styling the Cursor: the cursor property

This property specifies the type of cursor to be displayed for the pointing device when the cursor’s hotspot is within the element’s border edge.

Note: As per CSS Backgrounds 3 §5.1 Curve Radii: the border-radius properties, the border edge is affected by border-radius.

In the case of overlapping elements, which element determines the type of cursor is based on hit testing: the element determining the cursor is the one that would receive a click initiated from this position.

Note: The specifics of hit testing are out of scope of this specification. Hit testing will hopefully be defined in a future revision of CSS or HTML.

User agents may ignore the cursor property over native user-agent controls such as scrollbars, resizers, or other native UI widgets e.g. those that may be used inside some user agent specific implementations of form elements. User agents may also ignore the cursor property and display a cursor of their choice to indicate various states of the UA’s user interface, such as a busy cursor when the page is not responding, or a text cursor when the user is performing text selection.

Note: [HTML] defines special handling of image maps for the cursor property.

Values have the following meanings:

image cursors

<url>

The user agent retrieves the cursor from the resource designated by the URI. If the user agent cannot handle the first cursor of a list of cursors, it must attempt to handle the second, etc. If the user agent cannot handle any user-defined cursor, it must use the cursor keyword at the end of the list. Conforming User Agents may, instead of <url>, support <image> which is a superset.

The UA must support the following image file formats:

• PNG, as defined in [PNG]
• SVG, as defined in [SVG11], in secure static mode [SVG2], if it has an intrinsic size.
• any other non-animated image file format that they support for <image> in other properties, such as the the background-image property

In addition, the UA should support the following image file formats:

• SVG, as defined in [SVG11], in secure animated mode [SVG2], if it has an intrinsic size.
• any other animated image file format that they support for <image> in other properties, such as the the background-image property

The UA may also support additional file formats, including SVG, as defined in [SVG11], in secure static mode or secure animated mode [SVG2], even if it does not have an intrinsic size.

Note: The CSS Working group initially indented support for all SVG, intrinsically sized or not. Support for non intrinsically sized SVG was downgraded from mandatory to optional due to lack of implementations.

Note: At the time of writing this specification (spring 2015), the only file formats supported for cursors in common desktop browsers are the .ico and .cur file formats, as designed by Microsoft. For compatibility with legacy content, UAs are encouraged to support these, even though the lack of an open specification makes it impossible to have a normative requirement about these formats. Some information on these formats can be found on Wikipedia.

The default object size for cursor images is a UA-defined size that should be based on the size of a typical cursor on the UA’s operating system.

The concrete object size is determined using the default sizing algorithm. If an operating system is incapable of rendering a cursor above a given size, cursors larger than that size must be shrunk to within the OS-supported size bounds, while maintaining the cursor image’s intrinsic ratio, if any.

The optional <x> and <y> coordinates identify the exact position within the image which is the pointer position (i.e., the hotspot).

<x>

<y>

Each is a <number>. The x-coordinate and y-coordinate of the position in the cursor’s coordinate system (left/top relative) which represents the precise position that is being pointed to.

Note: This specification does not define how the coordinate systems of the various types of <image> are established, and defers these definitions to [CSS4-IMAGES].

If the values are unspecified, then the intrinsic hotspot defined inside the image resource itself is used. If both the values are unspecific and the referenced cursor has no defined hotspot, the effect is as if a value of "0 0" were specified.

If the coordinates of the hotspot, as specified either inside the image resource or by <x> and <y> values, fall outside of the cursor image, they must be clamped (independently) to fit.

general purpose cursors

auto

The UA determines the cursor to display based on the current context. Specifically, auto behaves as text over selectable text or editable elements, and default otherwise.

default

The platform-dependent default cursor. Often rendered as an arrow.

none

No cursor is rendered for the element.

links and status cursors

context-menu

A context menu is available for the object under the cursor. Often rendered as an arrow with a small menu-like graphic next to it.

help

Help is available for the object under the cursor. Often rendered as a question mark or a balloon.

pointer

The cursor is a pointer that indicates a link.

progress

A progress indicator. The program is performing some processing, but is different from wait in that the user may still interact with the program. Often rendered as a spinning beach ball, or an arrow with a watch or hourglass.

wait

Indicates that the program is busy and the user should wait. Often rendered as a watch or hourglass.

selection cursors

cell

Indicates that a cell or set of cells may be selected. Often rendered as a thick plus-sign with a dot in the middle.

crosshair

A simple crosshair (e.g., short line segments resembling a "+" sign). Often used to indicate a two dimensional bitmap selection mode.

text

Indicates text that may be selected. Often rendered as a vertical I-beam. User agents may automatically display a horizontal I-beam/cursor (e.g. same as the vertical-text keyword) for vertical text, or for that matter, any angle of I-beam/cursor for text that is rendered at any particular angle.

vertical-text

Indicates vertical-text that may be selected. Often rendered as a horizontal I-beam.

drag and drop cursors

alias

Indicates an alias of/shortcut to something is to be created. Often rendered as an arrow with a small curved arrow next to it.

copy

Indicates something is to be copied. Often rendered as an arrow with a small plus sign next to it.

move

Indicates something is to be moved.

no-drop

Indicates that the dragged item cannot be dropped at the current cursor location. Often rendered as a hand or pointer with a small circle with a line through it.

not-allowed

Indicates that the requested action will not be carried out. Often rendered as a circle with a line through it.

grab

Indicates that something can be grabbed (dragged to be moved). Often rendered as the backside of an open hand.

grabbing

Indicates that something is being grabbed (dragged to be moved). Often rendered as the backside of a hand with fingers closed mostly out of view.

resizing and scrolling cursors

e-resize, n-resize, ne-resize, nw-resize, s-resize, se-resize, sw-resize, w-resize

Indicates that some edge is to be moved. For example, the se-resize cursor is used when the movement starts from the south-east corner of the box.

ew-resize, ns-resize, nesw-resize, nwse-resize

Indicates a bidirectional resize cursor.

col-resize

Indicates that the item/column can be resized horizontally. Often rendered as arrows pointing left and right with a vertical bar separating them.

row-resize

Indicates that the item/row can be resized vertically. Often rendered as arrows pointing up and down with a horizontal bar separating them.

all-scroll

Indicates that the something can be scrolled in any direction. Often rendered as arrows pointing up, down, left, and right with a dot in the middle.

zooming cursors

zoom-in, zoom-out

Indicates that something can be zoomed (magnified) in or out, and often rendered as a magnifying glass with a "+" or "-" in the center of the glass, for zoom-in and zoom-out respectively.

:link,:visited {
    cursor: url(example.svg#linkcursor),
      url(hyper.cur),
      url(hyper.png) 2 3,
      pointer
}

6.1.1.1. Cursor of the canvas

The document canvas is the infinite surface over which the document is rendered [CSS21]. Since no element corresponds to the canvas, in order to allow styling of the cursor when not over any element, the canvas cursor re-uses the root element’s cursor. However, if no boxes are generated for the root element (for example, if the root element has display: none), then the canvas cursor is the platform-dependent default cursor.

Note: An element might be invisible, but still generate boxes. For example, if the element has visibility: hidden but not display: none, boxes are generated for it and its cursor is used for the canvas.

6.2. Insertion caret

6.2.1. Coloring the Insertion Caret: the caret-color property

auto

User agents should use currentColor. User agents may automatically adjust the color of caret to ensure good visibility and contrast with the surrounding content, possibly based on the currentColor, background, shadows, etc.

Note: When caret-shape is block, ensuring good visibility and contrast is best achieved with a UA determined color other than currentColor.

<color>

The insertion caret is colored with the specified color.

The caret is a visible indicator of the insertion point in an element where text (and potentially other content) is inserted by the user. This property controls the color of that visible indicator.

Note: UAs might have additional things that count as “carets”. For example, some UAs can show a “navigation caret”, which acts similarly to an insertion caret but can be moved around in non-editable text, and is functionally a caret. On the other hand, the cursor image shown when hovering over text when the cursor property is auto, or when hovering over an element where the cursor property is text or vertical-text, though it sometimes resembles a caret, is not a caret (it’s a cursor).

6.2.2. Shape of the insertion caret: caret-shape

This property allows authors to specify the desired shape of the text insertion caret.

Within the context of this definition, character is to be understood as extended grapheme cluster, as defined in [UAX29], and visible character means a character with a non-zero advance measure.

auto

The UA determines the shape of the caret. It should match platform conventions, and may be adjusted based on context. For example, if a UA switches between insert mode and overtype mode when the user presses the insert key on their keyboard, it may show a bar caret in insert mode, and a block caret in overtype mode.

bar

The UA must render the text insertion caret as a thin bar placed at the insertion point. This means it is between, before, or after characters, not over them. It should be perpendicular to the inline progression direction, although UAs may render it slanted when inserting italic or oblique text.

block

The UA must render the text insertion caret as a rectangle overlapping the next visible character following the insertion point. If there is no visible character after the insertion point, the UA must render the caret after the last visible character. UAs may render it as a slanted rectangle when inserting italic or oblique text.

underscore

The UA must render the text insertion caret as a thin line under (as defined in [CSS-WRITING-MODES-3] the next visible character following the insertion point. If there is no visible character after the insertion point, the UA must render the caret after the last visible character.

The width of the block and underscore carets should be the advance measure of the next visible character after the insertion point, or 1ch if there is no next visible character or if this information is impractical to determine.

When determining the orientation and appearance of the caret, UAs must take into account the writing mode [CSS-WRITING-MODES-3] and must apply transformations [CSS-TRANSFORMS-1]. If the edited text is laid out out on a path, for instance by using the SVG textPath element, UAs should also account for this.

The stacking position of the caret is left undefined, within the following constraints:

• The caret must not be obscured by the background of the element
• UAs must render block carets so that the character it overlaps with is not obscured by the caret

.console {
  caret-shape: underscore;
  background: black;
  color: white;
  font-family: monospace;
  padding: 1ex;
}

user@host:css-ui-4 $ ls -a
. .. Overview.bs Overview.html
user@host:css-ui-4 $  

user@host:css-ui-4 $ ls -a
. .. Overview.bs Overview.html
user@host:css-ui-4 $

6.2.3. Insertion caret shorthand: caret

This property is a shorthand for setting caret-color and caret-shape in one declaration. Omitted values are set to their initial values.

6.3. Keyboard control

6.3.1. Directional Focus Navigation: the nav-up, nav-right, nav-down, nav-left properties

auto

The user agent automatically determines which element to navigate the focus to in response to directional navigational input.

<id>

The <id> value is an ID selector [SELECT]. In response to directional navigation input corresponding to the property, the focus is navigated to the first element in tree order matching the selector.

If this refers to the currently focused element, the directional navigation input respective to the nav- property is ignored — there is no need to refocus the same element.

If no element matches the selector, the user agent automatically determines which element to navigate the focus to.

If the focus is navigated to an element that was not otherwise focusable, it becomes focusable only as the result of this directional navigation, and the :focus pseudo-class matches the element while it is focused as such.

Note: there were other options under consideration for such "not otherwise focusable" elements, including focus to the next otherwise focusable element in the document tree (including descendants of such a not otherwise focusable element). Input on such other options is welcome and explicitly solicited, especially from implementation experiences and author experience using the directional navigation properties in their content.

<target-name>

The <target-name> parameter indicates the target frame for the focus navigation. It is a <string> and it MUST NOT start with the underscore "_" character. Error handling: if it does start with an underscore, "_parent" navigates to the parent frame, "_root" is treated as root, and other values navigate to a frame by that name if it exists. If the specified target frame does not exist, the parameter will be treated as the keyword current, which means to simply use the frame that the element is in. The keyword root indicates that the user agent should target the full window.

User agents for devices with directional navigation keys respond by navigating the focus according to four respective nav-* directional navigation properties (nav-up, nav-right, nav-down, nav-left). This specification does not define which keys of a device are directional navigational keys.

Note: Typical personal computers have keyboards with four arrow keys. One possible implementation would be to use those four arrow keys for directional navigation. For accessibility and user convenience, user agents should allow configuration of which keys on a keyboard are used for directional navigation.

button { position:absolute }

button#b1 {
  top:0; left:50%;
  nav-right:#b2; nav-left:#b4;
  nav-down:#b2; nav-up:#b4;
}

button#b2 {
  top:50%; left:100%;
  nav-right:#b3; nav-left:#b1;
  nav-down:#b3; nav-up:#b1;
}

button#b3 {
  top:100%; left:50%;
  nav-right:#b4; nav-left:#b2;
  nav-down:#b4; nav-up:#b2;
}

button#b4 {
  top:50%; left:0;
  nav-right:#b1; nav-left:#b3;
  nav-down:#b1; nav-up:#b3;
}

button#foo { nav-left: #bar "sidebar"; }

6.3.2. Obsolete: the ime-mode property

ime-mode is a property somewhat implemented in some browsers, that is problematic and officially obsoleted by this specification and its predecessor [CSS-UI-3].

There is documentation of non-interoperability of these implementations.

User agents should not support the ime-mode property.

Authors must not use the ime-mode property.

Users may use the ime-mode property only for repair use-cases where they have to work around bad sites and legacy implementations, e.g. with a user style sheet rule like:

input[type=password] {  ime-mode: auto !important;
}

This example CSS may be placed into a user style sheet file to force password input fields to behave in a default manner.

This specification deliberately does not attempt to document the functionality of legacy ime-mode implementations nor what they specifically support because it does not make sense to pursue or recommend any such path.

7. User Interaction

7.1. Controlling content selection

The user-select property enables authors to specify which elements in the document can be selected by the user and how. This allows for easier interactions when not all elements are equally useful to select, avoiding accidental selections of neighbouring content.

User Agents must not apply the user-select property to the ::first-line and ::first-letter pseudo elements.

Note: The UA may apply this property to the ::before and ::after pseudo elements. If it does, auto value computes to none on these pseudos, but other values can be specified. This preserves the legacy behavior of generated content not being selectable or copyable, which is appropriate when these pseudos are used for decorative purposes. However, this property allows them to become selectable and copyable, as the user would expect in cases where they are used to generate part of the content, such as the issue numbers in this document. To the extent possible, authors should avoid using generated content for non decorative purposes, and should prefer including all the content in the DOM. This feature is at risk.

if we allow user-select to change to a value other than none, we need to figure out what this means for copyability, and DOM APIs.

When generated content in pseudo elements becomes selectable through this mechanism, UAs should also make this content findable through their search function.

Should it also apply to ::marker? To page-margin boxes? Should the computed value of auto also be none, or would text be more appropriate?

The computed value is the specified value, except:

1. on editable elements where the computed value is always contain regardless of the specified value
2. when the specified value is auto, which computes one of the other values as defined below

For the purpose of this specification, an editable element is either an editing host or a mutable form control with textual content, such as textarea.

Should there be constraints on what happens to the computed value on elements that are editable descendants of editing hosts? The semantics are not obvious. Maybe none should compute to text, or maybe all values should compute to text.

auto

The computed value of auto is determined as follows:

• On the ::before and ::after pseudo elements, the computed value is none
• If the element is an editable element, the computed value is contain
• Otherwise, if the computed value of user-select on the parent of this element is all, the computed value is all
• Otherwise, if the computed value of user-select on the parent of this element is none, the computed value is none
• Otherwise, the computed value is text

Note: This unusual combination of a non inherited property with an initial value of auto whose computed value depends on the parent element makes it possible to create what is effectively selective inheritance. This was initially proposed by Microsoft in IE to introduce a behavior similar to inheritance except that the contain value does not inherit.

text

The element imposes no constraint on the selection.

none

The UA must not allow selections to be started in this element.

A selection started outside of this element must not end in this element. If the user attempts to create such a selection, the UA must instead end the selection range at the element boundary.

Note: As of the time of writing, experimental implementations do not all behave like this. Firefox does. Chrome and Safari almost do: for a selection started after the element and trying to go backwards into the element they behave as specified here, but for a selection started before the element and trying to go into the element they behave as if the element has all and select it entirely. IE does not restrict selections started outside of the element from going into it at all. Another difference is that in Chrome and Safari, if the user attempts to start a selection inside a user-select: none, and to end the selection out of it, a selection will be created from the boundary of the element to the user-designated end-point. Firefox and Internet explorer behave as prescribed in this specification and do not create a selection at all.

However, if this element has descendants on which the computed value of user-select is not none, selections that start and end within these descendants are allowed.

The UA must allow selections to extend across this element, and must exclude this element from such a selection. An exception is made for UAs which do not support multiple ranges per selection, and they may include this element. If the element has descendants on which user-select does not compute to none, these descendants must be included in a selection extending across the element. This specification makes no normative requirement about the behavior of the clipboard. however, UAs are encouraged to keep the visual selection consistent with what would get copied to the clipboard when copying. Copying text that does not appear to be selected, or vice versa, is highligh confusing to users.

Attempting to start a selection in an element where user-select is none, such as by clicking in it or starting a drag in it, must not cause a pre-existing selection to become unselected or to be affected in any way.

As user-select is a UI convenience mechanism, not a copy protection mechanism, the UA may provide an alternative way for the user to explicitly select the text even when user-select is none.

Note: none is not a copy protection mechanism, and using it as such is ineffective: User Agents are allowed to provide ways to bypass it, it will have no effect on legacy User Agents that do not support it, and the user can disable it through the user style sheet or equivalent mechanisms on UAs that do anyway. Instead, none is meant to make it easier for the user to select the content they want, by letting the author disable selection on UI elements that are not useful to select. Tools such as CSS validators, linters or in-browser developer tools are encouraged to use heuristics to detect and warn against incorrect or abusive usage that would hamper usability or violate common user expectations.

Authors should be careful about not constraining the user unnecessarily. For example, setting user-select: none on the root element, and relaxing that restriction on a handful of elements the author judges useful to select can be frustrating to users:

• Clicking in empty areas to undo the current selection will no longer be effective

• The author may have overlooked some areas which should be selectable

• Areas may be added later to the page without remembering to make them selectable

• The user may want to select pieces of text other than the main content, for instance to look them up in a dictionary or in some translation tool, or to look for warnings and error messages in a search engine…

Instead, a good practice is for authors is to selectively apply user-select: none to elements which seem likely to be accidentally selected when doing so would interfere with the more likely intended action. Accidentally leaving parts of the page that are unlikely to be interacted with selectable will likely cause much less frustration to users than the opposite.

contain

UAs must not allow a selection which is started in this element to be extended outside of this element.

A selection started outside of this element must not end in this element. If the user attempts to create such a selection, the UA must instead end the selection range at the element boundary.

The UA must allow selections to extend across this element, and such selections must include the content of the element.

Note: At the time of writing, experimental implementations behave differently from eachother about selections started outside and selections going into the element. The behavior can be observed even on browsers that do not explicitly support contain by trying to select into a textarea or a contenteditable element.

Note: This was initially introduced as an experimental feature in Internet Explorer, under the name user-select: element.

all

The content of the element must be selected atomically: If a selection would contain part of the element, then the selection must contain the entire element including all its descendants. If the element is selected and the computed value of user-select on its parent is all, then the parent must be included in the selection, recursively.

If this element has descendants on which the computed value of user-select is not all and if a selection is entirely contained in these descendants, then the selection is not extended to include this whole element.

Note: Selections can include more than just text and extend over images, tables, videos, etc. The behavior when copying and pasting a such selections is out of scope for this specification.

The following additions are made to the UA stylesheet for HTML:

button, meter, progress, select { user-select: none; }

the list above is incomplete, and needs to include at least the various button-like variants of input.

For compatibility with legacy content, UAs that support user-select must also support -webkit-user-select as an alias.

Note: The details of the aliasing aliasing mechanism is intentionally left up to the UA. Making -webkit-user-select a shorthand property of user-select is known to be an effective approach, and new implementers should consider it. However UAs supporting -webkit-user-select as an alias of user-select through other means exist, without adverse consequences to compatibility, so this specification allows flexibility.

8. Form Control Styling

8.1. Switching appearance

input { appearance: none; }
input[type=checkbox] { appearance: checkbox; }

While the way most elements in a document look can be fully controlled by CSS, form controls are typically rendered by UAs using native UI controls of the host operating system, which can neither be replicated nor styled using CSS.

This specification introduces the appearance property to provide some control over this behavior. Using appearance: none allows authors to suppress the native style of form controls, so that CSS can be used to fully restyle them.

Should auto compute to none on regular elements? I would say no: If we did that, to be consistent, every time we introduced a new value, we would change what auto computes to on some elements, which doesn’t sounds desirable.

auto

UAs may render form controls using native controls of the host operating system or with a look and feel not otherwise expressible in CSS.

Elements other than form controls must be rendered as if none had been specified.

none

The element is rendered following the usual rules of CSS. Replaced elements other than form controls are not affected by this, and remain replaced elements. Form controls are not made to look like native controls of the host operating system. See below for details.

Shouldn’t this be called "normal" instead? none makes it sound like the targeted element will disappear.

On form control elements where the computed value is auto UAs may disregard some CSS properties to ensure that the intended appearance is preserved, or because these properties may not be meaningful for the chosen appearance. However, the following properties must not be disregarded:

• appearance
• display
• visibility
• position
• top
• right
• bottom
• left
• float
• clear
• margin and related long-hand properties
• unicode-bidi
• direction
• all This does not imply a requirement to apply all the properties reset by all
• cursor
• z-index

Are there more properties should include in this list? Should we remove some? Should whitelist the properties that are ok to ignore instead of blacklisting those that are not? Either way, UAs do ignore some properties when rendering form controls, so this specification needs to say something about this.

All decorative visual aspects of a form control which are not caused by a CSS style rule must be suppressed when the appearance is changed using appearance, even if the UA’s internal representation for this element was composed of multiple elements or pseudo elements combined together. For example, select is often rendered with an arrow on the right side indicating that the list will be expanded if the element is clicked. If the computed value of appearance is none, this must disappear.

However, the UA must preserve aspects of the form control which are necessary to operate the control with its original semantics. This does not include aspects of a control which are merely needed to observe the state the control is in, only those that are needed for the user to be able to modify the state of the control. The UA may however give them a different look and feel as long as it remains possible to operate the control. For example, the slider of an <input type=range> is preserved (or replaced by an equivalent mechanism) even if appearance is none as it would otherwise not be possible to change the value of the range with the mouse or touchscreen. On the other hand, the check mark in an <input type=checkbox checked> must be suppressed, as it merely indicates the state the element is in, which can be styled in different ways using the :checked selector.

UAs are inconsistent about the preceding two paragraphs. What is specified here attempts to give some logic to what is preserved and what is not, based on the use-cases for appearance.

UAs should include in their User Agent stylesheet style rules to give form controls a recognizable shape when appearance is none.

Note: Authors may therefore need to override these UA style rules to get the styling they intended, possibly by using all: unset.

This usage of the all property would remove focus indicators created by the outline property, which seems undesirable. Using all: revert would not solve it, as it would fail to remove the UA styles as intended. How can we mitigate this?

The behavior and semantics of elements remain defined by the host language, and are not influenced by this property. For example, regardless of the computed value of appearance:

• Elements which can be in different states keep this ability, and the relevant pseudo-classes match as usual.
• If a select element is activated, the UI to choose one of the associated option elements is shown (although it may look different).
• Event handlers attached to the element trigger as usual.

Conversely, changing the appearance of an element must not cause it to acquire new semantics, pseudo classes, event handlers, etc that it did not initially have. The ability for an element to be focused is also not changed by the appearance property.

input[type=checkbox] {
  appearance: none;
  all: unset;
  width: 1em;
  height: 1em;
  display: inline-block;
  background: red;
}
input[type=checkbox]:checked {
  border-radius: 50%;
  background: green;
}

Appendix A. Acknowledgments

This appendix is informative.

This specification builds upon [CSS-UI-3], which was edited and written for the most part by Tantek Çelik from 1999 to the present, first while representing Microsoft, then as an Invited Expert, and most recently while representing Mozilla.

Thanks to feedback and contributions from Rossen Atanassov, Tab Atkins, L. David Baron, Bert Bos, Matthew Brealey, Rick Byers, Ada Chan, James Craig, Michael Cooper, Axel Dahmen, Michael Day, Micah Dubinko, Elika E., Steve Falkenburg, Andrew Fedoniouk, Al Gilman, Ian Hickson, Bjoern Hoehrmann, Alan Hogan, David Hyatt, Richard Ishida, Sho Kuwamoto, Yves Lafon, Stuart Langridge, Susan Lesch, Peter Linss, Kang-Hao Lu, Masayuki Nakano, Mats Palmgren, Brad Pettit, Chris Rebert, François Remy, Andrey Rybka, Simon Sapin, Alexander Savenkov, Sebastian Schnitzenbaumer, Lea Verou, Etan Wexler, David Woolley, Frank Yan, Boris Zbarsky, and Domel.

Appendix B. Changes

This appendix is informative.

This appendix describes functional changes from the 22 Sep 2015 First Public Working Draft.

• The caret-animation property have been removed. It may be reintroduced later given sufficient evidence for use cases.
• The resize property now also applies to replaced elements representing images or videos, and to iframes.
• The string values and dual values of the text-overflow property were moved to this specification from level 3.
• Move directional focus navigation properties from level 3 to level 4
• Content from the now stable [CSS-UI-3] has been merged in.

Appendix C. Considerations for Security and Privacy

This appendix is informative.

The W3C TAG is developing a Self-Review Questionnaire: Security and Privacy for editors of specifications to informatively answer.

Per the Questions to Consider

Does this specification deal with personally-identifiable information?

No.

Does this specification deal with high-value data?

No.

Does this specification introduce new state for an origin that persists across browsing sessions?

No.

Does this specification expose persistent, cross-origin state to the web?

No.

Does this specification expose any other data to an origin that it doesn’t currently have access to?

No.

Does this specification enable new script execution/loading mechanisms?

Yes to loading, but not to execution.

The cursor property accepts <image> values which may include URLs to be loaded. These may be SVG documents which may contain scripts, but this specification requires that scripts must not be run.

Does this specification allow an origin access to a user’s location?

No.

Does this specification allow an origin access to sensors on a user’s device?

Yes.

The directional focus navigation properties indirectly allow access to the device’s keyboard navigation input mechanism such as arrow keys.

Does this specification allow an origin access to aspects of a user’s local computing environment?

No.

Does this specification allow an origin access to other devices?

No.

Does this specification allow an origin some measure of control over a user agent’s native UI?

Yes.

The cursor and caret property and sub-properties enable the page to change the display of the cursor and text insertion caret of the user agent’s native UI. In addition the outline-style property’s auto value (and thus outline shorthand) enable the page to potentially display a native focused element outline presentation around any element.

The appearance property also allows author to turn off native styling and replace it with css based styling.

Does this specification expose temporary identifiers to the web?

No.

Does this specification distinguish between behavior in first-party and third-party contexts?

No.

How should this specification work in the context of a user agent’s "incognito" mode?

No differently.

Does this specification persist data to a user’s local device?

No.

Does this specification have a "Security Considerations" and "Privacy Considerations" section?

Yes.

Does this specification allow downgrading default security characteristics?

No.

Appendix D. Default style sheet additions for HTML

This appendix is informative.

Potential additions to the base style sheet to express HTML form controls, and a few dynamic presentation attributes:

:enabled:focus {
  outline: 2px inset;
}

button,
input[type=button],
input[type=reset],
input[type=submit],
input[type=checkbox],
input[type=radio],
textarea,
input,
input[type=text],
input[type=password],
input[type=image] {
  display: inline-block;
}

input[type=button],
input[type=reset],
input[type=submit],
input[type=checkbox],
input[type=radio],
input,
input[type=text],
input[type=password],
input[type=image] {
  white-space: nowrap;
}

button {
  /* white space handling of BUTTON tags in particular */
  white-space:normal;
}

input[type=reset]:lang(en) {
  /* default content of HTML input type=reset button, per language */
  content: "Reset";
}

input[type=submit]:lang(en) {
  /* default content of HTML input type=submit button, per language */
  content: "Submit";
}

/* UAs should use language-specific Reset/Submit rules for others. */

input[type=button],
input[type=reset][value],
input[type=submit][value] {
  /* text content/labels of HTML "input" buttons */
  content: attr(value);
}

textarea {
  /* white space handling of TEXTAREA tags in particular */
  white-space:pre-wrap;
  resize: both;
}

input[type=hidden] {
  /* appearance of the HTML hidden text field in particular */
  display: none !important;
}

input[type=image] {
  content: attr(src,url);
  border: none;
}

select[size] {
  /* HTML4/XHTML1 <select> w/ size more than 1 - appearance of list */
  display: inline-block;
  height: attr(size,em);
}

select,select[size=1] {
  /* HTML4/XHTML1 <select> without size, or size=1 - popup-menu */
  display: inline-block;
  height: 1em;
  overflow: hidden;
}

select[size]:active {
  /* active HTML <select> w/ size more than 1 - appearance of active list */
  display: inline-block;
}

optgroup, option {
  display: block;
  white-space: nowrap;
}

optgroup[label],option[label] {
  content: attr(label);
}

option[selected]::before {
  display: inline;
  content: check;
}

/* Though FRAME resizing is not directly addressed by this specification,
   the following rules may provide an approximation of reasonable behavior. */

/*

frame { resize:both }
frame[noresize] { resize:none }

*/