IE8 Bugs
IE8 RTW (69 CSS, 5 HTML, 2 DOM, 1 Javascript) Bugs
If you believe you’ve found a new bug, please , providing a description and a test case (in Standards Compliance mode with valid HTML). If it is indeed a valid bug, it’ll appear on this list along with an author credit.
02/11/11
- Added new CSS escaped tab character bug and test case
01/03/10
- Added new CSS
:first-letter float text-transformbug and test case - Added new CSS generated content position bug test case
CSS Bugs
| # | Description | Notes | Fix | ||
|---|---|---|---|---|---|
| 1 | Inherited margin with inherit keyword: computed value’s parent should be inherited | Test Case | Bug Ticket | ||
| 2 | No support for Print media descriptor when used in conjunction with the @import rule | Test Case | Bug Ticket | ||
| 3 | font shorthand incorporating inherit keyword is parsed. |
Test Case | Bug Ticket | ||
| 4 | font-family string must not contain parentheses ( ), brackets [ ], braces { }, exclamation marks !, etc. |
If an unquoted font family name contains parentheses, brackets, and/or braces, they must still be escaped per CSS grammar rules. Similarly, quotation marks (both single and double), semicolons, exclamation marks, commas, and leading slashes within unquoted font family names must be escaped. |
Test Case | Bug Ticket | |
| 5 | Illegal parsing of a z-index decimal value, instead of an integer | Bug Ticket | |||
| 6 | White-space text nodes adjacent to inline elements (with white-space:nowrap applied) are ignored when those inline elements contain a text node and adjacent space character (U+0020 (which is first or last in the elements content)) |
Regression: Bug not present in IE7 | Bug Ticket | ||
| 7 | A background image on a box with overflow:auto (or scroll) wrongly moves 1 pixel up and down while scrolling | Test Case | Bug Ticket | ||
| 8 | A media attribute that contains a recognised Media Descriptor isn’t truncated before it encounters a succeeding unknown or invalid Media Descriptor |
Bug Ticket | |||
| 9 | Bullet (filled disc) list-markers misrendered when using MS Sans Serif font | Regression: Bug not present in IE7 | Test Case | Bug Ticket | |
| 10 | Only partial implementation of CSS3 writing-mode property |
Contrary to documentation release by Microsoft, there are several rendering bugs present. | Bug Ticket | ||
| 11 | Styles within :hover block are not applied to SELECT elements |
Test Case included in Bug Ticket | Bug Ticket | ||
| 12 | ‘Flash of Un-Styled Content’ (FOUC) when using @import rule | Initial analysis suggests that using the @import rule triggers FOUC, similar to the issues seen in IE6 & IE7. | Bug Ticket | ||
| 13 | A background-color value specified isn’t distributed to multiple lines on :hover |
Credit goes to Vasil Rangelov (a.k.a. boen_robot) for bringing this bug to my attention.
Regression: Bug not present in IE7 |
Bug Ticket | ||
| 14 | Specifiying list-style-image property (with valid image path) in global selector (*) block, triggers a Tab crash | Credit must go to markmck for spotting this issue and to AlexDawson for investigating the issue further | Bug Ticket | ||
| 15 | Descender space of an inline box containing an IMG (with vertical-align:baseline, and no sibling line boxes containing text) nested in an A element, is drawn only on :hover |
Regression: Bug not present in IE7
Transitional Doctype only. Thanks to Kevin Bush for notifying me of some additions to the initial bug |
Bug Ticket | ||
| 16 | Initial first-child element retains :first-child styling even after a new preceding (same-type) element is dynamically inserted |
Credit goes to the-dees for reporting this issue | Test Case | Bug Ticket | |
| 17 | A selector incorporating the :hover pseudo-class immediately followed by a combinator (whitespace (denoting a descendant selector), “>” or “+”) followed by another simple selector, does not match the selectors intended subject. |
Credit goes to the-dees for reporting the initial issue relating specifically to when the adjacent sibling is the subject of the selector. Preliminary Test Case only displays effect of the Adjacent Sibling combinator. | Preliminary Test Case | Bug Ticket | |
| 18 | BODY element automatically obtains :focus |
Credit goes to the_dees for reporting this issue. Although not technically not a bug, since CSS doesn’t define which elements may be in the [:hover, :focus & :active] states, or how the states are entered and left, it is an unusual implementation, nonetheless. |
WIP | WIP | |
| 19 | Limited stylesheet nesting level when using @import |
Support for only three levels deep | WIP | Bug Ticket | |
| 20 | background-position:top coordinate of an inline element (with vertical-align:top) is incorrectly calculated when its sibling is a replaced element (with greater line-box height) within the same line box |
Credit goes to François Remy (FremyCompany) for reporting this issue and creating an initial test case. | Bug Ticket | ||
| 21 | A selector incorporating the :hover pseudo-class immediately followed by a General Sibling Combinator matches only first 298 elements in pattern when document is reflowed. |
Credit goes to François Remy (FremyCompany) for reporting this issue and creating an initial test case. | Bug Ticket | ||
| 22 | content property doesn’t return the value of an attribute when attribute is dynamically inserted |
Credit goes to François Remy (FremyCompany) for reporting this issue and creating an initial test case. | Bug Ticket | ||
| 23 | Multiple Type ID selector statement is incorrectly ignored | Regression: Bug not present in IE7 | Bug Ticket | ||
| 24 | Inline boxes suffer from mispositioned outline when text-align:justify applies to parent |
Bug Ticket | |||
| 25 | Computed width of a content area of a table-cell isn’t re-calculated when a padding (percentage) value is applied |
Bug Ticket | |||
| 26 | Vertical margin of child element, whose parent has a percentage-based padding value in the same direction, is ignored and instead applied to the aforementioned parent element |
Regression: Bug not present in IE7 | Apply an overflow value other than visible to the element to which the percentage padding value is applied |
Bug Ticket | |
| 27 | visibility:visible applied to inline element doesn’t override inherited visibility:hidden value applied to it’s block-level parent |
Regression: Bug not present in IE7 | Bug Ticket | ||
| 28 | Marker box is placed inside of principal box upon indentation (using the text-indent property) of that element. |
Bug Ticket | |||
| 29 | Intermittent margin collapse of last sibling element upon selecting text within preceding same-type sibling.NEEDS UPDATING | Regression: Bug not present in IE7 Another example of the bug in a seperate instance |
Bug Ticket | ||
| 30 | Content box of table-cell isn’t recalculated to reflect a percentage padding value applied to that element |
Bug Ticket | |||
| 31 | Marker box doesn’t inherit styling from element to which its attached, when importing those style rules using @import | Regression: Bug not present in IE7 | Bug Ticket | ||
| 32 | Attribute selector (which matches the attribute, ‘colspan’ for TD and TH elements) matches every TD and TH element | Bug Ticket | |||
| 33 | User-defined outline is drawn inside the border box |
Regression: Bug not present in IE7 | |||
| 34 | text-align value of ascendant isn’t inherited by the TH element |
Regression: Bug not present in IE7
This bug affects the presentation of these bug tables. Credit to mvdleij for originally reporting this |
Add text-align:inherit to TH element |
Bug Ticket | |
| 35 | Normal flow position of an empty block element (with a clear value other than none) is incorrectly honoured |
NOTE: Connect ticket incorrectly refers to margin collapsing as the issue; this isn’t the case | Bug Ticket | ||
| 36 | Only-child replaced element (with explicit dimensions, and max-width percentage value) within a floated parent, isn’t rendered |
Regression: Bug not present in IE7 | Bug Ticket | ||
| 37 | Re-computed height of a replaced element (with an intrinsic ratio, and both ‘width’ and ‘height’ specified as ‘auto’) based on a max-width value, isn’t true based on the revised width | Regression: Bug not present in IE7 | Bug Ticket | ||
| 38 | Rule set that matches an element, which includes a float value, a max-height value, and an overflow value of scroll, triggers complete page blankness (forced Standards Mode), or Compatibility View fallback |
Regression: Bug not present in IE7
Thanks to Hilbrand Edskes for pointing this one out to me. |
Bug Ticket | ||
| 39 | max-width value (other than none, which triggers content overflow) is ignored when element is floated, and has an overflow value of either auto or scroll |
Regression: Bug not present in IE7
Thanks to Hilbrand Edskes for pointing this one out to me. |
Bug Ticket | ||
| 40 | When overflow:scroll and max-height are applied to an element (in normal flow), max-height behaves like height |
Regression: Bug not present in IE7
Credit to SvenGroot for discovering this one |
Bug Ticket | ||
| 41 | text-align value isn’t inherited by :before and :after pseudo elements |
Add text-align:inherit rule to aforemenioned pseudo-element |
Bug Ticket | ||
| 42 | When overflow:auto (or overflow:scroll) is applied to element, and content within the affected element triggers horizontal overflow due to width value, max-height behaves like height |
Credit to EricWatson for spotting this one
Similar to #41 but with extra conditions |
Bug Ticket | ||
| 43 | When block-level non-replaced element in normal flow is preceded by a floated sibling element, the horizontal hover area of the floated element is reduced and based on the ’shrink-to-fit’ model | Regression: Bug not present in IE7 | Bug Ticket | ||
| 44 | Space between short line segments of a dashed border (with 1 pixel length) are rendered as white, not transparent |
Regression: Bug not present in IE7
Credit to the_dees for originally reporting this |
Bug Ticket | ||
| 45 | ‘Shrink-to-fit’ width of parent based on intrinsic width of replaced child element (when ‘width’ is specified as ‘auto’) is wrongly retained when value of max-width of child is smaller than its intrinsic width |
Regression: Bug not present in IE7
Credit to the_dees for originally reporting this |
Bug Ticket | ||
| 46 | Selector which matches a boolean attribute value incorrectly matches the minimized form of the attribute | Regression: Bug not present in IE7 | Bug Ticket | ||
| 47 | Underline which is applied to the inline box of a text node -which has a sibling element with a greater box height, so that the aforementioned inline box is vertically aligned with the top edge of the line box- is incorrectly drawn at the bottom edge of the line box | Regression: Bug not present in IE7 | Bug Ticket | ||
| 48 | ‘underline‘ text-decoration effect isn’t drawn across inline boxes generated by replaced elements |
Bug Ticket | |||
| 49 | ‘border-color’ of ascendant isn’t inherited (using ‘border-color:inherit’) when that element has a ‘color’ value defined | Bug Ticket | |||
| 50 | When an ascendant has a ‘padding’ (applied to one or more sides) value of ‘0′ or higher, and a 1px-wide bottom border is drawn on :hover (or :active) on the child (‘A’ element), the aforementioned bottom border is drawn in the incorrect position | Credit to Jaroslav Jelinek for spotting this one first and for creating an | Bug Ticket | ||
| 51 | overflow:* isn’t honoured when background-image is applied to inline descendant, and the user performs certain cursor actions |
Credit to Simon Lessard for spotting this one first and for creating an initial test case | Bug Ticket | ||
| 52 | When the , and when hovering over this whitespace node, styles applied to the :hover state of sibling elements match the first child sibling element in that series. |
Regression: Bug not present in IE7
Credit to Ray H for spotting this issue. |
Bug Ticket | ||
| 53 | Empty element with fixed width and computed height isn’t drawn when floated in the opposite direction of it’s adjacent sibling, and when it’s parent uses the shrink-to-fit computation | Issue first spotted here | Bug Ticket | ||
| 55 | Declaration which includes !important keyword is ignored when used within a :first-letter or :first-line rule |
Regression: Bug not present in IE7
Test uses single-colon pseudo element syntax |
Bug Ticket | ||
| 56 | Stacking context of relatively-positioned descendant (with lower stacking level than it’s containing block) isn’t re-calculated on ascendant :hover (or :active) when stacking level of containing block is reset to reflect default back-to-front order in document tree |
Regression: Bug not present in IE7 | Bug Ticket | ||
| 58 | List-item which has an overflow value (other than visible) applied, wrongly clips marker box when marker box is positioned outside the principal box |
Regression: Bug not present in IE7
Test also fails in Safari 4.0.3, Firefox 3.5.5 and Opera 10.0 |
Bug Ticket | ||
| 59 | background-image assigned to table-row is instead wrongly applied to first-child table-cell |
Credit to Sendeman for initially spotting this issue and for Paul O’B for investigating the issue further.
Test also fails in Safari 4.0.3 |
Bug Ticket | ||
| 60 | Text inside an anonymous table object equivalent to a table-cell element isn’t selectable |
Thanks to Vladimir Zhuravlev for writing in about this issue | Test Case | Bug Ticket | |
| 61 | A floated (in the same writing direction) child of a list-item’s principal box wrongly influences the position of the marker box positioned outside the principal box | Test Case | Bug Ticket | ||
| 62 | An element which is semi-transparent and whose contents are clipped (overflow:hidden applied to parent) when it overflows it’s parents box, experiences a 1 pixel transparency-related bleed along the outside edges that provide boundaries for the overflow |
Regression: Bug not present in IE7 Thanks to Sergey Fetiskin for writing in about this issue |
Test Case | Bug Ticket | |
| 63 | An element’s opacity value isn’t propogated to it’s positioned descendants |
Thanks to Paul O’B for writing in about this issue | Test Case | Bug Ticket | |
| 64 | padding-top value set by padding property appears below content box of INPUT (text control type) when computed width is based on intrinsic width and no keystrokes have been entered into the element |
Regression: Bug not present in IE7 Thanks to for writing in about this issue |
Test Case | Reporting system unavailable | |
| 65 | When a background-image is applied to the :first-child pseudo-element, a different background-image applied to its associated element does not display |
Regression: Bug not present in IE7 Credit to ultranerds for initially spotting this issue and Paul O’B for investigating the issue further |
Test Case | Reporting system unavailable | |
| 66 | When the inline ancestor of an absolutely-positioned descendant has border-width:0 and a border-style not equal to none, that border-width value is incorrectly computed so that it affects layout of subsequent content |
Regression: Bug not present in IE7 | Test Case | Reporting system unavailable | |
| 67 | Layer position of generated content with a higher stack level than its associated element isn’t respected when attempting to position the generated content in front of system text within the associated element | Thanks to Nicolas Gallagher for writing in about this issue | Test Case | Reporting system unavailable | |
| 68 | Applying float to :first-letter wrongly influences effect of text-transform:capitalize |
Thanks to Nicolas Gallagher for writing in about this issue | Test Case | Reporting system unavailable | |
| 69 | A character escape sequence denoting a tab character (\n), in the context of a property string value, is incorrectly parsed |
Thanks to Mathias Bynens for writing in about this issue | Test Case | Reporting system unavailable |
HTML Bugs
Please note, the list below isn’t definitive. If you want to find more HTML bugs, a good place to start is Gerard Talbot’s MSIE8 Bug List
| # | Description | Notes | ||
|---|---|---|---|---|
| 1 | Clicking an image nested within a LABEL element (with a FOR attribute, explicity binding it with another control) does not transfer focus to its associated form control | Section 17.9.1 of the HTML 4.01 spec states When a LABEL element receives focus, it passes the focus on to its associated control. |
Test Case | Bug Ticket |
| 2 | ‘Quirks’ mode trigger | An HTML comment inserted before the Doctype declaration will trigger Quirks Mode | Test Case | Bug Ticket |
| 3 | Parsing of selectors that succeed an (invalid) end-tag open delimiter within the STYLE element (HTML4 spec violation) |
Section 6.2 of the HTML 4 spec states The first occurrence of the character sequence </” (end-tag open delimiter) is treated as terminating the end of the element’s content. NOTE: Only paragraph #5 in the test case relates to this example |
Test Case | Bug Ticket |
| 4 | OBJECT element with image/* content-type isn’t visible when printed |
Credit to for initially spotting this issue | Bug Ticket | |
| 5 | Inaccessible (using ‘screen‘ device) A element title attribute when A element features descendant element(s) and all are using the inline formatting context |
Regression: Bug not present in IE7 Transitional Doctype only. Credit to Martin Sirs for submitting this issue to me. |
Test Case | Bug Ticket |
DOM Bugs
Please note, the list below isn’t definitive. If you want to find more DOM bugs, a good place to start is Gerard Talbot’s MSIE8 Bug List
| # | Description | Notes | ||
|---|---|---|---|---|
| 1 | White-space text nodes are missing from DOM | Thanks to Steve Clay for bringing this one to my attention | Test Case | Bug Ticket |
| 2 | Unexpected white-space child text node rendered after last child inline-level element |
Credit to Ray H for spotting this issue and creating an
It is accepted that for HTML, whitespace-only text nodes that appear between the parents start-tag and start-tag of the first child, and the end-tag open delimeter of the last child and the parents end-tag, shouldn’t affect rendering. |
Bug Ticket |
Javascript Bugs
Please note, the list below isn’t definitive. If you want to find more Javascript bugs, a good place to start is Gerard Talbot’s MSIE8 Bug List
| # | Name | Description | ||
|---|---|---|---|---|
| 1 | Utilising querySelectorAll within a statement that breaks out of a for loop, triggers a full program crash. |
Thanks to for pointing this one out | Bug Ticket |