- Disabled mobile mode on Windows Phone, since it’s not needed there (#408).
- Fixed broken percentage constants.
- Constants can now be defined as functions or percentage offsets (#148, #404).
breaking: When using a constant of value
100 together with percentage offsets like
the value was implicitly handled as percentage value
100p. Starting with this version you need to explcitly use
100p if you want percentage constants.
On the plus side, you can now mix an absolute constant with a percentage keyframe or a percentage constant with an absolute keyframe.
- Fixed scrolling on input elements (#397).
- Fixed keyboard not appearing on some mobile browsers when an input was focused (#349).
z-index:auto;not working because it was always coerced to an integer (#351).
- Fixed clicking on links (and other elements) on mobile (#263, #303, #338).
- Fixed the
directionparameter that’s passed to the render events (#339).
- Added support for percentage offsets (#185).
- Made the mobile deceleration configurable and lowered the default (#222, #229).
- Fixed bug which caused IE to perform an endless loop (#271).
- Improved overall performance for mobile (#249).
- Added a new option
- Changed the default value of
set. There are too many cases where
easewas not wanted and unexpected.
- Fixed IE plugin not working. This was caused by assigning
skrollr.setStyleto a local variable inside the skrollr core. Since the IE plugin monkey-patches the skrollr.setStyle function, the core didn’t notice the change (#199 comment 18986949).
- Fixed crash in IE < 9 because the detected prefix was
- Fixed that some elements got the
skrollable-afterclass at the same time.
- When resizing the browser, the scroll position was reset to 0 (#217)
- When resizing the browser,
forceHeightwas colliding with the already forced height (#216).
- Allow numbers inside of easing function names (#152).
Expect things to break when coming from 0.5! Read through the changelog. Migration is not hard.
- [breaking] There’s no more
skrollr.mobile.jsfile. You only need
skrollr.js. You no longer need to conditionally include
- You can configure how skrollr detects mobile browsers using the
mobileCheckoption (check out the documentation).
- [possibly breaking] The meaning of the
#skrollr-bodyelement changed. Put all static elements inside of it and all absolute/fixed elements outside. It doesn’t need to be the first child of the body anymore.
- [breaking] The
unrenderedclasses where renamed because they were confusing and wrong. They’re now called
skrollable-after, because that’s their meaning (the element with these classes is before/after the first/last keyframe).
- Added a new class
skrollable-between, because symmetry. That’s why.
- Added a new class
- Easing functions are now applied when exactly at a keyframe (#132).
- [possibly breaking] The behavior changed for the case when the scroll position is before/after the first/last keyframe (I’m just gonna use “before first” from now on, because “after last” is analog). In 0.5 the behavior was not exactly specified and buggy (see item above regarding #132). Skrollr was applying the styles of the first keyframe to the element for all scroll position that were before the first keyframe. E.g. when
data-100="top:200px;"was the first keyframe, the element had
top:200px;at all scroll positions before (all from
99). From now on you can specify the behavior you want (see
edgeStrategyoption for details, set it to
setfor old behavior). Note: 0.6.7 and up use
- Add a
skrollr-mobileclass to the html element when the mobile script is included.
- #131: Use a cross browser approach for getting the body scroll-height.
- #133: Use the maximum of document height or the max keyframe for edge cases where absolute keyframes are used in a relative-mode-like document and
data-endwas calculated wrong.
- #121: Fixed prefix detection in Safari.
- #126: When calling refresh(), force rerendering, even if the scrollbar didn’t move.
- #104: Fixed the most annoying bug on mobile. There was a large blank space at the bottom of the page.
- #118: Fixed broken prefix detection. Was broken all the time, but worked before Firefox 18.
- #116 + #117: SVG support was broken for relative mode.
- #103: skrollr no longer depends on being added to the bottom of the document.
- #105: Fixed inconsistent behaviour for adding
unrenderedclass on page load
- #100: Fixed relative-mode not working correctly in IE < 9 due to crippled getBoundingClientRect
- #80: When resizing the browser window the calculation of relative mode was wrong when the element’s vertical position was changed before.
- #66: Make IE 7 support a light version of
- #78: Fixed that new parser didn’t allowed omitting the last semicolon in a keyframe property list.
animateTonot working because iScroll uses negative offset.
- breaking the
pluginapi has been removed (the IE plugin has been updated to a new, hidden api).
- Full mobile support using iscroll.
- #73: Fixed parser to not mess up colons inside URLs
- #74: Fixed parser to not treat single periods as decimal numbers
- #76: Fixed dummy element overlaping the content, even though it should be unobtrusive
forceHeightnow handles relative mode like a boss.
- #59: Make
scaleoption only affect absolute mode.
- #64: Setting
scaleoption does not affect
- Allow smooth scrolling on element level using
- Added experimental smooth scrolling (no more CSS transitions. WORKS IN IE.).
- Updated the requestAnimationFrame polyfill for smoother animations
- Updated the way requestAnimationFrame is used for even smoother animations
- New method
relativeToAbsolutewhich was formerly private
- New method
isAnimatingToto check if an animation caused by
- Experimental mobile support using https://github.com/zynga/scroller
skrollrclass is added to the HTML element and a
no-skrollrclass is removed when
initis called. Useful for fallback styling.
- Added new feature “constants”.
- Added new feature “anchor-target” which allows elements to react to other elements leaving/entering the viewport.
- Fixed a bug which broke skrollr in IE caused by wrong regular expression behavior
- breaking the
data-end-[offset]syntax changed. It’s now
- Fixed a bug where white spaces between style declarations were not ignored.
- Added support for anchors. Animations can now be specified relative to the elements position within the viewport.
- Added support for SVG elements.
- Added new method