Initial commit #1245

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2327: Fix file mode settings. Reviewed By: Derek Anderson<derek.anderson@lge.com> Integrated-By: Ryan Duffy (ryan.duffy@lge.com)

The refactoring of moon/History (including creation of the new enyo/History singleton) fixes multiple known issues and is a significant improvement, but I'm not confident that we have things quite right yet. I therefore want to mark enyo/History as an experimental / work-in-progress API for the time being to allow us to make breaking changes or replace the API altogether in subsequent releases. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

Fixes ENYO-2083 Reviewed-By: Gray Norton (gray.norton@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Enyo-DCO-1.1-Signed-off-by: Roy Sutton <roy.sutton@lge.com>

ENYO-2205 Initialize selected items when collection changed Reviewed By: Derek Anderson (derek.anderson@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

* Fix the gulpfile to actually test the source in lib. * Refactored Control, floatingLayer, and fullscreen to be in a single module to resolve a hidden circular dependency issue. * Fixed various jshint warnings Issue: ENYO-2356 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

At least in cases where the image size was not explicitly specified and the sizing option was not being used, the placeholder image was not properly scaled to the size of the image element. Some simple CSS refactoring takes care of that. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

ENYO-2313: In enyo/Image, ensure that placeholder image fits Reviewed-By: Art Dahm (art.dahm@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Animation transformation using matrix values. Enyo-DCO-1.1-Signed-off-by: Anish Ramesan <anish.ramesan@lge.com>

When false, enyo/History will only maintain its internal history stack and will not update window.history. Issue: ENYO-2368 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Fixes ENYO-2368 Reviewed-By: Gray Norton (gray.norton@lge.com) Integrated-By: Gray Norton (gray.norton@lge.com)

This reverts commit 65c07bd, reversing changes made to ee55e3c.

Revert "Merge pull request #1256 from enyojs/ENYO-2313-graynorton" Reviewed-By: Aaron Tam (aaron.tam@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Fixes ENYO-2356 Reviewed-By: Aaron Tam (aaron.tam@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

update for 0.5.1 tools release

Adopts a new design that uses otherwise inert properties to apply the necessary a11y attributes. Also adds more comprehensive aria support to the core Enyo controls. Issue: ENYO-2176 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Issue: ENYO-2414 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

fix path to js in gulpfile

… in a certain case ## Issue If user selects REW/FF when content is paused, setPlaybackRate should be called ahead of play. ## Fix added flag, "isNeedPlay", to make node play after setPlaybackRate . Enyo-DCO-1.1-Signed-off-by: Suhyung Lee (suhyung2.lee@lge.com)

ENYO-1725: enyo.Video: setPlaybackRate should be called ahead of play in a certain case

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

refactors accessibility implementation … Reviewed-By: Seungho Park (seunghoh.park@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

ENYO-2384: Use panel container wrapper for panels. Reviewed-By: Cole Davis (cole.davis@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2421: Prevent animating a direct transition. Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Roy Sutton <roy.sutton@lge.com>

ENYO-2257 DataList/DataRepeater does not update selection with POJOs Reviewed-By: Derek Anderson (derek.anderson@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

…e if travis is happier

…until fixes have been made for issues

Rather than forcing a tabindex of 0 for any control with content or aria-label, add the tabIndex property which maps to the tabindex attribute. This will allow saner keyboard navigation rather than virtually everything being navigable via tab. Issue: ENYO-2430 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Fix for: ENYO-2434

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Ankur Mishra<ankur.mishra@lge.com>

Without the notification, the aria observer cannot update the role of the active LightPanel. Issue: ENYO-1994 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2442: Preserve panel animation in RTL mode. Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

Fixes ENYO-1994 Reviewed-By: Aaron Tam (aaron.tam@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

…obal controller

…arily execute this method but no cases currently exist) ensure that any cached bubble targets are cleaned up

…le transitioning, tighten up selectors. Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Fix for: ENYO-2460

Fix for: ENYO-2459

…rton"" This reverts commit f94e6c2.

Enyo-DCO-1.1-Signed-off-by: Roy Sutton <roy.sutton@lge.com>

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Issue: ENYO-2430 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2465: Allow for declared component block and initial index. Reviewed-By: Ryan Duffy (ryan.duffy@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

Enyo 2313 roysutton - unrevert placeholder change Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

ENYO-2444: Complete animation when tearing down a cached view. Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

Fixes ENYO-2430

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2504: TouchScrollStrategy should require touch.js Reviewed-By: Cole Davis (cole.davis@lge.com) Integrated-By: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Cole Davis (cole.davis@lge.com)

VirtualDataRepeater has, until now, routinely made refresh() an async operation. This behavior was inherited (via copy/paste) from DataRepeater when VDR was first created. It's not apparent that making refresh() async is necessary or useful for VDR, and the async operation is causing problems in at least one instance -- specifically, when the data underlying a NewDataList is modified while the list is hidden, the list isn't refreshed until the list is reshown, and at that time the "old" list contents flash momentarily before being refreshed. We address this issue by making refresh() synchronous for VDR and its subkinds. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2491: Ensure that panel transition runs.

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Reflowing a Control when it's hidden generally doesn't make sense, since a) it's wasteful to spend cycles laying out something that can't be seen; and b) the layout may depend on measuring the Control's children, and they can't be accurately measured if they're not showing. We therefore defer reflow in the case where we're hidden, and wait until we're shown. In the case where we are explicitly hidden (i.e., `showing` is `false`), we defer immediately. We can't inexpensively detect the case where we're showing but an ancestor is hidden, so to handle that case we rely on the layout to do its measurements and let us know (by returning `true`) if we need to reflow again the next time we are shown. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

ENYO-2258: Defer reflow when hidden, do it when shown

Fix for: ENYO-2310

NewDataList has some logic to defer refreshing and resetting when the list is not showing. This logic would apply equally to VirtualDataRepeater, so we'll move it there. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

ENYO-2514: VDR: Always refresh immediately Reviewed-By: Cole Davis (cole.davis@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

Enyo-DCO-1.1-Signed-off-by:Madala Satyanarayana<cholanmadala1510@gmail.com>

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

https://jira2.lgsvl.com/browse/PLAT-8119 Enyo-DCO-1.1-Signed-off-by: Bongsub Kim bongsub.kim@lgepartner.com Change-Id: Ia5e83dc042d55cbfec3f44c00764f8cc56c32ec0

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Art Dahm art.dahm@lge.com

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Ankur Mishra<ankur.mishra@lge.com>

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2536: Guard against initialization edge cases Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

In `NewDataList`, the logic for constraining the minimum scroll threshold value was susceptible to failure in the case where the values being compared were long floating point values (we were checking for equality, but in the failure case two values that were "conceptually equal" for the purpose of our algorithm were actually different in digits far to the right of the decimal). We replace the equality operator with a call to `Math.min()` to constrain the threshold value, which gives us the desired result. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Some time ago, we added some logic to prevent resize events from waterfalling through hidden UI components. The intention was to prevent unnecessary layout from happening. This approach generally worked, but there were some problem cases, as noted in a comment that went in with the original change. Specifically, a component that is hidden and therefore never receives a resize event has no way of "knowing" that a resize has been requested, so won't reflow its layout when it is later shown. We recently added some different but related logic to allow components to defer reflows when hidden and perform them later upon being shown. This newer mechanism has the advantage of avoiding the problem described above. Since the expensive part of dealing with a resize event is almost always performing a reflow, it seems that we can safely remove the prevent-resize-event logic and rely on the new defer-reflow logic to address the original problem. I did some quick tests on a range of Moonstone samples with instrumentation in place to ensure that we weren't substantially increasing the number of resize events flowing through the UI. Based on these tests, it appears that the number of additional resize events is not significant, and since virtually all of the processing that would have been caused by the additional events is short-circuited by the defer-reflow logic, I feel good about this change. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

ENYO-2565: Bullet-proof logic for constraining min threshold

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2441 Fix failing Enyo unit tests and re-enable testing on Travis

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

PLAT-6855: Allow overriding of cacheViews property. Reviewed-By: Krishna Rangarajan (krishna.rangarajan@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

== Issue The Video component would not always successfully start playing the video after a rewind operation. == Solution This problem was, at least in part, caused by the need to emulate rewind on platforms that don't support negative `playbackRate`s. Additionally, a problem was discovered with handling the switch between slow and fastForward. The slowForward playback rate array was updated to remove the rate of '1' (normal speed) and the fastforward mechanism is updated to move to play when reaching the end of the playback rate array (Previously, it rotated to the fast forward array after playing the '1' rate). There should be no change for apps that did not alter the playback rate array. The player has also been updated so that rewinding after a slowForward will now result in a slowRewind. The converse also applies.

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

PLAT-8119: Add or modify transition related functions for safari browser Reviewed-By: Seungho Park (seunghoh.park@lge.com) Reviewed-By: Aaron Tam (aaron.tam@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2592: Add replaceAt method.

ENYO-2556: Allow for reversing animation in RTL. Reviewed-By: Art Dahm (art.dahm@lge.com) Integrated-By: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Issue: ENYO-2619 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Issue: ENYO-2622 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Issue: ENYO-2619 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2625: Prevent overriding of panel content directionality.

Enyo-DCO-1.1-Signed-off-by: Roy Sutton <roy.sutton@lge.com>

ENYO-2416 Fix issues with VideoPlayer Reviewed-By: Krishna Rangarajan (krishna.rangarajan@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

ENYO-2557: Allow resize events to waterfall freely Reviewed-By: Art Dahm (art.dahm@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

…ld run linter and unit tests, but each can still be executed separately as they are separate by concern Enyo-DCO-1.1-Signed-Off-By: Cole Davis (cole.davis@lge.com)

Fixes ENYO-2619 Reviewed-By: Cole Davis (cole.davis@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Fixes ENYO-2622 add back linting for npm test Reviewed-By: Art Dahm (art.dahm@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

ENYO-2541: Remove popOnForward. Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

enyo/Scrollable provides a `scrollToControl()` method that has a generalized implementation for calculating a Control's position and scrolling to it. In the case of enyo/NewDataList, it's possible to provide an alternative implementation that is more performant and more accurate (more performant because it doesn't need to query the DOM for bounds, more accurate because the generalized implementation is influenced by current scroll position and may give slightly different values when animating vs. not). We tweak enyo/Scrollable to defer to an alternative method when one is provided, and we provide an alternative method in enyo/NewDataList. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

Enyo-DCO-1.1-Signed-off-by:Rajkumar Selvam<rajkumar.selvam@lge.com>

Enyo-DCO-1.1-Signed-off-by: Blake Stephens <blake.stephens@lge.com>

ENYO-2667: Account for queue being modified during iteration.

ENYO-2674: Exit early if addClass receives no classname

…le for the easing functions Enyo-DCO-1.1-Signed-off-by:Madala Satyanarayana<madala.satyanarayana@lge.com>

…s of Bezier Curve for easing animation. Enyo-DCO-1.1-Signed-off-by:Rajkumar Selvam<rajkumar.selvam@lge.com>

ENYO-2644: Override scrollToControl() in enyo/NewDataList Reviewed-By: Art Dahm (art.dahm@lge.com) Integrated-By: Ryan Duffy (ryan.duffy@lge.com)

ENYO-2673: Prevent bounds of zero when restoring a cached view. Reviewed By: Derek Anderson <derek.anderson@lge.com> Integrated-By: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-off-by: Stephen Choi <stephen.choi@lge.com>

functions is updated after unit testing to handle invalid parameters. Enyo-DCO-1.1-Signed-off-by:Rajkumar Selvam<rajkumar.selvam@lge.com>

When re-rendering a scrolled scroller using TranslateScrollStrategy with translateOptimized: true, the startX/Y values would be updated to be a non-zero value matching the prior values of scrollLeft/Top. However, the scrollNode was also translated the same amounts. As a result, when the user scrolled, the content would jump the amount of startX/Y resulting in whitespace above the content. Since the purpose of startX/Y is to offset the transform the amount of the scrollLeft/Top of the scrollNode but those values are never updated with translateOptimized: true, we'll reset them to 0 in rendered() so no offsetting occurs. Issue: ENYO-2691 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Fixes ENYO-2691 Reviewed-By: Gray Norton (gray.norton@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

…hile clearing

ENYO-2698: Reset queue and history instead of enqueueing drop when clearing history

…et executed Enyo-DCO-1.1-Signed-off-by: Stephen Choi <stephen.choi@lge.com>

This reverts commit 37b81a6. The original fix for ENYO-2644, which we're reverting here, didn't really address the issue. It made the symptoms disappear, but not in the way originally intended because the issue wasn't properly diagnosed. It also introduced some new issues because it unintentionally skipped some logic related to scrolling options. There were actually some performance- and customization-related benefits to the change, so we'll likely want to reintroduce a version of it at some point in the future, but we'll simply revert it for now since we are at a point in the release cycle where we need to minimize risk of regressions. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

Enyo-DCO-1.1-Signed-off-by: Stephen Choi <stephen.choi@lge.com>

Often, if you're scrolling to a target near the edge of a scrollable region, it makes sense to scroll all the way to the edge. To support this use case, we add an (experimental, for now) option called `boundarySnapThreshold` to ScrollMath. At the same time, we do a bit of minor code cleanup on another experimental feature of a similar nature (which supports snapping to the boundaries of items within the scrollable region in the specific case where the items are evenly sized and spaced). Due to time constraints and the fact that these features are not ready for wide use, leaving docs as a TODO for the moment. This change was precipitated by an issue filed against Moonstone, where 5-way scrolling to the first / last item in a list was not scrolling all the way to the scroller boundary; as a result, the up / down arrows in the scroll controls were not being disabled, making it unclear that the end of the list had been reached. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

Enyo-DCO-1.1-Signed-off-by:Ankur Mishra<ankur.mishra@lge.com>

…imate as expected Enyo-DCO-1.1-Signed-off-by:Rajkumar Selvam<rajkumar.selvam@lge.com>

Enyo-DCO-1.1-Signed-off-by:Rajkumar Selvam<rajkumar.selvam@lge.com>

Enyo-DCO-1.1-Signed-off-by:Madala Satyanarayana<madala.satyanarayana@lge.com>

Our old logic for updating `childrenByIndex` worked in most cases, but wasn't bullet-proof. In particular, it didn't do well in cases where the indices of existing children changed in "irregular" ways. The issue that exposed the problem was one in which a filter was applied to the underlying collection, causing a potentially random set of items to "disappear" and the indices of remaining items to change, but similar problems could also have occurred when non-contiguous items were added or removed, or when items were reordered. The new logic should work in all such cases. Also did some minor cleanup and added some inline comments describing the logic in the infamous `doIt()` method. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

ENYO-2644: Revert flawed initial fix, try another approach Reviewed-By: Ryan Duffy (ryan.duffy@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

ENYO-2669: VDR: Fix issue with child index Reviewed-By: Derek Anderson (derek.anderson@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

ENYO-2697: Stop job when clearing to ensure drop handlePop does not g… Reviewed-By: Derek Anderson (derek.anderson@lge.com) Integrated-By: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

…ls.replaceAt() Enyo-DCO-1.1-Signed-off-by: Stephen Choi <stephen.choi@lge.com>

ENYO-2705: Change function from popPanel to removePanel for LightPane… Reviewed-By: Derek Anderson (derek.anderson@lge.com) Integrated-By: Gray Norton (gray.norton@lge.com)

Enyo-DCO-1.1-Signed-off-by: Anish Ramesan<anish.ramesan@lge.com>

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Stephen Choi <stephen.choi@lge.com>

In the original fix for ENYO-2514, we modified VDR.refresh() to make it work synchronously, as the historical reasons for making it async (inherited from DataRepeater) seemed not to apply, and the async behavior was causing a problem in one case: when the refresh was deferred because the repeater was hidden, the "stale" state of the repeater would flash momentarily on screen after the repeater was re-shown but before the refresh occurred. Making refresh() synchronous fixed that case, but it appears that there are still some cases where async refresh has significant perf benefits, so in this change we're restoring the async behavior and taking a different approach to fixing ENYO-2514. The new issue (ENYO-2696) that prompted the change back to async was one in which the collection underlying the repeater was emptied and immediately repopulated. In this case, the async approach results in debouncing the refresh; it only refreshes once, after repopulating, rather than once after emptying and again after repopulating. To solve the flashing-stale-content issue, we now use the CSS visibility property to make the repeater invisible while it's in the process of performing a deferred refresh. While making this change, also did some cleanup and added some inline comments. Finally, note that the logic for deferring refresh() and reset() originally lived in NewDataList, not in VirtualDataRepeater. In the process of reviewing and integrating the original fix for ENYO-2514, we determined that it made sense to move it into VDR and wrote up a new issue (ENYO-2533) to track that work. Since we're back in the same code now, we're making that change, too. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

The placeholder feature in enyo/Image currently has logic to remove the placeholder when the image itself loads. This logic was intended to provide support for images that contain transparent or semi-transparent pixels, by preventing the placeholder from "showing through" those pixels. However, supporting this use case can cause expensive style recalculation under some circumstances -- specifically, it was causing grid list regeneration on mid-range TV hardware (M16) to be half a second slower. Since the 80+% use case for placeholder images involves traditional opaque, rectangular images, we'll simply remove the sometimes- expensive logic and document the limitation. We can consider reintroducing placeholder support for transparent images at some future point, either with a flag to explicitly turn it on or with an as-yet-not-thought-of, better-performing implementation. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

Enyo-DCO-1.1-Signed-off-by: Anish Ramesan(anish.ramesan@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by:Ankur Mishra<ankur.mishra@lge.com>

ENYO-2719: Remove panel in reverse order when replacing panels

Enyo-DCO-1.1-Signed-off-by: Madala Satyanarayana<madala.satyanarayana@lge.com>

ENYO-2533, ENYO-2696, ENYO-2514: Overhaul refresh / reset in VDR, NewDataList

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2768: Set visibility to a valid property value. Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

Issue: ENYO-2766 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

When not using sizing, the placeholder would not be set because a change notification doesn't fire for it during create. Adding a check for a falsey prop so placeholder will be set during create. Issue: ENYO-2734 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

We have recently started getting more aggressive about view virtualization for performance reasons. One tactic in this area involves tearing down the DOM representation of a Control and then restoring it later as needed ("view caching"). In the case where the view being cached / restored contains a DataList, the state of the view was not being preserved as expected because DataList effectively resets upon being (re-)rendered. The proper fix for this issue would involve rethinking the DataList lifecycle with view-virtualization scenarios in mind, but this would be a significant effort and the current DataList is on a relatively short road to replacement, so for now we'll go with a lighter-weight, Moonstone-specific solution. That solution does, however, require a number of minor tweaks and enhancements in Enyo: * A minor change to the delegation interface between enyo/Control and enyo/HTMLStringDelegate, allowing `teardownChildren()` to be more easily extended or overridden. * A new `callback` argument for `enyo/DataList.scrollToIndex()`, allowing for actions to be performed after this potentially asynchronous operation is completed. * A change in `enyo/ViewPreloadSupport.cacheView()` to defer removing the view's DOM node until after its children have been torn down. This enables views to hook `teardownChildren()` and do things in preparation for being torn down that take current DOM state into account. * To support the above change in enyo/ViewPreloadSupport, a minor refactoring of some cross-platform node-removal logic, moving it from enyo/Control to enyo/dom. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

ENYO-2734: Remove transparent-image placeholder support

Fixes ENYO-2766 Reviewed-By: Stephen Choi ( stephen.choi@lge.com ) Integrated-By: Cole Davis (cole.davis@lge.com)

ENYO-2744: Tweaks to allow restoration of list state on render Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

A prior change updated clear() to only reset the internal history so this change corrects clear() to both reset the internal history and keep the browser history in sync. Issue: ENYO-2699 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-off-by: Roy Sutton <roy.sutton@lge.com>

Enyo-DCO-1.1-Signed-off-by: Aaron Tam <aaron.tam@lge.com>

ENYO-2780: Disable user input monitoring, by default.

In the fix for ENYO-2744, we made a tweak to the interface between enyo/Control and enyo/HTMLStringDelegate to make it possible for Controls to override `teardownChildren()`. In doing so, we failed to preserve the `cache` argument to `teardownChildren()`, which caused a regression for Controls that use the 'renderOnShow' feature. This regression could have been addressed by simply making sure we passed the `cache` argument properly, but on further reflection it seems safer and more future-proof to instead revert the previous change in favor of an explicit `beforeTeardown()` lifecycle method that Controls can implement as needed. Doing that now. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

ENYO-2788: Provide an explicit `beforeTeardown()` hook for Control Reviewed-By: Cole Davis (cole.davis@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Enyo-DCO-1.1-Signed-off-by: Ankur Mishra<ankur.mishra@lge.com>

Fixes ENYO-2699 Reviewed-By: Stephen Choi (stephen.choi@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Issue: ENYO-2771 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

…ramework Enyo-DCO-1.1-Signed-off-by: Rajkumar Selvam<rajkumar.selvam@lge.com>

Fixes ENYO-2771 Reviewed-By: Blake Stephens (blake.stephens@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Enyo-DCO-1.1-Signed-off-by: Roy Sutton <roy.sutton@lge.com>

This reverts commit 14fe001, reversing changes made to 7b738e9.

This reverts commit d7dba4c, reversing changes made to 05c5484.

The `accessibilityPreventScroll` feature prevents the browser engine from doing bad auto-scrolling things when an item is focused in one of our scrollers. It works by listening for native scroll events and forcibly resetting native scrollTop and scrollLeft properties as needed. Our logic for resetting scrollLeft did not previously account for RTL, so we fix that. As noted inline, the current fix is specific to Blink and WebKit; we'll need to do something clever to make this work cross-browser, but not attempting that now because cross-browser support is not urgent for the imminent internal release, and this is not our only cross-browser scrolling issue. Wrote up ENYO-2841 to track future work. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

The logic for scrolling to a child of Scrollable has until now been essentially monolithic, but we now refactor it to allow parts of the logic to be selectively overridden. Specifically, there are now separate methods for getting the offset coordinates of the child (`getChildOffsets()`), calculating the target scroll coordinates (`getTargetCoordinates()`), and actually effecting the scroll (`scrollToTarget()`). The immediate use case is to provide a custom implementation of `getChildOffsets()` for NewDataList, since NewDataList can get child offsets more efficiently than the default implemenation does, and can also properly account for the spacing of list items in its own implementation. (Not properly accounting for spacing was the actual root cause of ENYO-2644 and ENYO-2783, the issues that led to this work. As part of this pull request, we are also reverting earlier, flawed solutions to the problem described in ENYO-2644). Note that we also have a use case for overriding `scrollToTarget()` (see ENYO-2710 and ENYO-2729), but that work is not required to fix ENYO-2783 so we aren't doing it now. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

It has taken a few tries, but I think this finally gets us to the bullet-proof child-indexing solution we've been chasing. We simply rebuild the index from scratch each time we run the virtualization logic in doIt(). Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

This reverts commit 6d477c1. Reverting this change because it turns out not to be necessary to address any currently known issue, and we're at a point in both the product release cycle and the lifespan of enyo/DataList where minimizing risk of new regressions is our top priority. Leaving in the git log, though, in case we run into problems in this area and want to retrace our steps. Enyo-DCO-1.1-Signed-Off-By: Gray Norton (gray.norton@lge.com)

ENYO-2840: Make `accessibilityPreventScroll` work in RTL Reviewed-By: Cole Davis (cole.davis@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

ENYO-2808 Add warning about incorrect fixedChildSize Reviewed-By: Gray Norton (gray.norton@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Issue: ENYO-2844 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

ENYO-2816: Fix VDR child indexing logic (again) Reviewed-By: Blake Stephens (blake.stephens@lge.com) Integrated-By: Cole Davis (cole.davis@lge.com)

ENYO-2783: Refactoring in enyo/Scrollable to allow customization

Issue: ENYO-2858 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Fixes ENYO-2858 prevent scrolling when focusing on a panel while transitioning in Reviewed-By: Roy Sutton (roy.sutton@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by:Madala Satyanarayana<madala.satyanarayana@lge.com>

Issue: PLAT-13387 Enyo-DCO-1.1-Signed-off-by: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Fixes PLAT-13387 Reviewed-By: Gray Norton (gray.norton@lge.com) Integrated-By: Aaron Tam (aaron.tam@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

…-dev

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Fixes ENYO-2844 Integrated-By: Ryan Duffy (ryan.duffy@lge.com)

Enyo-DCO-1.1-Signed-Off-By: Jim Tang (jim.tang@lge.com)

Enyo-DCO-1.1-Signed-off-by:Madala Satyanarayana<madala.satyanarayana@lge.com>

…O-2174-animationmodule

Enyo-DCO-1.1-Signed-off-by: Ankur Mishra<ankur.mishra@lge.com>

…/enyo into ENYO-2174-animationmodule

Enyo-DCO-1.1-Signed-off-by: Anish Ramesan<anish.ramesan@lge.com>

Enyo-DCO-1.1-Signed-off-by:Arzoo Rai<arzoo.rai@lge.com>

Enyo-DCO-1.1-Signed-off-by: Ankur Mishra<ankur.mishra@lge.com>

Enyo-DCO-1.1-Signed-off-by:Arzoo Rai<arzoo.rai@lge.com>

Commits on Aug 14, 2015

Update version string to 2.6.0-pre.14.1

Jenkins committed Aug 14, 2015

Configuration menu

View commit details

Copy full SHA for c9acd0d

Browse repository at this point

Copy the full SHA

c9acd0d View commit details

Browse the repository at this point in the history

Commits on Aug 15, 2015

Update version string to 2.6.0-pre.14.1

Jenkins committed Aug 15, 2015

Configuration menu

View commit details

Copy full SHA for 8ccaac0

Browse repository at this point

Copy the full SHA

8ccaac0 View commit details

Browse the repository at this point in the history

Commits on Aug 17, 2015

Add mixins for Interfaces

ankurmishra32 committed Aug 17, 2015

Configuration menu

View commit details

Copy full SHA for 6bf9f5b

Browse repository at this point

Copy the full SHA

6bf9f5b View commit details

Browse the repository at this point in the history

Commits on Sep 30, 2015

Update version string to 2.6.0-pre.17

Jenkins committed Sep 30, 2015

Configuration menu

View commit details

Copy full SHA for 78920cd

Browse repository at this point

Copy the full SHA

78920cd View commit details

Browse the repository at this point in the history

Commits on Nov 18, 2015

Update version string to 2.6.0-pre.20

Jenkins committed Nov 18, 2015

Configuration menu

View commit details

Copy full SHA for 6b2d06f

Browse repository at this point

Copy the full SHA

6b2d06f View commit details

Browse the repository at this point in the history

Commits on Nov 30, 2015

Update version string to 2.6.0-rc.1

Jenkins committed Nov 30, 2015

Configuration menu

View commit details

Copy full SHA for c1abdd8

Browse repository at this point

Copy the full SHA

c1abdd8 View commit details

Browse the repository at this point in the history

Initial commit #1245

Are you sure you want to change the base?

Initial commit #1245

Commits on Aug 14, 2015

Commits on Aug 15, 2015

Commits on Aug 17, 2015

Commits on Aug 18, 2015

Commits on Aug 19, 2015

Commits on Aug 20, 2015

Commits on Aug 21, 2015

Commits on Aug 22, 2015

Commits on Aug 24, 2015

Commits on Aug 25, 2015

Commits on Aug 26, 2015

Commits on Aug 27, 2015

Commits on Aug 28, 2015

Commits on Sep 1, 2015

Commits on Sep 2, 2015

Commits on Sep 3, 2015

Commits on Sep 4, 2015

Commits on Sep 5, 2015

Commits on Sep 7, 2015

Commits on Sep 8, 2015

Commits on Sep 9, 2015

Commits on Sep 10, 2015

Commits on Sep 11, 2015

Commits on Sep 14, 2015

Commits on Sep 15, 2015

Commits on Sep 16, 2015

Commits on Sep 17, 2015

Commits on Sep 21, 2015

Commits on Sep 22, 2015

Commits on Sep 23, 2015

Commits on Sep 24, 2015

Commits on Sep 25, 2015

Commits on Sep 28, 2015

Commits on Sep 29, 2015

Commits on Sep 30, 2015

Commits on Oct 1, 2015

Commits on Oct 2, 2015

Commits on Oct 3, 2015

Commits on Oct 5, 2015

Commits on Oct 6, 2015

Commits on Oct 7, 2015

Commits on Oct 8, 2015

Commits on Oct 9, 2015

Commits on Oct 10, 2015

Commits on Oct 12, 2015

Commits on Oct 13, 2015

Commits on Oct 14, 2015

Commits on Oct 15, 2015

Commits on Oct 16, 2015

Commits on Oct 19, 2015

Commits on Oct 20, 2015

Commits on Oct 21, 2015

Commits on Oct 23, 2015

Commits on Oct 26, 2015

Commits on Oct 27, 2015

Commits on Oct 29, 2015

Commits on Oct 30, 2015

Commits on Nov 2, 2015

Commits on Nov 3, 2015

Commits on Nov 4, 2015

Commits on Nov 5, 2015

Commits on Nov 6, 2015

Commits on Nov 7, 2015

Commits on Nov 9, 2015

Commits on Nov 10, 2015

Commits on Nov 11, 2015

Commits on Nov 12, 2015

Commits on Nov 14, 2015

Commits on Nov 16, 2015

Commits on Nov 18, 2015

Commits on Nov 19, 2015

Commits on Nov 20, 2015

Commits on Nov 27, 2015

Commits on Nov 30, 2015

Commits on Dec 4, 2015

Commits on Dec 5, 2015

Commits on Dec 7, 2015