`Player#tech` can now be called without passing an object into it. It no longer requires passing an object into it, so, current code will not break.
If nothing is passed in, a warning will be logged about knowing what you're doing. If anything is passed in, the warning is silenced.
We previously hid the poster image when controls were disabled because
the poster image had a click handler on it. However, this meant that in
the case of Flash, which doesn't have a native poster image ability, we
lost the poster.
Fixes#1625.
This helps with debugging to know what a component's name is.
We try to look up the tech's name via the constructor's name property,
otherwise, we set it to 'Unknown Tech'. Can be overridden by setting
`this.name_` after calling `super()` in the constructor.
Fixes#1786.
npm start now runs `grunt dev` which in turn runs the connect server,
sets up babel watching, browserify watching (for both src and test
files), and css and alternate builds watching.
Additionally, it will copy over the dev build files into the dist folder
so that if you're linking videojs and you have a project that gets
video.js from the dist, you can link and develop with more easy without
rerunning all of `grunt dist` each time.
Additionally, this specifically does not run more time consumming
operations like minification.
In chrome 55, something changed which introduced a bug in videojs where
if you seeked back, it wouldn't work. This is because we always paused
the video in the mousedown handler.
Instead, we should create a timer for pausing that is cleared in the
mouseup handler or in the mouse move handler. This is so that if someone
is seeking by clicking and waiting the video pauses. As soon as we start
moving and we get a mousemove event, we can know that it's safe to pause
as well.
Fixes#3839.
If the videojs embed code (a video element) is wrapped in a div with the
'data-vjs-player' attribute on it, that element will be used for the
player div and a new one will not be created. In addition, on browsers
like iOS that don't support moving the media element inside the DOM, we
will not need to clone the element and we could continue to re-use the
same video element give to us in the embed code.
This could also be extended in the future to change our embed code to a
div-only approach if we so choose.
Chrome has a bug where if you add a remote text track and try to use it programmatically, you won't get any cues displayed. So, just switch to always emulated mode.
Also, add IS_SAFARI and IS_ANY_SAFARI to the browsers.
addRemoteTextTrack and removeRemoteTextTrack in Html5 tech assumed that
they are always called for native text tracks. However, Html5 tech can
have emulated text tracks. For those, we just exit early after the super
methods are called.
Safari 10 automatically dedupes duplicate class names in an element. So,
our test was failing because we had an extra "foo" in the check. This is
an unlikely scenario that has browser variations, so, better to just
remove it.
This is because IE returns 0 for both getComputedStyle and currentStyle.
However, offsetHeight and offsetWidth do contain the correct values we
want. So, if before returning in currentDimension the return value is
still zero, check the offset values.
This is used internally in the seek bar and mouse time display.
In firefox, in an iframe that is hidden with "display: none",
getComputedStyle() returns "null" which can break things.
See https://bugzilla.mozilla.org/show_bug.cgi?id=548397 for more
details.
5.12.6's changelog accidentally included all the 5.12.x changes and the
link to the comparison was from 5.10.7 to 5.12.6.
The comparison link in 5.12.3 was also updated to properly refer to
5.12.2 as the previous change.
This involves updating to latest version of uglify and setting
"screwIE8" to false since that is a default in 2.7. However, this makes
"preserveComments" "some" work as it is supposed to work.
Fixes#3707. This also comes from the stable branch after the 5.11.9
release.
If you try and require videojs in an environment that doesn't implement `document.createElement` properly -- like in Node.js -- you could potentially get an error. This checks that `window.document && window.document.createElement` is available before calling `createElement`.
We specifically check `window.document` so that `global` module won't cause issues with it's shimming of `document.createElement` in `global/document.
Fixes#3665
There was a potential breakage that was caused by #3349. This restores the timeupdate and loadedmetadata listeners in the duration display that were removed. As part of 6.0, they could be removed as durationchange should be the correct and only listener we need.
iOS still doesn't have native fullscreen API access. The video element uses the old webkit fullscreen events `webkitbeginfullscreen` and `webkitendfullscreen`. This makes it so both of those trigger `fullscreenchange` on the player always as opposed to only when `requestFullscreen` was called on the player.
Gary Katsevman