* Remove radio component
* "Voice room" → "video room"
* Remove interactivity from video room tiles
* Update connection state when joining via widget
* Simplify room header buttons for video rooms
* Split out video room creation into a separate menu option
* Simplify room options for video rooms
* Update video room tile layout
* Tell the Jitsi widget whether it's a video channel
* Update tests
* "Voice" → "video" in more places
* Fix tests
* Re-add frame to immersive Jitsi widgets
* Comment ack
* Make updateDevices more readable
* Type FacePile
* Add voice room labs flag
Signed-off-by: Robin Townsend <robin@robin.town>
* Add more widget actions for interacting with Jitsi
Signed-off-by: Robin Townsend <robin@robin.town>
* Factor out a more generic Jitsi creation utility
Signed-off-by: Robin Townsend <robin@robin.town>
* Add utilities for managing voice channels
Signed-off-by: Robin Townsend <robin@robin.town>
* Enable creation of voice rooms
Signed-off-by: Robin Townsend <robin@robin.town>
* Force a maximized view of voice channel widgets
Signed-off-by: Robin Townsend <robin@robin.town>
* Add voice channel store
Signed-off-by: Robin Townsend <robin@robin.town>
* Factor out a more generic FacePile
Signed-off-by: Robin Townsend <robin@robin.town>
* Implement room tile changes for voice rooms
Signed-off-by: Robin Townsend <robin@robin.town>
* Add interactive radio component to the left panel
Signed-off-by: Robin Townsend <robin@robin.town>
* Test voice rooms
Signed-off-by: Robin Townsend <robin@robin.town>
* Update name of call room type
Signed-off-by: Robin Townsend <robin@robin.town>
* Clarify that voice rooms are under development
Signed-off-by: Robin Townsend <robin@robin.town>
* Use readonly
Signed-off-by: Robin Townsend <robin@robin.town>
* Move acks to the end of handlers
Signed-off-by: Robin Townsend <robin@robin.town>
* Add comment about avatar URLs coming from Jitsi
Signed-off-by: Robin Townsend <robin@robin.town>
* Don't use unicode ellipses
for translation reasons?
Signed-off-by: Robin Townsend <robin@robin.town>
* Fix tests
Signed-off-by: Robin Townsend <robin@robin.town>
* Fix tests, again
Signed-off-by: Robin Townsend <robin@robin.town>
* Remove unnecessary export
Signed-off-by: Robin Townsend <robin@robin.town>
* Ack Jitsi events when we wait for them
Signed-off-by: Robin Townsend <robin@robin.town>
Currently we have a lot of unnecessary layers being created in the app,
which is extremely problematic on Windows when using Chrome, as it
deoptimises fonts rendering due to layers mismatch; chrome's algorithm
can't figure out if the layers won't resize or overlap so it creates a
lot of additional layers, which has a side effect of turning off
sub-pixel antialiasing on Windows.
This is a first step to improve the layering in the entire app.
This follows some polish time with a designer.
The placeholder text on the two panels was tracking up/down when the width was changed. This is fixed by adjusting some of the flexbox properties to center it more safely.
We also spent some time making the notifications panel more legible while we wait for the overhaul to land.
It's become obvious that these random floating points everywhere
are unwieldy. Now they're all in one place with some fairly logical
variable names which will help out in design->implementation phase.
Font size of the whole app would ideally be controlled by a single
value. This value is currently hard coded using the :root CSS selector.
It is the intention to make this value configurable within riot. In the
interim all font-sizes have been converted to rem by the simple process
of regex. Replacing px values with their equivalent rem values assuming
a font size of 15px and then rounded to three decimal places, which was
the base at the time of this transformation.
I'm expecting another commit cleaning up rem values but I thought it
best to leave that to review.
This commit doesn't address any scaling issues. I thought it better to
land this unwieldy, mechanical, invisible change before the others
otherwise the pr would be impossible to review thoroughly.
Removes one of the two places we use Velocity, so we're one step
closer to getting rid of it for good.
Should therefore fix the fact that Velocity is leaking data entries
and therefore <hr> elements.
Hopefully also makes the logic in getEventTiles incrementally simpler,
if still somwewhat byzantine.