Giving out a pointer implied excessive error checking in user code.
There's a new hasApplication() accessor that can be used to check for
application presence, moreover the ScreenedApplication is now
convertible to a pointer to provide backward compatiblity. This
conversion is marked as deprecated and will be removed in a future
release.
On Emscripten as well, however I'm keeping the Configuration::setTitle()
a no-op because the title is usually set by the HTML markup already and
so dynamic code implicitly changing it to something else doesn't make
much sense.
* Wrap the text like in a console because nobody wants to read
kilometers-long horizontal lines with a horizontal scrollbar
* Make the output fill at least the space given to it but let it expand
beyond because a vertical scrollbar in an area of this size is
useless as well.
Because not all applications implement these, it's done via a "mixin". I
never did such a thing before, just got an idea that it could work and
I'm equally amazed and horrified that it actually DOES WORK. The Screen
will now expose the MouseScrollEvent, TextInputEvent and
TextEditingEvent typedefs only if the underlying application has them
too and provides the overrideable mouseScrollEvent(), textInputEvent()
and textEditingEvent() also only if the underlying application has them
-- that also means you can't `override` those if the app doesn't provide
such APIs, which acts as a nice check against accidental overgenericity.
The code was taken straight from Sdl2Application where it assumed the
main render loop polls for canvas size changes and so it would pick it
up automagically. Not the case here, so the viewport event was never
fired after this.
If building with deprecated features enabled, the buildsystem checks if
the option is still set and is inconsistent with what Corrade reports
and reports a deprecation warning. For backwards compatibility the
MAGNUM_BUILD_MULTITHREADED CMake variable and preprocessor macro are
still provided as well.
By mistake I thought it's the same as in Emscripten's SDL, but there
Emscripten does some emulation to ensure windowSize() ==
framebufferSize().
For the resize event it's possible to hook into the window resize
callback instead of polling for the size every frame, That's much more
efficient.