I attempted to make it private only to discover it was used by Magnum
Player to make the workflow with opt-in tweakable constants more
efficient. So let's document that.
Unlike the Key enum, which shows what a user would perceive as given key
in a particular layout, the scancode is a layout-independent identifier
for e.g. WASD movement in games.
Unfortunately the API availability is wildly different among the
toolkits -- SDL's is the most complete, GLFW is second, and then there's
Emscripten / HTML5 which provides just string identifiers. I tried to
add these for X11 as well, but quick googling led to a SO question where
it was left unanswered. Not worth my time.
There's the obvious advantage of them now being shorter to type, as one
no longer needs to prefix them with KeyEvent::. But the main reason I
did this was to allow various direct keyboard state queries to be
implemented, such as isKeyPressed(). With them being hidden in the event
class the only way would be to put the query directly there as well,
which isn't nice and is also not very discoverable.
A similar case was with mouse buttons, but that was already fixed with
the PointerEvent rework that happened in earlier commits. There the
additional complication was that MouseEvent::Button and
MouseMoveEvent::Button were incompatible enums. Application::Pointer
fixes that now as well.
Partially reverts 589b3e594d -- I got it
backwards. It actually makes more sense from a user perspective to have
layout-dependent key identifiers, because for for e.g. using Ctrl-Z as
the "Undo" shortcut it's the most important what key the user perceives
as Z, not where that key is placed on some (again, from the user PoV
basically arbitrary) layout. For game-y usage, e.g. the WASD movement,
scancodes matter (and appropriate getters will get added), but for 90%
of actual user interfaces the local layout is how the keys should be
defined. This is consistent with e.g. how Qt does it (and it has
scancodes as an additional, completely untyped getter), another data
point that's in favor of keycodes is that SDL's SDLK_* defines are far
shorter than the corresponding SDL_SCANCODE_* scancodes.
As usual, the most trash fire platform of them all. Ugh. I chose to
ignore certain aspects and suggestions and made it behave more like
Emscripten and SDL2, because that makes more sense to me.
Co-authored-by: nodoteve <nodoteve@yandex.com>
The impossible-to-reliably-disable behavior with compatibility mouse
events is quite a headache. I wish Emscripten implemented pointer events
already so I could ditch this mess -- especially the array of 32 touches
where all of them but one will be unchanged is stupid.
For the internals unfortunately, EmscriptenMouseEvent and
EmscriptenTouchEvent have no common base, so I had to give up on the
current way of querying the event struct directly from event getters, as
that'd be too nasty with the branching and casts. Instead the relevant
fields are put directly into the events themselves.
HTML5 also doesn't provide any relative pointer position. For the mouse
it was rather straightforward, but for the up-to-32 touches I have to
maintain an array of per-finger positions and match them by ID.
Hopefully the linear lookup is fine. I'll probably use the same approach
for the AndroidApplication.
This makes 2.0.6 as the oldest supported because in older versions it's
not possible to disable touch to mouse event translation, and it'd be
too annoying to have it special-cased there. The version bump should be
fine as Ubuntu 18.04 has 2.0.8.
All the new pointer events have float positions, this one was the odd one
out. And I didn't like the name anymore, so I took that as an opportunity
to change the position() data type without introducing a breaking change
for everyone.
Another considered change was adding Z offset to it, since HTML5 APIs
have that. However, all my googling led to just a single SO question from
2015, where someone said it's for trackballs that can navigate in 3D
space. I'm not sure if *scroll* is actually the best way to report those,
and since SDL3 didn't bother adding that and neither Android nor WINAPI
have anything like that, I'm not bothering either.
Unlike the previous commits, this is done for all apps at once, because
it's a comparatively simpler change. The only odd one out is
AbstractXApplication, where I introduced MouseScrollEvent just a few
commits back, so I simply renamed it without leaving a deprecated copy.
Then, ScreenedApplication needed some extra logic to handle the case of
apps not implementing any scroll event at all.
Because this overrides the base pointer*Event() implementations, it
additionally has to call into the parent implementation in order to fall
back to the deprecated mouse event if the new pointer event isn't
handled.
Pointer events are an unified abstraction over mouse, touch, pen and
potential other yet-to-be-invented pointer-like input methods. Their goal
is to expose all such input methods under a single interface so the
application side doesn't need to explicitly make sure that it's
touch-aware or pen-aware. This abstraction is already present in HTML5,
in Qt6 and in WINAPI as well, and is also what I adopted for the new UI
library because it *just makes sense*.
Unfortunately not even SDL3 took the opportunity to introduce that and
instead added a *third* separate event type for pen input in SDL3. At
first I thought that I wouldn't introduce any extra abstractions in the
Application classes (because that's what they are designed to be, as
lightweight as possible), but midway through introducing TouchEvent
classes and fighting SDL's touch->mouse and mouse->touch compatibility
translation (yes, it's both ways, depending on the platform) I realized
that a much simpler solution that doesn't require any event translation
or the users duplicating their event handling logic for several possible
input types is to introduce a single new event type that covers all.
Which is what this commit does -- it doesn't introduce anything
touch-related so far, just creates a new PointerEvent and
PointerMoveEvent class and corresponding virtual functions. Additionally,
I took this as an opportunity to make the position floating-point, since
that's what SDL3 does now as well, and GLFW did so since ever.
Plus, the Pointer and Pointers enums are directly on the Sdl2Application
class, to allow me to *finally* introduce pointer state queries. Which
weren't possible until now, because there were mutually incompatible
MouseEvent::Button and MouseMoveEvent::Button enums and putting them on
the base class would mean one would have to be translated and the other
not. With Pointer it's translated always, because there isn't any similar
enumeration in SDL that would cover mouse, touch and pen at the same
time.
I wanted to add this for GlfwApplication, only to realize the timer there
is a double, in seconds, so accepting *integral* milliseconds there felt
very weird. Let's use the fancy new time types instead.
Also updated the docs to (hopefully) clarify that setSwapInterval() has
to be called in order for the interaction between the two to work
properly.
As usual, the old variant taking untyped milliseconds is a deprecated
alias to this one.
In short this means the users no longer need to care about FindSDL2 etc.,
it'll just get done automatically. They don't need to drag along
FindMagnum / FindCorrade anymore either, but having them gives a much
more reasonable error message if the library cannot be located at all, so
I still recommend having them.
This is just documentation-facing, instead of documenting an overload as
having a default-constructed argument simply show just one constructor
with a default value.
Also simplify the wording. They either create an OpenGL context or not
(or not implicitly).
This makes them consistent with window and framebuffer size queries,
that are also not cached but queried every time. It fixes a case where a
global UI scaling change in the OS triggered a viewport event but the
event didn't actually have the DPI scaling value updated.
It doesn't however handle actual explicit DPI change events yet, that's
another nightmare altogether.
I'm getting kinda pissed at the strange defaults. Should I be using a
different toolkit altogether because SDL IS FOR GAMES ONLY? Given how
many bugreports and complaints there is about "Dosbox blocking
screensaver" I'm beginning to think it's SDL's fault, not mine.
Let the users control what they want, ffs, don't enable problematic
features like blocking powersave or disabling compositor by default.
Those should be a runtime option anyway, similarly to how video players
block powersaving only when *an actual video is playing*.
Similar to the change done in Corrade, see the commit for details:
878624ac36
Wow, this is probably the most backwards-compatibility code I've ever
written. Can't wait until I can drop all that.
Which allows to get rid of a now-unneeded ArrayView include in the
header. Also, now that we're returning a StringView, it's useful to
have the view always null-terminated. In SDL2 and Emscripten it was
already like that, GLFW needed a minor change.
Spent several hours trying to figure out why magnum-player built against
static SDL (which uses the flags documented here) was hellishly laggy.
Didn't expect it was due to timers missing. OTOH, why SDL does even
allow itself to be built in such an unusable state?
Except CGL, iOS, AndroidApplication and AbstractXApplication which are
either too crappy or don't have the needed scaffolding for specifying
context flags yet.
With this flag set (which is done implicitly for all windowless apps
and, conversely, not done for all windowed apps), the default
framebuffer state isn't touched in any way, which should avoid potential
race conditions with default framebuffer on another thread.
This removes one unnecessary allocation from each application startup.
In some cases of the windowless apps the Platform::GLContext could be
put directly into the class, in other cases it had to be wrapped in an
Optional because we need delayed construction and/or earlier
destruction.
Disabling engine startup log or modifying enabled extensions /
workarounds from the application side was one of the common pain
points and this should *finally* solve the problem. This Configuration
is now inherited by the usual Platform::*Application::GLConfiguration /
Platform::Windowless*Application::Configuration classes people are used
to, so for the end user it's just as if these classes got a bunch new
options.
Having this, I also extended the ContextGLTest to verify that the
Configuration and command-line options do what's expected because that
hadn't automated tests until now. The test is mostly a copy of what I
did for Vulkan already, nothing special. Additionally all
Platform*ApplicationTest executables gained a new --quiet option to
verify that the GL::Context::Configuration subset gets correctly passed
from the Application code, because that's something we can't really
verify in an automated way.
I don't see any use case for just specifying "all the flags", mainly
because the flags are mutually conflicting and whatnot. Also, these go
outdated *fast*.
There's four more new cursors and the _cursors array was too small.
At first I got confused because I thought the assertion on top is done
against the CursorMap, which didn't contain the Hidden cursors. So to
avoid confusing myself again in the future, I moved the assert after the
special cases and made both arrays the same size since it doesn't make
sense to have always-empty fields in there.
Similar change is done in Sdl2Application, and an assert is added to
avoid a nondescript crash if the window is not created yet.
Vladimír Vondruš