The infinite loop here doesn't really work at all for an application
process that expects to be able to exit. Check against
Core::EventLoop::current() to see if it's time to
exit, and return 0 from the thread function if so.
The thread will be joined in its destructor, which doesn't assert
anymore now that Thread is a jthread.
Just constructing one of these guys on the stack willy nilly will leak
the first reference to them. There might be other C_OBJECTs that have
public constructors, seems like a good place for some static analysis
checks :).
Force users to call the construct() method for it.
Since the vast majority of message boxes should be modal, require
the parent window to be passed in, which can be nullptr for the
rare case that they don't. By it being the first argument, the
default arguments also don't need to be explicitly stated in most
cases, and it encourages passing in a parent window handle.
Fix up several message boxes that should have been modal.
During app teardown, the Application object may be destroyed before
something else, and so having Application::the() return a reference was
obscuring the truth about its lifetime.
This patch makes the API more honest by returning a pointer. While
this makes call sites look a bit more sketchy, do note that the global
Application pointer only becomes null during app teardown.
This commit adds some actions for creating and cycling through tracks.
set_octave_and_ensure_note_change() was refactored to allow switching
tracks to implement the same behaviour.
KnobsWidget is getting pretty bad.
This commit adds multi-track functionality without exposing it to the
user.
All I really did was rename AudioEngine to Track and allow more than one
Track in TrackManager. A lot of the changes are just changing widgets to
take a TrackManager and use current_track().
The TrackManager creates Tracks and gives them a read-only reference to
the global time value. When the TrackManager wants to fill a sample in
the buffer (in fill_buffer()), it calls fill_sample() on each Track.
The delay code is slightly different - a Track will fill its
m_delay_buffer with the sample it just created rather than the most
recent sample in the buffer (which used to be the same thing).
TrackManager manages the current octave.
Other than those few things, this is a pretty basic separation of
concerns.
This allows us to construct menus in a more natural way:
auto& file_menu = menubar->add_menu("File");
file_menu.add_action(...);
Instead of the old way:
auto file_menu = GUI::Menu::construct();
file_menu->add_action(...);
menubar->add_menu(file_menu);
This patch allows roll notes to be of different sizes. This necessitates
a new internal representation of time. BPM and time signatures are
mostly implemented but not exposed.
Roll notes are now sample-accurate and the grid is aligned to 60 BPM
4/4. The roll is divided by the time signature raised to some power of
2, giving the musical divisions of (in the case of 4/4) 16, 32, 64 etc.
Before, our timing was derived from the buffer size and we relied on
that to implement delay. Delay has been rewritten to be sample-granular.
It's now exposed as the proper "divisions of a beat".
Something to be wary of is that the last buffer in the loop is also used
for the start of the next loop. In other words, we loop mid-buffer. This
means we write WAVs with a tiny bit of silence due to breaking the loop
after filling half a buffer.
The data structure for the roll is an array of SinglyLinkedLists of
RollNotes. Separating by pitch (via the array layout) makes insertion
much simpler and faster. Using sorted lists (and thus
SinglyLinkedListIterators) to do lookups is very quick as you know the
sample of the next note and can just compare it to the current sample. I
implemented this with HashMaps and the cost of lookups was abysmal. I
also tried a single SinglyLinkedList and the insertion code got even
more complicated than it already is.
I started adding things to a Draw namespace, but it somehow felt really
wrong seeing Draw::Rect and Draw::Bitmap, etc. So instead, let's rename
the library to LibGfx. :^)
I've been wanting to do this for a long time. It's time we start being
consistent about how this stuff works.
The new convention is:
- "LibFoo" is a userspace library that provides the "Foo" namespace.
That's it :^) This was pretty tedious to convert and I didn't even
start on LibGUI yet. But it's coming up next.
Goals:
- Switch to a more typical LibGUI arrangement
- Separate GUI (MainWidget) and audio (AudioEngine)
- Improve on existing features while retaining the same feature set
Improvements:
- Each GUI element is a separate widget
- The wave (WaveWidget) scales with the window
- The piano roll (RollWidget) scales horizontally and scrolls vertically
- The piano (KeysWidget) fits as many notes as possible
- The knobs (KnobsWidget) are now sliders
- All mouse and key events are handled in constant time
- The octave can be changed while playing notes
- The same note can be played with the mouse, keyboard and roll at the
same time, and the volume of the resulting note is scaled accordingly
- Note frequency constants use the maximum precision available in a
double
As suggested by Joshua, this commit adds the 2-clause BSD license as a
comment block to the top of every source file.
For the first pass, I've just added myself for simplicity. I encourage
everyone to add themselves as copyright holders of any file they've
added or modified in some significant way. If I've added myself in
error somewhere, feel free to replace it with the appropriate copyright
holder instead.
Going forward, all new source files should include a license header.
GEventLoop was just a dummy subclass of CEventLoop anyway. The only
thing it actually did was make sure a GWindowServerConnectionw was
instantiated. We now take care of that in GApplication instead.
CEventLoop is now non-virtual and a little less confusing. :^)
Okay, I've spent a whole day on this now, and it finally kinda works!
With this patch, CObject and all of its derived classes are reference
counted instead of tree-owned.
The previous, Qt-like model was nice and familiar, but ultimately also
outdated and difficult to reason about.
CObject-derived types should now be stored in RefPtr/NonnullRefPtr and
each class can be constructed using the forwarding construct() helper:
auto widget = GWidget::construct(parent_widget);
Note that construct() simply forwards all arguments to an existing
constructor. It is inserted into each class by the C_OBJECT macro,
see CObject.h to understand how that works.
CObject::delete_later() disappears in this patch, as there is no longer
a single logical owner of a CObject.
Now that we support more than 2 clients per shared buffer, we can use them
for window icons. I didn't do that previously since it would have made the
Taskbar process unable to access the icons.
This opens up some nice possibilities for programmatically generated icons.
This behavior and API was extremely counter-intuitive since our default
behavior was for applications to never exit after you close all of their
windows.
Now that we exit the event loop by default when the very last GWindow is
deleted, we don't have to worry about this.
As a consequence, move to use an explicit handshake() method rather than
calling virtuals from the constructor. This seemed to not bother
AClientConnection, but LibGUI crashes (rightfully) because of it.
