Fixes: https://todo.sr.ht/~eliasnaur/gio/567
Signed-off-by: Aman Karmani <[email protected]>

Window events are no longer asynchronous, so deadlocks are no longer
possible when calling Run.
...

The updates to the gioui.org/app package are described by #555.

Signed-off-by: Elias Naur <mail...

Magic incantations lifted from

https://thume.ca/2019/06/19/glitchless-metal-window-resizing/

S...

Signed-off-by: Elias Naur <[email protected]>

Current implementation relies on change event callback to return. However, if no file is selecte...

I had accidentally committed a modified program in its place.

Signed-off-by: Chris Waldon <chri...

Earlier js.CopyBytesToJS(value, b) was used on a blank destination array. As per go docs, it cop...

People often end up unkowingly using `go get -u` for updating modules,
however that's not safe t...

Signed-off-by: Egon Elbre <[email protected]>

Signed-off-by: Egon Elbre <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

StageEvent served only redundant purposes:

- To detect whether the window has focus. That is co...

Signed-off-by: Elias Naur <[email protected]>

A uniform type allows convenient nil checks and for future window
backends on platforms other th...

Signed-off-by: Elias Naur <[email protected]>

Instead of sending ViewEvents once at construction and once at destruction,
it's better to send ...

Signed-off-by: Elias Naur <[email protected]>

Notifications don't require a list of windows nor an app delegate.

Signed-off-by: Elias Naur <m...

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

Their absense didn't make a practical difference so far, but we're about
to refactor the macOS e...

While here, don't use trySetPrivateCursor for the public openHandCursor
and closedHandCursor.

S...

The dealloc method is where we're guaranteed the NSView is no longer
used anywhere.

Signed-off-...

This is mostly a refactor, but there are two user-visible effects:
- Window.NextEvent may be cal...

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

This commit updates the colorpicker to process events using the new event filters
API and also r...

Signed-off-by: Chris Waldon <[email protected]>

This commit updates the constraint debugger to use event filters. Sadly, this
appears to tickle ...

This commit makes package richtext compatible with the new usptream event
filtering while also u...

Gio upstream no longer supports this strategy for gathering graphics profiling
information, so t...

This commit eliminates the eventx package. With the new Gio event filtering API,
the spy and rel...

Signed-off-by: Chris Waldon <[email protected]>

Signed-off-by: Chris Waldon <[email protected]>

This commit updates all remaining outlay layouts to use the new API and
splits the Grid type's U...

We're about to need per-handler state related to neither pointer nor
key input. This change merg...

It's semantically problematic that a zero Kind matches Cancel, and
outweighs the downside of hav...

If the client asks for the focus to be set to a tag, allow it. There is a
check at the end of Ro...

The method no longer returns anything, and thus does not actually report
whether any events matc...

Replace the key.Filter.Target field with a Focus field that matches only
of the specified tag ha...

Similar to how events are processed one at a time, change Clickable to
report clicks one at a ti...

Refactor delivery of reset events to be resolved and delivered as part of
Source.Events. This is...

Widgets have themselves as tags, by convention, and so it's possible to
replace the per-widget F...

Change the semantics of commands to execute immediately. In cases where
execution of a command i...

Until now, every event has had a particular target. We're about to simplify
key event delivery t...

Replace the per-event event queues with a single queue of events, each
marked with the target ta...

We're about the refactor this quite subtle code, and the optimization
doesn't seem to carry its ...

Processing one event at a time allows a widget to execute commands after
the event that triggere...

New handlers receive reset events the first time Source.Event is called.
However, in case the fi...

Refactor the pointer and key filter unions into the handler state struct.
This is a preparation ...

Signed-off-by: Elias Naur <[email protected]>

This commit eliminates (*widget.Editor).Events() in favor of making
(*widget.Editor).Update() re...

We're about to replace the per-widget Focus methods with the client
executing FocusCmd themselve...

Now that event delivery can be interleaved with commands, tests can be
made more precise.

Signe...

Curiously, InvalidateCmd is probably the only command that is appropriate
to call during layout....

Signed-off-by: Elias Naur <[email protected]>

This change defers event routing from the time the event is queued until
the time Events is call...

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

Events are now delivered one at a time, and this change makes the
corresponding change to gestur...

Signed-off-by: Elias Naur <[email protected]>

With the introduction of filters, it is now possible to have one tag per
widget by convention. N...

The empty key.Filter.Name now means matching every key name. This is a
replacement for the previ...

Now that widgets by convention may be focused by issuing FocusCmd
directly, remove the now redun...

Now (*widget.Selectable).Update() returns whether the selection changed during
event processing,...

Make the internal changes to support fine-grained event delivery.

Signed-off-by: Elias Naur <ma...

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

This change gets rid of the event.Queue interface by replacing it with
input.Source values. Sour...

This change is required to to replace event.Queue with a concrete
input.Source.

Signed-off-by: ...

Signed-off-by: Elias Naur <[email protected]>

We're about to make Context.Queue a concrete type, and this change
replaces code that relies on ...

We're about to make the Queue field of FrameEvent (and layout.Context)
a concrete type. Remove t...

The input name better matches its purpose, in particular when we
introduce input.Source.

Signed...

4 billion semantic IDs should be enough for everyone.

Signed-off-by: Elias Naur <mail@eliasnaur...

It was a design mistake to make profiling data available to programs.
Rather, profiling should e...

Modeling focus change as an operation is awkward, because focus changes
logically happen during ...

Signed-off-by: Elias Naur <[email protected]>

We're about to replace key.InputOp with a filter; this change separates
the input hint into its ...

Signed-off-by: Elias Naur <[email protected]>

In the early days of Gio, FrameEvent was part of package app. It was
moved to package system to ...

They're only useful at the top-level event loop in combination with an
app.Window.

Signed-off-b...

Without the dependency, tests builds much faster.

Signed-off-by: Elias Naur <[email protected]>

We're about to replace the interface Queue with a concrete input.Source.
This change renames the...

Without the dependency, tests builds much faster.

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>

Also delete two tests that are no longer relevant.

Signed-off-by: Elias Naur <[email protected]>

Signed-off-by: Elias Naur <[email protected]>