- Blur no longer expands the shadow geometry, but instead is clamped to
avoid overlapping drawing calls.
- Overlay now handles hit tests correctly with regards to the original
relative widget.
- Align was using an Into conversion that wasn't actually correct,
causing the contents to not actually get aligned in some situations.
Rather than clip or cause shadows to be weirdly shaped, the shape is
adjusted if the blur radius is too large for the given geometry. This
ensures the shadows always draw correctly, but it also forces shadows to
render their full blur + spread gradients.
Font families are now cached. There probably should be a mechanism to
refresh the cache, but we don't currently have a "signal" to notify us
when a font is installed. Presumably, this could be something that
fontdb would eventually add, rather than us building our own.
Closes#48
- On Linux, `fm-match` is used to query for the default fonts.
- DynamicComponents now have their own trait and can now be specified
with a constant or dynamic.
- Roboto Flex is now always loaded when the feature is enabled.
Overriding the default sans serif font prefers the overridden value,
then roboto, then the result of fc-match/fontdb's default.
- Button now supports background colors being set on a transparent
button.
Debug printing widgets was quite verbose. While developing a widget, you
often want to see a full debug printout, but this feature assumes that
debug printing a WidgetInstance should show a summary of the widget, not
a full debug printout containing cached glyph information of every
label.
By default, summarize just calls Debug, but this extra layer allows
widgets to provide a more condensed summary and exclude details like
caches.
Originally, adding dbg!() around the theme example's UI yielded a
whopping 20,324 lines of text. The summary code only prints 3,858
lines.
Installing a callback now returns a CallbackHandle. All map-style APIs
install this handle automatically on the created dynamic, which keeps
the callback installed until the dynamic is freed. All other APIs
return the handle for the caller to either call persist() or store
somewhere.
Now, the dynamic system can be used for application-long data with
almost no fear of leaking data due to how callbacks are being installed.
Technically cycles are still possible by moving clones into the
callbacks, so a WeakDynamic type might be worth exposing.
- Caching font family resolution to avoid scanning the database over and
over. The db should still be cached, but this makes repeated setting
free.
- into_switcher rename for Dynamic<WidgetInstance> to avoid conflicting
with Switchable::switcher()
- Dynamic debugging is less verbose
- IntoDynamic<Validation> for Result<T,E>
- Input no longer blinks cursor when disabled.
This expansion only triggers if the root widget measures larger than the
window's current size. We can't set a minimum size explicitly unless a
Resize widget is present, as we don't have enough knowledge to ensure
that the exact measurement we received was indeed the smallest layout in
any given direction. We only know that given the current constraints,
the returned measurement was the smallest possible. All future queries
will have to still be done again, as any change to the constraints could
impact the measured size.
Closes#60
Stepping in sliders is a compromise due to the flexibility of the
current slider implementation. I don't want to force types to implement
Add, and I don't like forcing types to require a Step (ie, what's the
appropriate value for f32 to specify as its next value?). Using a
percentage combined with lerp keeps the implementation fairly
straightfoward, although I remember experiencing this type of
configuration in another UI framework a long time ago and thinking it
was a little annoying to work with.
Ultimately, setting actual step boundaries can be done by customizing
the type that the slider is operating over. I feel like that's a much
more powerful design than I've experienced in previous frameworks, so
I'm hoping this percent step behavior is a reasonable compromise.
Inset outlines may not be the right approach, but it simplifies
potential alignment issues caused by insetting all filled background
drawing by half the width of the focus ring.
After trying to run Gooey again on my Mac for the first time in a few
weeks, I found that I ran into the Condvar issue again. Rather than
pasting AssertUnwindSafe in those files, I've both reported the
discrepency in unwind safety (rust-lang/rust#118009) and moved the
workaround into a type that only uses AssertUnwindsafe when compiling
for Apple.
The issue was that my last set of changes were causing the animations to
restart, causing the animation to keep being extended to another 150ms.
I think the only way for this to work is to switch to an event
mechanism to notify widgets once they've been invalidated. This event
could include a parameter stating whether it was a direct invalidation
or an invalidation due to another widget in the hierarchy. Button
doesn't really care about the rest of the hierarchy, it only cares about
its own state, and the cache key was including too many changes.
This also fixes some inconsistencies that arose when the focus widget
was "stuck" on a removed widget. Button previously handled it hackily in
a redraw function, but now Gooey handles it automatically without
needing to wait for a repaint.
Given the goal of this function, I'm not sure it can get more optimal
than this even with specialized data structures like KD-trees. The
problem is that we want all widgets that are hovered, not just some, and
that makes nearest neighbor useless.
The main optimizations here are simple:
- Group up all the render data we need in a single vec to help cache.
- Precompute the rect's extents to make the contains check at most 4
comparisons.
This had a noticable effect on the "wiggle the mouse frantically"
performance, where Gooey isn't actually repainting but is routing mouse
events.
Asserting condition on PercentBetween, Color lerping now works
correctly according to testing with gray shades, but due to rounding
errors, no unit test is being checked in at the moment.
I noticed that the new default theme settings were causing the "wrong"
text color to be picked. After reviewing why the contrast values were
the way they were, I reasoned that the less lightness, the less the
ColorSource contrast matters. So I've applied a multiplication of the
average ligntess between the two colors being compared.
Introducing two new colors:
- ColorTheme::color_dim, for dimmed/disabled primary colors
- SurfaceTheme::opaque_widget, for buttons.
In material design, a button's background color uses the Highest
Container role, which seems incorrect because then buttons wouldn't have
a different color when placed inside of the highest level container.
Rather than remove a container level, I added one more tone using the
neutral variant.
Other changes are just gut feelings to have a slightly richer dark
theme. I feel like material is a little muddy in dark mode.
I started making more changes, and I lost track of what was merge
related and what wasn't, so I committed the merge with what was staged,
but it probably didn't compile.
This is the remaining changes to get things back to a state I'm happy
with, mostly tweaking the constrast behavior.
Scroll was previously taking the graphics region as its control size as
opposed to the constraints. This was due to this code originally living
in redraw. This fixes scroll areas being able to scroll their contents
fully when sharing window space with other widgts.
This also means that if an animation is animating over discrete values
and the actual value has not changed, the Dynamic will no longer detect
a change because it's now using update instead of set.
