The problem to fix is that the arrow drawn by draw_arrow1() in src/fl_symbols.cxx
displays a faint clear line between the stem and head of the arrow with the Cairo
graphics driver.
This occurs because draw_arrow1() draws the arrow in 2 steps (a rectangle +
a triangle) and the Cairo driver is configured to use antialiasing when filling
polygons. The antialiasing produces the faint line between stem and head.
Why does draw_arrow1() draw a rectangle + a triangle rather than a
7-vertex polygon? That's because the X11 graphics driver fails with its polygon-
drawing function when the polygon is also rotated: the polygon is drawn
empty.
We want to keep using antialiasing under Cairo for polygons because
the result is better with non horizontal/vertical polygon edges.
This implementation changes function draw_arrow1() which draws
the arrow as a 7-vertex filled polygon except when the graphics driver
returns false for its virtual member function can_fill_non_convex_polygon().
In that situation, draw_arrow1() draws, as before, a rectangle + a triangle.
The new, virtual member function can_fill_non_convex_polygon() returns
true except for the X11 graphics driver. Therefore, draw_arrow1() is effectively
unchanged under the X11 driver.
For a very simple test program see STR 3222, File 2:
https://www.fltk.org/strfiles/3222/value_slider.cxx
Known issues:
- range checking of the value box size is very limited
- resizing the widget does not change the value box size
- drawing issues are possible if the value box size and the widget
size are not compatible (e.g. value box too large)
"Arrows" in widgets are those GUI elements mostly represented by
triangles pointing in a particular direction as in scrollbars,
choice widgets, some menus, valuators and Fl_Counter widgets.
The code has been simplified and standardized such that all these
GUI elements are drawn identically per FLTK scheme.
Widget authors no longer need to write code to calculate arrow sizes
and draw polygons etc.
Different schemes can and do implement different drawing functions.
Todo: see comments "FIXME_ARROW" in src/Fl_Menu_Button.cxx and
src/Fl_Menu.cxx
Selecting a text range programmatically would not sync
some variables with the actual selection. This also fixes
a crash bug in macOS when dragging text that was
selected by buffer()->select() only.
While adding the docs, noticed some things that need modification
for proper public use. These are highlighted as \todo items and
\warning items, which will be fixed in a separate commit forthcoming. -erco
Started with @wcout's patch (see Issue #153, STR 3290), modified
patch for CMP conformance, restructured methods, and finally:
- used Fl_Int_Vector instead of 'fl_data_container'
- removed fl_data_container
Thanks for the working patch to @wcout.
Both bugs can happen if a widget doesn't have an associated window()
or in similar situations. These fixes returns NULL to prevent crashes.
Bugs observed in special test scenarios, not real-life programs.
This fixes the scrollbar order when children are inserted rather than
at some indefinite time later, for instance in draw().
This commit will very likely make Fl_Scroll::fix_scrollbar_order()
obsolete but this method is kept as is for tests and verification.
If the resizable() of an Fl_Group is deleted it is automatically
removed from the group.
New: to avoid dangling resizable() pointers the resizable widget
is set to the group itself.
Some functions didn't document the handling of arguments properly,
particularly Fl::has_timeout() and Fl::remove_timeout().
This is now fixed by documenting the correct behavior that was
preserved (re-implemented) from FLTK 1.3.x in the new class Fl_Timeout.
Unfortunately there have been some inconsistencies (likely unexpected
behavior) which have been preserved and which are now documented.
Doxygen uses a <div> with center alignment for images, hence
adding HTML center alignment for image is not only superfluous
but was also inconsistent.
Also removed some more unnecessary HTML tags, updated copyright year.
This commit allows to switch between FL_DOUBLE / FL_SINGLE modes
in widget-containing GL3 windows.
Demo program examples/OpenGL3test is modified to show FLTK widgets
even if the platform does not support OpenGL 3.
Under non-macOS platforms, the key is to call glUseProgram(0); after having used OpenGL 3
which allows to then use OpenGL 1 and draw FLTK widgets over the OpenGL3 scene.
Under macOS, this is impossible because macOS GL3 contexts are not compatible
with GL1. The solution implemented here is to create an additional Fl_Gl_Window
placed above and sized as the GL3-based window, to give it a non opaque,
GL1-based context, and to put the FLTK widgets in that additional window.
Function fl_read_image() obliges to keep a minimal use of global variable fl_window
also under the Wayland platform, even if its type (Window) makes little sense for
the hybrid library because it has its X11 value (given by X11/X.h) which is not
meaningful for the Wayland leg of the hybrid platform.
Virtual member function Fl_Surface_Device::as_image_surface() becomes useless.
1) add Wayland code that prevent subwindows from leaking outside their parent.
This does not cover GL subwindows.
2) add macOS code that prevent GL subwindows from leaking outside their parent.
This fixes issue #494 for the macOS platform.
N.B.: Wayland GL subwindows are not prevented from leaking because no solution
that would not require any change in client applications was found. Code that
would cover Wayland GL subwindows but would require client applications to always
use the FL_ALPHA flag is included in this commit in commented out form.
