- Migrate to Xenial
 - Use whitelisted PPA instead of hardcoded deb URL
 - Only add X11 dependencies for X11 builds
 - Only set sudo on matrix entries

(cherry picked from commit ff7d42a)

The monitor content scale was missing from the monitor properties
paragraph.

(cherry picked from commit be71147)

Closes glfw#1470.

(cherry picked from commit 15af302)

(cherry picked from commit 42a3e44)

Fixes glfw#1480.

(cherry picked from commit 090b16b)

GLFW does not require the X11 extension libraries to build or run, so
only fail if the headers are unavailable.

(cherry picked from commit 5c5963f)

This replaces the manual ad-hoc version variables we have used since
CMake 2.x (and GLFW 2.x lite).

(cherry picked from commit a255e7a)

This replaces the earlier manual logic for dependent CMake options with
the cmake_dependent_option function from CMakeDependentOption.

(cherry picked from commit cd290f7)

(cherry picked from commit c0eabc5)

(cherry picked from commit 7dc36dd)

This was left over from an old workaround for Travis CI having an
ancient version of CMake pre-installed.

(cherry picked from commit 535491c)

(cherry picked from commit 0ccb690)

The X11_Xinput_* variables have been removed in recent CMake, leaving
only the X11_Xi_* set.

Related to glfw#1480.

(cherry picked from commit ceb16cb)

Related to glfw#1480.

(cherry picked from commit 57bf6b2)

(cherry picked from commit 700d1f2)

Most context related hint and attribute links had copypaste errors.
The GLFW_CONTEXT_RELEASE_BEHAVIOR and GLFW_CONTEXT_NO_ERROR attributes
did not have guide documentation.

(cherry picked from commit 3cf7645)

(cherry picked from commit feaa532)

Files built for Win32 must use C89 style declarations for compatibility
with VS 2010 and 2012, which are still supported by GLFW.

(cherry picked from commit 56aad76)

(cherry picked from commit 3fd4e79)

This should have been done when the WGL extension members were moved
from the context struct to the library struct.

(cherry picked from commit 22a6c02)

(cherry picked from commit bb6945a)

(cherry picked from commit 267e06a)

Disabled cursor mode interfered with some non-client actions.

(cherry picked from commit d0c3fa9)

(cherry picked from commit 84ec99b)

(cherry picked from commit e1d9e2b)

This will let higher-level projects override GLFW CMake options with
normal variables instead of having to use cache variables.

This means with CMake 3.13 and later you can now do:

set(GLFW_BUILD_TESTS ON)
add_subdirectory(path/to/glfw)

Instead of the more verbose:

set(GLFW_BUILD_TESTS ON CACHE BOOL "" FORCE)
add_subdirectory(path/to/glfw)

(cherry picked from commit 71e6ff3)

(cherry picked from commit 51bb76c)

Thank you, Travis CI, for reminding me that one cannot disable a header
with inclusion guards if it doesn't exist.

(cherry picked from commit 537ea4c)

Fixes glfw#1477.

(cherry picked from commit d834f01)

I found some tabs where there should be spaces for consistency.
Closes glfw#1496.

(cherry picked from commit 2db3b96)

(cherry picked from commit 0c6b505)

The EWMH feature detection atoms are now named and loaded the same way
as other X11 atoms.  Detection is now performed after all
non-conditional atoms have been loaded.  The EWMH detection now has
hopefully more readable comments.

(cherry picked from commit fad9896)

(cherry picked from commit 1f50853)

MSVC 2019 complains that the code at line 1744
(`GetMonitorInfo(window->monitor->win32.handle, &mi);`) can potentially
dereference a null pointer. The compiler is wrong in this case (it has
not spotted that `monitor` and `window->monitor` must be equal), but
I think it makes sense for our non-NULL test to be on the variable we
actually use rather than the one it was set from.

Related to glfw#1491.

(cherry picked from commit 1d62157)

This makes the Cocoa _glfwPlatformSetWindowMonitor consistent with its
X11 and Win32 counterparts.

(cherry picked from commit 3262c29)

(cherry picked from commit 3a5e99e)

Closes glfw#1513.

(cherry picked from commit e463e85)

The filter condition had ended up below the action.

(cherry picked from commit c6b95e3)

For users with multiple keyboard layouts configured, glfwGetKeyName
works fine only with the primary layout.  Switching layouts results in
changing the group index.  This commit querries the current group index
when initializing keyboard input and keeps track of any change to it.

As a result the scancode -> keyname mapping may change while the program
is running (needs to be documented).

Fixes glfw#1462.
Closes glfw#1528.

(cherry picked from commit 36f9080)

Related to glfw#1528.

(cherry picked from commit d232bcf)

Related to glfw#1528.

(cherry picked from commit 5bea122)

When both GLFW_INCLUDE_VULKAN and VK_USE_PLATFORM_WIN32_KHR were
defined, the GLFW header would define replacement versions of APIENTRY
and WINGDIAPI /before/ including the Vulkan header, which would include
windows.h, which (justifiably) defines APIENTRY and WINGDIAPI blindly.

Fixes glfw#1524.

(cherry picked from commit 773f449)

(cherry picked from commit 62b7fe8)

Closes glfw#1500.

(cherry picked from commit 84b1311)

(cherry picked from commit 7bd0af3)

This fixes the cursor enter event being emitted after the first cursor
position event on Windows.

Closes glfw#1490.

(cherry picked from commit 4f0b8b0)

[ci skip]

(cherry picked from commit a4d910b)

(cherry picked from commit 062a1c2)

The default symbol visibility was not set to hidden on macOS.

(cherry picked from commit b430bc4)

When using GLFW with CMake and installed GLFW binaries, `-DGLFW_DLL` is
passed on Linux, which should not happen.

Closes glfw#1530.

(cherry picked from commit d7fb01750ece465cfaae0493dfe3003da1e5f500)

(cherry picked from commit b0692d73ddb66461549cda0a009f867dab9de0fd)

The Bazel build system may define DEBUG when compiling on macOS, which
caused the glfwinfo test program to fail to build.

Fixes glfw#1537.

(cherry picked from commit 6abad2e)

The command buffer handle was not reset after being freed on window
resize, leading to a segfault when the stale handle was used.

(cherry picked from commit 65748fb)

This fixes a mismerge by @elmindreda in
65748fb.

(cherry picked from commit efda4af)

The window rect adjustment for content scale broke the initial, correct
maximization performed when creating the window with WS_MAXIMIZE.  This
switches to updating the restored rect instead of the current rect.

Fixes glfw#1499.
Closes glfw#1503.

(cherry picked from commit 711b969)

(cherry picked from commit 2c7ef5b)

(cherry picked from commit f764836)

(cherry picked from commit 4cc5d2e)

This makes key names per-key static strings for all supported platforms.

Fixes glfw#1200.

(cherry picked from commit 56ca0cb)

This way is perhaps also more readable for humans.

(cherry picked from commit cbf23e5)

This fixes the bug of video modes being discarded if they had
a different resolution but the same area as another mode.

Fixes glfw#1555.
Closes glfw#1556.

(cherry picked from commit 2777f6a)

Related to glfw#1556.

(cherry picked from commit 26aac53)

This fixes warnings emitted by Doxygen 1.8.16.

(cherry picked from commit aa8e8ab)

This makes glfwSetWindowSize use the top-left corner as anchor point
instead of the bottom-left corner.

Fixes glfw#1553.

(cherry picked from commit 7f02898)

Found via `codespell -q 3 -S ./deps -L fo,numer,te,uint,wille`

(cherry picked from commit 7105ff2)

The `size` member in the `GLFWgammaramp` struct is of type `unsigned int`, so the `for` loop iterating over it should also use the type `unsigned int`.

Closes glfw#1541.

(cherry picked from commit 243b1bc)

(cherry picked from commit 3ec8f4a)

This replaces some magic numbers with the corresponding winuser.h
provided macros and unifies how the MSB from Get*KeyState is tested.

(cherry picked from commit 3d2540c)

VS 2010 and MinGW are the two trickiest Windows environments so this
should hopefully catch more mistakes.  This is probably not the right
way to set up any of this, but it works for now.

This is also a first step towards building our release binaries with CI.

[skip travis]

(cherry picked from commit 1bd0a55)

The `name` member in the `_GLFWmonitorWayland` struct is used in two places. It is assigned the value from a variable of type `uint32_t` and is compared to another variable of type `uint32_t`, so `name` should also have the same type.

Closes glfw#1569.

(cherry picked from commit d252483)

(cherry picked from commit f71b55f)

This enables compilation as C99 where supported by the compiler.
A workaround with per-compiler hardcoded flags is used for CMake 3.0,
which does not support the C_STANDARD target property.

Fixes glfw#1560.
Closes glfw#1576.

(cherry picked from commit 8f852e0)

Related to glfw#1576.

(cherry picked from commit 545f409)

(cherry picked from commit 1e20218)

The `monitorCount` member in the `_GLFWlibrary` struct is of type `int`, so the `for` loop iterating over it should also use the type `int`.

Closes glfw#1572.

(cherry picked from commit 7749aac)

Closes glfw#1440.

(cherry picked from commit 3a37a08)

(cherry picked from commit 692754a)

(cherry picked from commit 36c5478)

Replaces `VERSION_GREATER_EQUAL` with `VERSION_EQUAL OR
VERSION_GREATER`.  `VERSION_GREATER_EQUAL` was only added in CMake 3.7.

Fixes glfw#1584.
Closes glfw#1585.

(cherry picked from commit 4d0ae4f)

Related to glfw#1585.

(cherry picked from commit 04f7f55)

The additional '>' characters were appended to the compiler option.

Related to glfw#1576.

(cherry picked from commit 33683ec)

This fixes the enabling of window decorations after creation.  Instead
of removing the _MOTIF_WM_HINTS property, we now set or unset the
MWM_DECOR_ALL bit of the decorations field.

Fixes glfw#1566.

(cherry picked from commit 5fc4c01)

KDE sometimes removes the Xft.dpi resource when it would be set to the
X11 default value of 96, causing GLFW to fall back to a value calculated
from the core display sizes in pixels and mm in a desktop environment
that supports Xft.dpi.

This moves to a hardcoded fallback value of 96 on the assumption that
there are more people running KDE with 96 DPI than there are people
running desktop environments that do not support Xft.dpi.

All of this is terrible please send help.

Fixes glfw#1578.

(cherry picked from commit 7529446)

Fixes glfw#1571.

(cherry picked from commit 28d8507)

This is a temporary fix while we determine if the color key workaround
should be removed completely.  See issue thread for discussion.

Related to glfw#1512.

(cherry picked from commit c88ee1c)

The SDL2 2.0.5+ controller GUID 03000000790000000600000000000000 matches
many devices with different layouts and element counts but with the same
chipset.  This issue is still being resolved upstream.  In the meantime
this removes those mappings from GLFW to avoid confusion and errors.

SDL upstream issue: https://bugzilla.libsdl.org/show_bug.cgi?id=4545

Related to glfw#1583.

(cherry picked from commit b69fb99)

Closes glfw#1595.

(cherry picked from commit c5f1ca3)

(cherry picked from commit 20e522c)

Related to glfw#1593.

(cherry picked from commit bc5523e)

Closes glfw#1593.

(cherry picked from commit 70a3104)

(cherry picked from commit 953a1c3)

(cherry picked from commit 94cb034)

On macOS a destroyed window remained on screen until the next time
events were processed.  This makes the behavior more consistent with
other platforms.

Fixes glfw#1412.

(cherry picked from commit c819f27)

This fixes a warning due to us using bare ids and
NSCollectionLayoutAnchor having a message with the same name.

(cherry picked from commit bac15f9)

Readability fix, does not affect generated code.

(cherry picked from commit bd45201)