Update GLM to latest version (0.9.9.3). This includes GLM's change of matrices no longer default initializing to the identity matrix. This commit thus also includes the update of all of LearnOpenGL's code to reflect this: all matrices are now constructor-initialized to the identity matrix where relevant.

includes/glm/ext/quaternion_double_precision.hpp

@@ -0,0 +1,42 @@
+/// @ref ext_quaternion_double_precision
+/// @file glm/ext/quaternion_double_precision.hpp
+///
+/// @defgroup ext_quaternion_double_precision GLM_EXT_quaternion_double_precision
+/// @ingroup ext
+///
+/// Exposes double-precision floating point quaternion type with various precision in term of ULPs.
+///
+/// Include <glm/ext/quaternion_double_precision.hpp> to use the features of this extension.
+
+#pragma once
+
+// Dependency:
+#include "../detail/type_quat.hpp"
+
+#if GLM_MESSAGES == GLM_ENABLE && !defined(GLM_EXT_INCLUDED)
+#	pragma message("GLM: GLM_EXT_quaternion_double_precision extension included")
+#endif
+
+namespace glm
+{
+	/// @addtogroup ext_quaternion_double_precision
+	/// @{
+
+	/// Quaternion of double-precision floating-point numbers using high precision arithmetic in term of ULPs.
+	///
+	/// @see ext_quaternion_double_precision
+	typedef qua<double, lowp>		lowp_dquat;
+
+	/// Quaternion of medium double-qualifier floating-point numbers using high precision arithmetic in term of ULPs.
+	///
+	/// @see ext_quaternion_double_precision
+	typedef qua<double, mediump>	mediump_dquat;
+
+	/// Quaternion of high double-qualifier floating-point numbers using high precision arithmetic in term of ULPs.
+	///
+	/// @see ext_quaternion_double_precision
+	typedef qua<double, highp>		highp_dquat;
+
+	/// @}
+} //namespace glm
+