0.9.9.5 Release

Builds/MSVC/Installer/Product.wxs

@@ -170,9 +170,6 @@
       <Component Id="Qt5Gui.dll" Guid="7f93dcc2-55db-4920-83d7-e06c23f7719a">
         <File Id="Qt5Gui.dll" Source="$(env.QTDIR)\bin\Qt5Gui.dll" KeyPath="yes" Checksum="yes" ProcessorArchitecture="x64"/>
       </Component>
-      <Component Id="Qt5OpenGL.dll" Guid="dad0a1ca-366c-4df8-aaa1-ae0de31f157a">
-        <File Id="Qt5OpenGL.dll" Source="$(env.QTDIR)\bin\Qt5OpenGL.dll" KeyPath="yes" Checksum="yes" ProcessorArchitecture="x64"/>
-      </Component>
       <Component Id="Qt5Widgets.dll" Guid="006bb2f1-7a38-426f-ba2c-5196d1d6c24d">
         <File Id="Qt5Widgets.dll" Source="$(env.QTDIR)\bin\Qt5Widgets.dll" KeyPath="yes" Checksum="yes" ProcessorArchitecture="x64"/>
       </Component>

Data/Version History.txt

@@ -1,6 +1,38 @@
+0.9.9.5 03/04/2016
+--User changes
+	-None.
+	
+--Bug fixes
+	-Fix crash when using Lanczos2 filter with color curves due to negative numbers. Fix by passing abs() of first argument to pow() in Renderer::GammaCorrection().
+	-Fix crash in hexes with SP. Rounding error caused out of bounds index.
+	-Do not reset locked affine scale on new flame or undo.
+	-Gracefully recover from reading a bad palette file.
+	-Gracefully recover from reading flame file with unknown params.
+	-Available memory size checking in the final render dialog was accidentally removed during the multi-gpu work last year. Add it back in.
+
+--Code changes
+	-Migrate to Qt 5.5.1
+	-Remove dependency projects from the solution. Move to a project design where they are built once via command line and stored in a /Deps folder.
+		--Because of this, libpng is built as a static lib instead of a dll because that's what its command line makefile supports.
+	--Remove libpng16.dll from Wix installer.
+	-Remove all configs except for debug and release.
+	-Remove all platforms except for x64.
+	-Add #define __TBB_NO_IMPLICIT_LINKAGE 1 to prevent tbb from looking for tbb_debug.lib by default on debug builds.
+	-Put Ember.rc in a filter folder in the project.
+	-Place pragma warning disable statements in PCH files rather than in projects. This makes it easier when using Qt Creator.
+	-Move all resource files to the MSVC folder.
+	-Set all targets in QtCreator project to their lowercase names to be more *nix friendly. Set -l link statements to use these new lowercase names for ember and embercl.
+	-Rework Qt projects to favor shadow builds.
+	-Remove the symlinks.sh file and all references to it. It was never needed. Instead, just specify multiple include and link paths in the .pro files.
+	-Change WIN32 to _WIN32.
+	-Fix a few code warnings.
+	-Make usage of VarFuncs more consistent by moving some global functions there.
+	-Make CircleTrans1 a little safer by calling Zeps() on m_Sc during precalc.
+	-Major work on Qt Creator builds.
+
 0.9.9.4 02/21/2016
 --User changes
-	-Allow for stopping the renderer in the main window. This makes is more efficient when entering many parameters, such as when following a tutorial.
+	-Allow for stopping the renderer in the main window. This makes it more efficient when entering many parameters, such as when following a tutorial.
 	-Add support for new variations: erf, gamma, jac_cn, jac_dn, jac_sn, logDB, pressure_wave, pRose3D, splits3D, w, waves2b, x, xerf, y, z.
 	-Inform user of the start and stop of file parsing in EmberAnimate because the files could potentially be very large.
 	-Move the follwing fields to a new table called Animation: Interpolation, Affine Interpolation, Temporal Samples, Temporal Filter Width, Temporal Filter Type.

Source/Ember/Utils.h

@@ -442,6 +442,38 @@ static inline T Clamp(T val, T min, T max)
 		return val;
 }
 
+template <>
+#ifdef _WIN32
+	static
+#endif
+float Clamp<float>(float val, float min, float max)
+{
+	if (val < min)
+		return min;
+	else if (val > max)
+		return max;
+	else if (!std::isfinite(val))
+		return min;
+	else
+		return val;
+}
+
+template <>
+#ifdef _WIN32
+	static
+#endif
+double Clamp<double>(double val, double min, double max)
+{
+	if (val < min)
+		return min;
+	else if (val > max)
+		return max;
+	else if (!std::isfinite(val))
+		return min;
+	else
+		return val;
+}
+
 /// <summary>
 /// Clamp and return a value to be greater than or equal to a specified minimum and less than
 /// or equal to a specified maximum. If lesser, the value is fmod(val - min, max - min). If greater,
@@ -458,6 +490,8 @@ static inline T ClampMod(T val, T min, T max)
 		return min + fmod(val - min, max - min);
 	else if (val > max)
 		return max - fmod(max - val, max - min);
+	else if (!std::isfinite(val))
+		return min;
 	else
 		return val;
 }
@@ -477,6 +511,34 @@ static inline void ClampRef(T& val, T min, T max)
 		val = max;
 }
 
+template <>
+#ifdef _WIN32
+	static
+#endif
+void ClampRef<float>(float& val, float min, float max)
+{
+	if (val < min)
+		val = min;
+	else if (val > max)
+		val = max;
+	else if (!std::isfinite(val))
+		val = min;
+}
+
+template <>
+#ifdef _WIN32
+	static
+#endif
+void ClampRef<double>(double& val, double min, double max)
+{
+	if (val < min)
+		val = min;
+	else if (val > max)
+		val = max;
+	else if (!std::isfinite(val))
+		val = min;
+}
+
 /// <summary>
 /// Similar to Clamp(), but clamps a reference value in place rather than returning.
 /// </summary>
@@ -489,6 +551,26 @@ static inline void ClampLteRef(T& val, T lte)
 		val = lte;
 }
 
+template <>
+#ifdef _WIN32
+	static
+#endif
+void ClampLteRef<float>(float& val, float lte)
+{
+	if (val > lte || !std::isfinite(val))
+		val = lte;
+}
+
+template <>
+#ifdef _WIN32
+	static
+#endif
+void ClampLteRef<double>(double& val, double lte)
+{
+	if (val > lte || !std::isfinite(val))
+		val = lte;
+}
+
 /// <summary>
 /// Clamp and return a value to be greater than or equal to a specified value.
 /// Useful for ensuring something is not less than zero.
@@ -502,6 +584,30 @@ static inline T ClampGte(T val, T gte)
 	return (val < gte) ? gte : val;
 }
 
+template <>
+#ifdef _WIN32
+	static
+#endif
+float ClampGte<float>(float val, float gte)
+{
+	if (val < gte || !std::isfinite(val))
+		return gte;
+	else
+		return val;
+}
+
+template <>
+#ifdef _WIN32
+	static
+#endif
+double ClampGte<double>(double val, double gte)
+{
+	if (val < gte || !std::isfinite(val))
+		return gte;
+	else
+		return val;
+}
+
 /// <summary>
 /// Similar to Clamp(), but clamps a reference value in place rather than returning.
 /// </summary>
@@ -514,6 +620,26 @@ static inline void ClampGteRef(T& val, T gte)
 		val = gte;
 }
 
+template <>
+#ifdef _WIN32
+	static
+#endif
+void ClampGteRef<float>(float& val, float gte)
+{
+	if (val < gte || !std::isfinite(val))
+		val = gte;
+}
+
+template <>
+#ifdef _WIN32
+	static
+#endif
+void ClampGteRef<double>(double& val, double gte)
+{
+	if (val < gte || !std::isfinite(val))
+		val = gte;
+}
+
 /// <summary>
 /// Thin wrapper around a call to ClampGte() with a gte value of zero.
 /// </summary>
@@ -546,28 +672,6 @@ static inline T Round(T r)
 	return (r > 0) ? static_cast<T>(Floor<T>(r + T(0.5))) : ceil(r - T(0.5));
 }
 
-/// <summary>
-/// Special rounding for certain variations, gotten from Apophysis.
-/// </summary>
-/// <param name="x">The value to round</param>
-/// <returns>The rounded value</returns>
-static inline float LRint(float x)
-{
-	int temp = (x >= 0 ? static_cast<int>(x + 0.5f) : static_cast<int>(x - 0.5f));
-	return static_cast<float>(temp);
-}
-
-/// <summary>
-/// Special rounding for certain variations, gotten from Apophysis.
-/// </summary>
-/// <param name="x">The value to round</param>
-/// <returns>The rounded value</returns>
-static inline double LRint(double x)
-{
-	glm::int64_t temp = (x >= 0 ? static_cast<int64_t>(x + 0.5) : static_cast<int64_t>(x - 0.5));
-	return static_cast<double>(temp);
-}
-
 /// <summary>
 /// Never really understood what this did.
 /// </summary>
@@ -584,30 +688,6 @@ static inline T Round6(T r)
 	return static_cast<T>(1e-6 * static_cast<int>(r + T(0.5)));
 }
 
-/// <summary>
-/// Return -1 if the value is less than 0, 1 if it's greater and
-/// 0 if it's equal to 0.
-/// </summary>
-/// <param name="v">The value to inspect</param>
-/// <returns>-1, 0 or 1</returns>
-template <typename T>
-static inline T Sign(T v)
-{
-	return (v < 0) ? static_cast<T>(-1) : (v > 0) ? static_cast<T>(1) : static_cast<T>(0);
-}
-
-/// <summary>
-/// Return -1 if the value is less than 0, 1 if it's greater.
-/// This differs from Sign() in that it doesn't return 0.
-/// </summary>
-/// <param name="v">The value to inspect</param>
-/// <returns>-1 or 1</returns>
-template <typename T>
-static inline T SignNz(T v)
-{
-	return (v < 0) ? static_cast<T>(-1) : static_cast<T>(1);
-}
-
 /// <summary>
 /// Return the square of the passed in value.
 /// This is useful when the value is a result of a computation
@@ -622,16 +702,16 @@ static inline T Sqr(T t)
 }
 
 /// <summary>
-/// Taking the square root of numbers close to zero is dangerous.  If x is negative
-/// due to floating point errors, it can return NaN results.
+/// Return the cube of the passed in value.
+/// This is useful when the value is a result of a computation
+/// rather than a fixed number. Otherwise, use the CUBE macro.
 /// </summary>
+/// <param name="v">The value to cube</param>
+/// <returns>The cubed value</returns>
 template <typename T>
-static inline T SafeSqrt(T x)
+static inline T Cube(T t)
 {
-	if (x <= 0)
-		return 0;
-
-	return std::sqrt(x);
+	return t * t * t;
 }
 
 template <typename T>
@@ -658,83 +738,6 @@ double SafeTan<double>(double x)
 	return std::tan(x);
 }
 
-/// <summary>
-/// If r < EPS, return 1 / r.
-/// Else, return q / r.
-/// </summary>
-/// <param name="q">The numerator</param>
-/// <param name="r">The denominator</param>
-/// <returns>The quotient</returns>
-template <typename T>
-static inline T SafeDivInv(T q, T r)
-{
-	if (r < EPS)
-		return 1 / r;
-
-	return q / r;
-}
-
-/// <summary>
-/// Return the cube of the passed in value.
-/// This is useful when the value is a result of a computation
-/// rather than a fixed number. Otherwise, use the CUBE macro.
-/// </summary>
-/// <param name="v">The value to cube</param>
-/// <returns>The cubed value</returns>
-template <typename T>
-static inline T Cube(T t)
-{
-	return t * t * t;
-}
-
-/// <summary>
-/// Return the hypotenuse of the passed in values.
-/// </summary>
-/// <param name="x">The x distance</param>
-/// <param name="y">The y distance</param>
-/// <returns>The hypotenuse</returns>
-template <typename T>
-static inline T Hypot(T x, T y)
-{
-	return std::sqrt(SQR(x) + SQR(y));
-}
-
-/// <summary>
-/// Spread the values.
-/// </summary>
-/// <param name="x">The x distance</param>
-/// <param name="y">The y distance</param>
-/// <returns>The spread</returns>
-template <typename T>
-static inline T Spread(T x, T y)
-{
-	return Hypot<T>(x, y) * ((x) > 0 ? 1 : -1);
-}
-
-/// <summary>
-/// Unsure.
-/// </summary>
-/// <param name="x">The x distance</param>
-/// <param name="y">The y distance</param>
-/// <returns>The powq4</returns>
-template <typename T>
-static inline T Powq4(T x, T y)
-{
-	return std::pow(std::abs(x), y) * SignNz(x);
-}
-
-/// <summary>
-/// Unsure.
-/// </summary>
-/// <param name="x">The x distance</param>
-/// <param name="y">The y distance</param>
-/// <returns>The powq4c</returns>
-template <typename T>
-static inline T Powq4c(T x, T y)
-{
-	return y == 1 ? x : Powq4(x, y);
-}
-
 /// <summary>
 /// Return EPS if the passed in value was zero, else return the value.
 /// </summary>
@@ -760,66 +763,6 @@ static inline T Lerp(T a, T b, T p)
 	return a + (b - a) * p;
 }
 
-/// <summary>
-/// Thin wrapper around a call to modf that discards the integer portion
-/// and returns the signed fractional portion.
-/// </summary>
-/// <param name="v">The value to retrieve the signed fractional portion of.</param>
-/// <returns>The signed fractional portion of v.</returns>
-template <typename T>
-static inline T Fabsmod(T v)
-{
-	T dummy;
-	return modf(v, &dummy);
-}
-
-/// <summary>
-/// Unsure.
-/// </summary>
-/// <param name="p">Unsure.</param>
-/// <param name="amp">Unsure.</param>
-/// <param name="ph">Unsure.</param>
-/// <returns>Unsure.</returns>
-template <typename T>
-static inline T Fosc(T p, T amp, T ph)
-{
-	return T(0.5) - std::cos(p * amp + ph) * T(0.5);
-}
-
-/// <summary>
-/// Unsure.
-/// </summary>
-/// <param name="p">Unsure.</param>
-/// <param name="ph">Unsure.</param>
-/// <returns>Unsure.</returns>
-template <typename T>
-static inline T Foscn(T p, T ph)
-{
-	return T(0.5) - std::cos(p + ph) * T(0.5);
-}
-
-/// <summary>
-/// Log scale from Apophysis.
-/// </summary>
-/// <param name="x">The value to log scale</param>
-/// <returns>The log scaled value</returns>
-template <typename T>
-static inline T LogScale(T x)
-{
-	return x == 0 ? 0 : std::log((fabs(x) + 1) * T(M_E)) * SignNz(x) / T(M_E);
-}
-
-/// <summary>
-/// Log map from Apophysis.
-/// </summary>
-/// <param name="x">The value to log map</param>
-/// <returns>The log mapped value</returns>
-template <typename T>
-static inline T LogMap(T x)
-{
-	return x == 0 ? 0 : (T(M_E) + std::log(x * T(M_E))) * T(0.25) * SignNz(x);
-}
-
 /// <summary>
 /// Thin wrapper around calling xmlStrcmp() on an Xml tag to tell
 /// if its name is a given value.
@@ -877,23 +820,6 @@ static inline T NormalizeDeg180(T angle)
 	return a;
 }
 
-/// <summary>
-/// Put an angular measurement in degrees into the range of 0 - 360.
-/// </summary>
-/// <param name="angle">The angle to normalize</param>
-/// <returns>The normalized angle in a range of 0 - 360</returns>
-template <typename T>
-static inline T NormalizeDeg360(T angle)
-{
-	if (angle > 360 || angle < -360)
-		angle = fmod(angle, 360);
-
-	if (angle < 0)
-		angle += 360;
-
-	return angle;
-}
-
 /// <summary>
 /// Return a lower case copy of a string.
 /// </summary>
@@ -979,7 +905,7 @@ static inline T Arg(char* name, T def)
 {
 	char* ch;
 	T returnVal;
-#ifdef WIN32
+#ifdef _WIN32
 	size_t len;
 	errno_t err = _dupenv_s(&ch, &len, name);
 #else
@@ -1001,7 +927,7 @@ static inline T Arg(char* name, T def)
 			returnVal = def;
 	}
 
-#ifdef WIN32
+#ifdef _WIN32
 	free(ch);
 #endif
 	return returnVal;
@@ -1036,7 +962,7 @@ string Arg<string>(char* name, string def)
 {
 	char* ch;
 	string returnVal;
-#ifdef WIN32
+#ifdef _WIN32
 	size_t len;
 	errno_t err = _dupenv_s(&ch, &len, name);
 #else
@@ -1052,7 +978,7 @@ string Arg<string>(char* name, string def)
 	else
 		returnVal = string(ch);
 
-#ifdef WIN32
+#ifdef _WIN32
 	free(ch);
 #endif
 	return returnVal;