• 13
• 14
• 27
• 9
• 9
• ### Similar Content

• By abarnes
Hello All!
I am currently pursuing a degree in video game programming, so far I have completed an intro to programming course and object oriented programming course. Both were taught using C++ as the programming langauge which I know is very popular for game development, but in these classes we do not actually do any game development. I would like to start to build my skills with C++ for game development as that is a common required thing for a job and am looking for ways to do this. Any recommendations such as books to read or youtube videos to watch will be greatly appreciated!
• By Orella
I'm having problems rotating GameObjects in my engine. I'm trying to rotate in 2 ways.
I'm using MathGeoLib to calculate maths in the engine.
First way: Rotates correctly around axis but if I want to rotate back, if I don't do it following the inverse order then rotation doesn't work properly.
e.g:
Rotate X axis 50 degrees, Rotate Y axis 30 degrees -> Rotate Y axis -50 degrees, Rotate X axis -30 degrees. Works.
Rotate X axis 50 degrees, Rotate Y axis 30 degrees -> Rotate X axis -50 degrees, Rotate Y axis -30 degrees. Doesn't.

Code:
void ComponentTransform::SetRotation(float3 euler_rotation) { float3 diff = euler_rotation - editor_rotation; editor_rotation = euler_rotation; math::Quat mod = math::Quat::FromEulerXYZ(diff.x * DEGTORAD, diff.y * DEGTORAD, diff.z * DEGTORAD); quat_rotation = quat_rotation * mod; UpdateMatrix();  } Second way: Starts rotating good around axis but after rotating some times, then it stops to rotate correctly around axis, but if I rotate it back regardless of the rotation order it works, not like the first way.

Code:
void ComponentTransform::SetRotation(float3 euler_rotation) { editor_rotation = euler_rotation; quat_rotation = math::Quat::FromEulerXYZ(euler_rotation.x * DEGTORAD, euler_rotation.y * DEGTORAD, euler_rotation.z * DEGTORAD); UpdateMatrix();  }
Rest of code:
#define DEGTORAD 0.0174532925199432957f void ComponentTransform::UpdateMatrix() { if (!this->GetGameObject()->IsParent()) { //Get parent transform component ComponentTransform* parent_transform = (ComponentTransform*)this->GetGameObject()->GetParent()->GetComponent(Component::CompTransform); //Create matrix from position, rotation(quaternion) and scale transform_matrix = math::float4x4::FromTRS(position, quat_rotation, scale); //Multiply the object transform by parent transform transform_matrix = parent_transform->transform_matrix * transform_matrix; //If object have childs, call this function in childs objects for (std::list<GameObject*>::iterator it = this->GetGameObject()->childs.begin(); it != this->GetGameObject()->childs.end(); it++) { ComponentTransform* child_transform = (ComponentTransform*)(*it)->GetComponent(Component::CompTransform); child_transform->UpdateMatrix(); } } else { //Create matrix from position, rotation(quaternion) and scale transform_matrix = math::float4x4::FromTRS(position, quat_rotation, scale); //If object have childs, call this function in childs objects for (std::list<GameObject*>::iterator it = this->GetGameObject()->childs.begin(); it != this->GetGameObject()->childs.end(); it++) { ComponentTransform* child_transform = (ComponentTransform*)(*it)->GetComponent(Component::CompTransform); child_transform->UpdateMatrix(); } } } MathGeoLib: Quat MUST_USE_RESULT Quat::FromEulerXYZ(float x, float y, float z) { return (Quat::RotateX(x) * Quat::RotateY(y) * Quat::RotateZ(z)).Normalized(); } Quat MUST_USE_RESULT Quat::RotateX(float angle) { return Quat(float3(1,0,0), angle); } Quat MUST_USE_RESULT Quat::RotateY(float angle) { return Quat(float3(0,1,0), angle); } Quat MUST_USE_RESULT Quat::RotateZ(float angle) { return Quat(float3(0,0,1), angle); } Quat(const float3 &rotationAxis, float rotationAngleRadians) { SetFromAxisAngle(rotationAxis, rotationAngleRadians); } void Quat::SetFromAxisAngle(const float3 &axis, float angle) { assume1(axis.IsNormalized(), axis); assume1(MATH_NS::IsFinite(angle), angle); float sinz, cosz; SinCos(angle*0.5f, sinz, cosz); x = axis.x * sinz; y = axis.y * sinz; z = axis.z * sinz; w = cosz; } Any help?
Thanks.
• By owenjr
Hi there!
I am trying to implement a basic AI for a Turrets game in SFML and C++ and I have some problems.
This AI follows some waypoints stablished in a Bezier Courve.
In first place, this path was followed only by one enemy. For this purpose, the enemy has to calculate his distance between his actual position
to the next waypoint he has to pick.
If the distance is less than a specific value we stablish, then, we get to the next point. This will repeat until the final destination is reached. (in the submitting code, forget about the var m_go)

Okay, our problem gets when we spawn several enemies and all have to follow the same path, because it produces a bad visual effect (everyone gets upside another).
In order to solve this visual problem, we have decided to use a repulsion vector. The calculus gets like this:

As you can see, we calculate the repulsion vector with the inverse of the distance between the enemy and his nearest neighbor.
Then, we get it applying this to the "theorical" direction, by adding it, and we get a resultant, which is the direction that
our enemy has to follow to not "collide" with it's neighbors. But, our issue comes here:

The enemys get sepparated in the middle of the curve and, as we spawn more enemys, the speed of all of them increases dramatically (including the enemies that don't calculate the repuslion vector).
1 - Is it usual that this sepparation occours in the middle of the trajectory?
2 - Is it there a way to control this direction without the speed getting affected?
3 - Is it there any alternative to this theory?

I submit the code below (There is a variable in Spanish [resultante] which it means resultant in English):

if (!m_pathCompleted) { if (m_currentWP == 14 && m_cambio == true) { m_currentWP = 0; m_path = m_pathA; m_cambio = false; } if (m_neighbors.size() > 1) { for (int i = 0; i < m_neighbors.size(); i++) { if (m_enemyId != m_neighbors[i]->GetId()) { float l_nvx = m_neighbors[i]->GetSprite().getPosition().x - m_enemySprite.getPosition().x; float l_nvy = m_neighbors[i]->GetSprite().getPosition().y - m_enemySprite.getPosition().y; float distance = std::sqrt(l_nvx * l_nvx + l_nvy * l_nvy); if (distance < MINIMUM_NEIGHBOR_DISTANCE) { l_nvx *= -1; l_nvy *= -1; float l_vx = m_path[m_currentWP].x - m_enemySprite.getPosition().x; float l_vy = m_path[m_currentWP].y - m_enemySprite.getPosition().y; float l_resultanteX = l_nvx + l_vx; float l_resultanteY = l_nvy + l_vy; float l_waypointDistance = std::sqrt(l_resultanteX * l_resultanteX + l_resultanteY * l_resultanteY); if (l_waypointDistance < MINIMUM_WAYPOINT_DISTANCE) { if (m_currentWP == m_path.size() - 1) { std::cout << "\n"; std::cout << "[GAME OVER]" << std::endl; m_go = false; m_pathCompleted = true; } else { m_currentWP++; } } if (l_waypointDistance > MINIMUM_WAYPOINT_DISTANCE) { l_resultanteX = l_resultanteX / l_waypointDistance; l_resultanteY = l_resultanteY / l_waypointDistance; m_enemySprite.move(ENEMY_SPEED * l_resultanteX * dt, ENEMY_SPEED * l_resultanteY * dt); } } else { float vx = m_path[m_currentWP].x - m_enemySprite.getPosition().x; float vy = m_path[m_currentWP].y - m_enemySprite.getPosition().y; float len = std::sqrt(vx * vx + vy * vy); if (len < MINIMUM_WAYPOINT_DISTANCE) { if (m_currentWP == m_path.size() - 1) { std::cout << "\n"; std::cout << "[GAME OVER]" << std::endl; m_go = false; m_pathCompleted = true; } else { m_currentWP++; } } if (len > MINIMUM_WAYPOINT_DISTANCE) { vx = vx / len; vy = vy / len; m_enemySprite.move(ENEMY_SPEED * vx * dt, ENEMY_SPEED * vy * dt); } } } } } else { float vx = m_path[m_currentWP].x - m_enemySprite.getPosition().x; float vy = m_path[m_currentWP].y - m_enemySprite.getPosition().y; float len = std::sqrt(vx * vx + vy * vy); if (len < MINIMUM_WAYPOINT_DISTANCE) { if (m_currentWP == m_path.size() - 1) { std::cout << "\n"; std::cout << "[GAME OVER]" << std::endl; m_go = false; m_pathCompleted = true; } else { m_currentWP++; } } if (len > MINIMUM_WAYPOINT_DISTANCE) { vx = vx / len; vy = vy / len; m_enemySprite.move(ENEMY_SPEED * vx * dt, ENEMY_SPEED * vy * dt); } } }
¡¡Thank you very much in advance!!

• Overview
Welcome to the 2D UFO game guide using the Orx Portable Game Engine. My aim for this tutorial is to take you through all the steps to build a UFO game from scratch.
The aim of our game is to allow the player to control a UFO by applying physical forces to move it around. The player must collect pickups to increase their score to win.
I should openly acknowledge that this series is cheekily inspired by the 2D UFO tutorial written for Unity.
It makes an excellent comparison of the approaches between Orx and Unity. It is also a perfect way to highlight one of the major parts that makes Orx unique among other game engines, its Data Driven Configuration System.
You'll get very familiar with this system very soon. It's at the very heart of just about every game written using Orx.
If you are very new to game development, don't worry. We'll take it nice and slow and try to explain everything in very simple terms. The only knowledge you will need is some simple C++.
I'd like say a huge thank you to FullyBugged for providing the graphics for this series of articles.

What are we making?
Visit the video below to see the look and gameplay of the final game:
Getting Orx
The latest up to date version of Orx can be cloned from github and set up with:
git clone https://github.com/orx/orx.git After cloning, an $ORX environment variable will be created automatically for your system which will help with making game projects much easier. It will also create several IDE projects for your operating system: Visual Studio, Codelite, Code::Blocks, and gmake. These Orx projects will allow you to compile the Orx library for use in your own projects. And the$ORX environment variable means that your projects will know where to find the Orx library.
For more details on this step, visit http://orx-project.org/wiki/en/tutorials/cloning_orx_from_github at the Orx learning wiki.
Setting up a 2D UFO Project
Now the you have the Orx libraries cloned and compiled, you will need a blank project for your game. Supported options are: Visual Studio, CodeLite, Code::Blocks, XCode or gmake, depending on your operating system.
Once you have a game project, you can use it to work through the steps in this tutorial.
Orx provides a very nice system for auto creating game projects for you. In the root of the Orx repo, you will find either the init.bat (for Windows) or init.sh (Mac/Linux) command.
Create a project for our 2D game from the command line in the Orx folder and running:
init c:\temp\ufo or
init.sh ~/ufo Orx will create a project for each IDE supported by your OS at the specified location. You can copy this folder anywhere, and your project will always compile and link due to the \$ORX environment variable. It knows where the libraries and includes are for Orx.
Open your project using your favourite IDE from within the ufo/build folder.
When the blank template loads, there are two main folders to note in your solution:
config src Firstly, the src folder contains a single source file, ufo.cpp. This is where we will add the c++ code for the game. The config folder contains configuration files for our game.
What is config?
Orx is a data driven 2D game engine. Many of the elements in your game, like objects, spawners, music etc, do not need to be defined in code. They can be defined (or configured) using config files.
You can make a range of complex multi-part objects with special behaviours and effects in Orx, and bring them into your game with a single line of code. You'll see this in the following chapters of this guide.
There are three ufo config files in the config folder but for this guide, only one will actually be used in our game. This is:
ufo.ini All our game configuration will be done there.
Over in the Orx library repo folder under orx/code/bin, there are two other config files:
CreationTemplate.ini SettingsTemplate.ini These are example configs and they list all the properties and values that are available to you. We will mainly concentrate on referring to the CreationTemplate.ini, which is for objects, sounds, etc. It's good idea to include these two files into your project for easy reference.
Alternatively you can view these online at https://github.com/orx/orx/blob/master/code/bin/CreationTemplate.ini and here: https://github.com/orx/orx/blob/master/code/bin/SettingsTemplate.ini

The code template
Now to take a look at the basic ufo.cpp and see what is contained there.
The first function is the Init() function.
This function will execute when the game starts up. Here you can create objects have been defined in the config, or perform other set up tasks like handlers. We'll do both of these soon.
The Run() function is executed every main clock cycle. This is a good place to continually perform a task. Though there are better alternatives for this, and we will cover those later. This is mainly used to check for the quit key.
The Exit() function is where memory is cleaned up when your game quits. Orx cleans up nicely after itself. We won't use this function as part of this guide.
The Bootstrap() function is an optional function to use. This is used to tell Orx where to find the first config file for use in our game (ufo.ini). There is another way to do this, but for now, we'll use this function to inform Orx of the config.
Then of course, the main() function. We do not need to use this function in this guide.
Now that we have everything we need to get start, you should be able to compile successfully. Run the program and an Orx logo will appear slowly rotating.

Great. So now you have everything you need to start building the UFO game.

Setting up the game assets
Our game will have a background, a UFO which the player will control, and some pickups that the player can collect.
The UFO will be controlled by the player using the cursor keys.
First you'll need the assets to make the game. You can download the file  assets-for-orx-ufo-game.zip which contains:
The background file (background.png):

The UFO and Pickup sprite images (ufo.png and pickup.png):

And a pickup sound effect (pickup.ogg):
pickup.ogg
Copy the .png files into your data/texture folder
Copy the .ogg file into your data/sound folder.
Now these files can be accessed by your project and included in the game.

Setting up the Playfield
We will start by setting up the background object. This is done using config.
Open the ufo.ini config file in your editor and add the following:

[BackgroundGraphic] Texture = background.png Pivot = center
The BackgroundGraphic defined here is called a Graphic Section. It has two properties defined. The first is Texture which has been set as background.png.
The Orx library knows where to find this image, due to the properties set in the Resource section:

[Resource] Texture = ../../data/texture
So any texture files that are required (just like in our BackgroundGraphic section) will be located in the ../../data/texture folder.
The second parameter is Pivot. A pivot is the handle (or sometimes “hotspot” in other frameworks). This is set to be center. The position is 0,0 by default, just like the camera. The effect is to ensure the background sits in the center of our game window.
There are other values available for Pivot. To see the list of values, open the CreationTemplate.ini file in your editor. Scroll to the GraphicTemplate section and find Pivot in the list. There you can see all the possible values that could be used.
top left is also a typical value.
We need to define an object that will make use of this graphic. This will be the actual entity that is used in the game:

[BackgroundObject] Graphic = BackgroundGraphic Position = (0, 0, 0)
The Graphic property is the section BackgroundGraphic that we defined earlier. Our object will use that graphic.
The second property is the Position. In our world, this object will be created at (0, 0, 0). In Orx, the coordinates are (x, y, z). It may seem strange that Orx, being a 2D game engine has a Z axis. Actually Orx is 2.5D. It respects the Z axis for objects, and can use this for layering above or below other objects in the game.
To make the object appear in our game, we will add a line of code in our source file to create it.
In the Init() function of ufo.cpp, remove the default line:
orxObject_CreateFromConfig("Object"); and replace it with:
orxObject_CreateFromConfig("BackgroundObject"); Compile and run.
The old spinning logo is now replaced with a nice tiled background object.

Next, the ufo object is required. This is what the player will control. This will be covered in Part 2.
• By yyam

Hey there! I released my game, Hedgehogs Can Fly, on GameJolt today! It's a cute, 2D physics-platformer where you try to fling a hedgehog through tricky levels to get to the finish line. I wrote it from scratch in C++ with SFML. There are multiple types of terrain each with different properties and effects making for some interesting level design. The physics/level code also allows for free-form terrain that isn't constrained to a tile grid. The levels are loaded from color-coded image files, I have an entry in the devlog on the GameJolt page explaining how it all works!
If these screenshots look cool, visit the GameJolt page here (With Trailer Video!)
Screenshots incoming...

Have a nice day

# C++ Well-defined shifts

## Recommended Posts

This is basically a repost of http://www.gamedev.ru/flame/forum/?id=228402 .

As many programmers know, there is quite a lot of undefined behavior in C++ shifts.

So let's make our own ones!

Since the task is relatively small, it seems worth doing WELL, to save everyone using it a bit of time.

Motivation:

0. Well-defined semantics rulez.

1. For many cases (e. g. shifts by a constant) the result should be exactly the same as built-in shifts, with no performance penalty.

2. For many other cases performance does not matter.

3. If old code (with built-in shifts) worked, then, presumably the shift amount was always in the allowed range, and the branch in the new code would have essentially 100% prediction.

4. For the cases, where shifts are variable, and performance matters so much, that even predicted branches make difference (e. g. coding/decoding) - well there built-in shifts still remain. Hopefully, it is <5% of code.

My present attempt:

// This software is in the public domain. Where that dedication is not
// recognized, you are granted a perpetual, irrevocable license to copy
// and modify this file as you see fit.

template<typename L,typename R>
L shl(L value,R amount)
{
if(amount>=R(sizeof(L))*8) return L(0);
if(amount<R(0))
{
if(amount<=-R(sizeof(L))*8)
{
if(value<L(0)) return L(-1);
return L(0);
}
return value>>(-amount);
}
return value<<amount;
}

template<typename L,typename R>
L shr(L value,R amount)
{
if(amount>=R(sizeof(L))*8)
{
if(value<L(0)) return L(-1);
return L(0);
}
if(amount<R(0))
{
if(amount<=-R(sizeof(L))*8) return L(0);
return value<<(-amount);
}
return value>>amount;
}

I subject it to scrutiny of the community.

Basically, pretty please, check that all corner cases actually work.

Quite a bit of trickery went into the writing this code, to satisfy the rules of C++. As you can imagine, there are A LOT of corner cases, and I am not confident everithing works correctly.

http://www.gamedev.ru/flame/forum/?id=228402 - original post (in Russian)

http://en.cppreference.com/w/cpp/language/operator_arithmetic - for language-lawering

http://rextester.com - for testing snippets

http://gcc.godbolt.org - for testing assembly output

P. S. Does anyone see a nice way to do it in C? Templates are C11, macros evaluate arguments multiple times, and lots of functions are ugly.

I'm asking because it seems like a good candidate for inclusion into stb_* (right along with stb_divide.h), and maybe if we ask really nicely, Sean Barrett can do just that.

##### Share on other sites

I'm not sure what you're on about.  This seems unnecessary.

They're undefined if you have a negative number of bits or if the shift is larger than the number of bits.  But you should never be in either situation.  That is a bug in the code, a bug in the logic. Writing code that solves a bug by wrapping it inside a function that hides the bug is nonsense.  That is a "Don't Do That" bug, the problem is that the programmer did something they shouldn't.

It also gets into trouble if you attempt to shift a signed negative value. Since the language does not specify exactly how numbers are encoded it would be nonsensical to enforce what happens on those situations.  That's another case of "Don't Do That'.

None of those are particularly problematic. Programmers shouldn't be doing that in the first place, and if they do, that's a bug that should be caught through testing.

As for your performance concerns, every CPU that matters implements a bit shift opcode. Every compiler I'm aware of either simplifies the expression to something even faster, or it generates the CPU-specific opcode.  It is hard to get much faster, especially compared to all your branching code.

##### Share on other sites

Bugfix: to prevent UB for left-shifting negative numbers

//return value<<amount;
return L((typename std::make_unsigned<L>::type)(value)<<amount);

//return value<<(-amount);
return L((typename std::make_unsigned<L>::type)(value)<<(-amount));

This requires C++11.

##### Share on other sites

So you impose a cost to ALL code using bitwise shift because someone once wrote a bug?

Most functions have rules about their use, such as not passing null values as parameters, or ensuring elements are within a legal range.  This is no different. Shift operations have a mandatory range between zero and the number of bits of the type.

Passing a negative shift distance is a bad parameter, exactly the same as if the programmer had passed an invalid object address or dereferenced a null pointer.  Don't do that in the first place, then you don't need to write special code to handle the buggy code.

##### Share on other sites

I could see adding some asserts, do the shifts assert?  I can't remember the last time I passed a bad parameter to a shift.

##### Share on other sites

I for one think this is a good idea. I would like to be able to shift something to the left some number of times as a way to divide it by a power of 2. If I shift by too much, the result should be 0 (at least for unsigned integer types).

The current C++ rules are written so shifts can be translated to machine-code shifts, and those have semantics that made the hardware easy to implement (usually several decades ago). There are natural semantics for shifts with any number of bits, and it would be good to have them available.

A similar "feature" of C++ that bothers me even more is the rules on integer modulo and division when it comes to signs. a % b should always return a number in the interval [0, abs(b)), and a / b should round down. The fact that 7 % 10 and -3 % 10 can be different boggles the mind. The whole point of arithmetic modulo 10 is that you should not see any difference between numbers that are a multiple of 10 apart. Just ask your mathematician friends.

##### Share on other sites
4 hours ago, frob said:

It also gets into trouble if you attempt to shift a signed negative value. Since the language does not specify exactly how numbers are encoded it would be nonsensical to enforce what happens on those situations.  That's another case of "Don't Do That'.

In practice, on every compiler that matters to gamedev, shifting an unsigned integer to the right is a zero-fill right shift (logical shift), and shifting a signed integer to the right is a sign-extending right shift (arithmetic shift).

The standard doesn't define this because the language needs to be able to run on a soviet toaster as well as a PC... but in practice, this is the defacto definition of what signed and unsigned shifts should do. Every sane CPU architecture has instructions for both arithmetic and logical shift, and every sane C/C++ compiler will map unsigned to logical and signed to arithmetic. You'll likely find code at every gamedev company that relies on this definition, even though it's not guaranteed by the standard, e.g.

#define MAKE_MASK_64(StartBit, NumBits) \
((u64) ((s64) 0x8000000000000000i64 >> (NumBits-1)) >> (64-NumBits-StartBit))

The cautious projects will include a static assertion that verifies that the compiler is doing the defacto standard behaviour

##### Share on other sites
41 minutes ago, Hodgman said:

the language needs to be able to run on a soviet toaster as well as a PC

Soviet toasters probably ran 8086 knockoffs, ironically.

Edited by TheChubu

##### Share on other sites
2 hours ago, alvaro said:

The fact that 7 % 10 and -3 % 10 can be different boggles the mind.

When either value is negative the resulting sign is implementation defined.

It is the polite way of saying: "The operation does whatever your hardware does on that opcode."

Various chipsets handle it differently, and the standards committee tries very hard to prevent systems from being slow in the name of uniformity.  Java tried doing that, and it backfired spectacularly. It took until Java 6 for them to work most of them out, but during the early years many operations, especially trickier floating point operations, were incredibly slow as they were processed in software rather than using the slightly different hardware operations.

In practicality it is not an issue.  In our field you typically have either a single platform, or you have up to four platforms with known specifications, and that's it. You don't need to write code for every conceivable platform, nor have thousands of variations for every compiler and chipset and operating system throughout history.

##### Share on other sites
14 hours ago, frob said:

I'm not sure what you're on about.

The core difference seems to be that you consider shift amount outside of [0; width) nonsensical. I do not.

Not necessarily trying to convince you, just hope this helps to explain where I'm coming from.

This (to me) is about providing sensible, uniform semantics.

It seems pretty obvious to me that the Right Thing(TM) semantics-wise (ignoring performance) is for 20<<100 to be 0, and for 20<<-2 to be the same as 20>>2, namely 5.

It is also looks like it makes more sense mathematically.

alvaro seems to agree.

At very least, when shifting bunch of numbers it would be nice not to special-case full-width shifts.

Also sometimes it does make sense to write a<<(b-c) with the sign of (b-c) not known at compile-time.

14 hours ago, frob said:

Most functions have rules about their use, such as not passing null values as parameters, or ensuring elements are within a legal range.  This is no different. Shift operations have a mandatory range between zero and the number of bits of the type.

Yes, this is the semantics of shift operations in C++ Standard. It does not necessarily follows that this is a good idea for C++. It SURELY does not follow that this is a good idea for my own bit-shifting function.

Also, ambiguous wording: "number of bits of the type" is actually excluded from range.

14 hours ago, frob said:

So you impose a cost to ALL code using bitwise shift because someone once wrote a bug?

I'm not advocating changing all shifts into functions calls, just considering it.

Suppose you have two functions to do some things: one is robust, and another is fast, but works only for restricted inputs.

You can usually do one of the the following:

1. Use robust one everywhere in code, and change it to fast, where it causes performance problems.

2. Use fast one everywhere in code, and change it to robust, where it causes correctness problems.

It does not seem too outrageous to prefer option 1.

Even if you go with the option 2, the shl/shr functions provide the robust functionality mentioned.

Also, for shifts by constant the cost should be exactly zero, assuming compiler is allowed to optimize at all.

8 hours ago, frob said:

When either value is negative the resulting sign is implementation defined.

Not since C99 and C++11 respectively. It is well-defined now.

So

8 hours ago, frob said:

In practicality it is not an issue.

it (result being implementation-defined) is indeed not an issue. Result being well-defined (to the value it is) is.

10 hours ago, alvaro said:

A similar "feature" of C++ that bothers me even more is the rules on integer modulo and division when it comes to signs. a % b should always return a number in the interval [0, abs(b)), and a / b should round down. The fact that 7 % 10 and -3 % 10 can be different boggles the mind. The whole point of arithmetic modulo 10 is that you should not see any difference between numbers that are a multiple of 10 apart. Just ask your mathematician friends.

Yes, C++ lack of Euclidean division is similar (I mention stb_divide.h in the opening post). However:

1. Division is actually well-defined (sans INT_MIN/-1), even if the definition is suboptimal. So it just produces garbage value, whereas UB nukes entire codepaths. I slightly prefer the former, I suppose.

2. Non-Euclidean division seems to be problematic more often, than restricted shifts.

3. Performance considerations are probably no concern, since the division itself is so slow.

10 hours ago, Hodgman said:

In practice, on every compiler that matters to gamedev, shifting an unsigned integer to the right is a zero-fill right shift (logical shift), and shifting a signed integer to the right is a sign-extending right shift (arithmetic shift).

In fact, in the code above I rely on implementation-defined behavior:

1. Signed right-shift sign-extends.

2. Unsigned->signed conversion produces the value with the same bitwise representation.

Both seem to be rock-solid in practice.

Edited by FordPerfect