Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Valid_Dark]

What about programs that only fit a very specific need for a very tiny community?
Because I have one of those, I'm thinking about putting up on github.

[Skyrunner]

Though working on personal projects that AREN'T games is generally a lot more attractive to employers.

I wonder if colleges look at personal projects too. :c

[Valid_Dark]

Though working on personal projects that AREN'T games is generally a lot more attractive to employers.

I wonder if colleges look at personal projects too. :c

Probably, I showed the college I wanted to go to the stuff I was working on, turns out I can't afford to go, but the guy that I showed it sends me cards at christmas and my birthday and emails me about how his classes are going (he's a teacher)
idk, I guess he really wanted me to go to that school.

[alway]

Do you guys put your personal projects on your resumes? I guess it couldn't hurt, but I'm not sure how professional it would be.

What employers care about, in this order:
1. Skills; what can you do for them; which languages can you do, what APIs do you have experience with, that sort of thing
2. Experience; what have you already done; this includes both work experience and programming projects, both group and solo. The order you list them in should be based on how impressive/relevant they are. Personal projects are a plus, since it shows initiative, saying things were class projects or required for something will subtract from the value of that item*. Unless you have experience in the industry, work experience goes after projects.
3. Education; where were you formally instructed, how well did/are you do/doing there, your major, and year level.

* As to the reasons behind the latter modifier, it's because required work implies very little about what you actually did. For example, RIT's Software Engineering major has a class in which the quarter long project is to create a pizza delivery management system. While the project might by impressive if it was a personal project, when an employer gets 20 resumes all talking about the pizza delivery management system, they kind of stop caring. Such projects also are less indicative of effort put into it. A class project with random group members will often have 1 or 2 people doing all the work for the group; employers know this.

[Valid_Dark]

so I'm working on something I haven't touched in a few months,
and it's crazy looking at how I used to code. Like seriously lulzworthy.

I have to look back at some of the code to figure out what is doing exactly what since I left no notes like a dumdum.  Made me smile / lol.

[cerapa]

Code: [Select]

void Planet::VariableBlurHumidity(int Strength) {

    map_point                   TempMap[800][400];

    for (int y=0; y<400; y++) {
        for (int x=0; x<800; x++) {
            if (DataMap[x][y].humidity < 255) {

                int AreaHumidity = 0;

                for (int x2 = -Strength; x2<Strength+1; x2++) {
                    for (int y2 = -Strength; y2<Strength+1; y2++) {
                        int rX = x+x2;
                        int rY = y+y2;

                        if (rX<0) rX +=800;
                        if (rX>799) rX -=800;

                        if (rY < 0) rY = 0;
                        if (rY > 399) rY = 399;

                        AreaHumidity += DataMap[rX][rY].humidity;
                    }
                }
                TempMap[x][y].humidity = AreaHumidity/( (Strength*2+1)*(Strength*2+1) );
            }
            else TempMap[x][y].humidity = DataMap[x][y].humidity;
        }
    }

    for (int y=0; y<400; y++) {
        for (int x=0; x<800; x++) {
            DataMap[x][y].humidity = TempMap[x][y].humidity;
        }
    }
}This is randomly broken. "Program received signal SIGSEGV, Segmentation fault." Through experimentation I found that "map_point TempMap[800][400];" causes the whole thing to fail. Even if its at the end of the function. Before it should be doing anything. And it worked perfectly before.

EDIT: Creating a 2d array in other functions causes the same problems, unless the 2nd dimension of the array is 251 or smaller. WTF? Again, this worked before, and I have an identical array sitting comfortably in the class without any worries.

EDIT2: Wut. map_point is a struct. Making the struct itself bigger broke the array. But again, the array of the planet class works perfectly. I can turn the array of structs into the relevant thing(I wasnt using anything besides humidity) to keep this from happening this time, but I'm afraid this problem might pop up again later.

EDIT3: Apparently even the class array cant handle a struct bigger than 6 bytes. Cant say I'm too happy about that. But 6 should be plenty.

[olemars]

TempMap is a local variable, which means it gets created on the stack. The stack is of system/compiler dependent size, and can be very small as it's not intended for that kind of data.

You will have to create it on the heap to avoid problems, which means making map_point a pointer and newing and deleting the array. If you call the function a lot you should just keep the array around instead of newing and deleting it over and over, as that's a slow operation.

[cerapa]

TempMap is a local variable, which means it gets created on the stack. The stack is of system/compiler dependent size, and can be very small as it's not intended for that kind of data.

You will have to create it on the heap to avoid problems, which means making map_point a pointer and newing and deleting the array. If you call the function a lot you should just keep the array around instead of newing and deleting it over and over, as that's a slow operation.

I had the class itself on the stack too. No wonder it ran out of memory.

Thanks for the help.

[Immortal]

Hello all,
Well recently I was pointed towards steering behaviours and the results are fantastic.
I have seeking, following, avoiding, and arrival facing forwards (reverse thrust is only 1/5 main engine).
If you are unfamiliar I am making an Asteroids like game.

The problem I have is arrival with the shortest trip possible. To arrive the fastest way, it has to accelerate forwards then spin around and full thrust in the opposite direction. I can't get my AI to account for the time it takes to spin the ship around, and when it does spin around at the correct time, it decides it's now going to slow and spins all the way back (it just ends up spinning in circles till impact).

If anyone has an idea/tutorial/pseudocode on how to solve this it would be helpful. I have tried splitting it into pre//post spin but that doesn't work (or at least hasn't for me)

[olemars]

Would be helpful to know what your current attempt looks like.

[GalenEvil]

Immortal: the issue sounds like it is in your "seek" behaviour. What you need to do is make a separate "Arrive" behaviour which will start slowing down at some distance from the target location that will allow it to stop at the location using a desired time to target variable to control how fast it attempts to slow down.

Spoiler: "Arrive Class" Pseudocode gleaned from "Artificial Intelligence for Games, 2nd Edition" (click to show/hide)

Code: [Select]

class Arrive:
  # Holds the kinematic data for the character and target
  character
  target

  # Holds the max acceleration and speed of the character
  maxAcceleration
  maxSpeed

  # Holds the radius for arriving at the target
  targetRadius

  # Holds the radius for beginning to slow down
  slowRadius

  # Holds the time over which to achieve target speed
  timeToTarget = 0.1

  def getSteering(target):

    # Create the structure to hold our output
    steering = new SteeringOutput()

    # Get the direction to the target
    direction = target.position - character.position
    distance = direction.length()

    # Check if we are there, return no steering
    if distance < targetRadius
      return None

    # If we are outside the slowRadius, then go max speed
    if distance > slowRadius:
      targetSpeed = maxSpeed

    # Otherwise calculate a scaled speed
    else:
      targetSpeed = maxSpeed * distance / slowRadius

    # The target velocity combines speed and direction
    targetVelocity = direction
    targetVelocity.normalize()
    targetVelocity *= targetSpeed

    # Acceleration tries to get to the target velocity
    steering.linear = targetVelocity - character.velocity
    steering.linear /= timeToTarget

    # Check if the acceleration is too fast
    if steering.linear.length() > maxAcceleration:
      steering.linear.normalize()
      steering.linear *= maxAcceleration

    # Output the steering
    steering.angular = 0
    return steering


And a demonstration of what it looks like in 3D: Arrive Demo

[HavingPhun]

I have been in and out of learning c++ and now i'm finally starting to understand it. But I'm having a problem with this code:

Code: [Select]


#include <stdio.h>
#include <iostream>
#include "human.h"
#include "World.h"


int main(int argc, char **argv)
{
enum STATE{INITIALIZE, ON, OFF, PAUSED, RUNNING, MENU, COMMAND};
int Response;
char Command;
char * PCommand;
PCommand = &Command;
int Counter;
int LastStop;
int NumberOfPeople;
int MoneyInCirculation;
char Name;
STATE state;

state = INITIALIZE;

printf("Economy Simulation Program. \n");
printf("Type one to Continue, Any other number to exit:\n");
std::cin >> Response;

if(Response == 1)
{
state = ON;
}
else
{
state = OFF;
}

while(state != OFF)
{
int Counter = 0;
int LastStop = 0;
int NumberOfPeople = 0;
int MoneyInCirculation = 0;

printf("Enter a command. Type help for a list of commands:\n");
std::cin >> *PCommand;

if(Command == "help")
{
printf("New Person \nClose \nStop \nContinue");
}
printf("Enter a command:\n");


Human bob;



bob.Age = 27;
bob.Clothing = 80;
bob.Hunger = 20;
bob.Thirst = 15;
bob.Blood = 100;
}


return 0;
}

I get this error: C:/Users/HavingPhun/Desktop/Economy Simulations/Simulation 1/main.cpp:46:19: error: ISO C++ forbids comparison between pointer and integer [-fpermissive]

On this line:

Code: [Select]

std::cin >> *PCommand;I was trying to do that earlier with:

Code: [Select]

std::cin >> Command;and it still gave me the same error above, and also said I cannot change a const char(Or something along the lines of that). So I tried to change it indirectly with a pointer. Right now I'm just screwing around with some simulations until I can redownload mingw on the 17th and get libtcod working. Sorry if its hard to read I am still working with wordpad until the 17th also. I also did not include the header files since the error and variables involved are all in the main file.Thanks.

[fergus]

Code: [Select]

std::cin >> *PCommand;

Try replacing

Code: [Select]

char Command;with

Code: [Select]

char Command[10];and

Code: [Select]

std::cin >> *PCommand;with

Code: [Select]

std::cin >> Command;

[Immortal]

Immortal: the issue sounds like it is in your "seek" behaviour. What you need to do is make a separate "Arrive" behaviour which will start slowing down at some distance from the target location that will allow it to stop at the location using a desired time to target variable to control how fast it attempts to slow down.

Thanks, but that is exaclty what I have for arriving facing forward. The problem is when I want to arrive facing away it always overshoots the target (the ship takes time to change direction). How would I be able to post a java for you to see if you're interested?

[GalenEvil]

Immortal: I suggest posting it to pastebin since it has decent permanence and is useful for moving code snippets around. If you have a way to do rotations without translating at all then I'd suggest have it cut forward thrust once it is on a collision course, spin around to face directly away, and then use forward thrust to slow down. You'd have to take into account the rotational acceleration of the ship to figure out a minimum safe distance for the start of the maneuver. Back to the code: Upload it to pastebin and send me a link to it via PM, or onto this thread since I'm sure others here would also be able to help you if I am slow getting back to you