Getting A List Of All Points In A Line

[Woodcutter11]

So, I want something that calculates the points between two other points. I tried this [URL=http://en.wikipedia.org/wiki/Bresenham%27s_line_algorithm]"Bresenham's line algorithm"[/URL] but it didn't work out. Here's the function from it (C# with the XNA framework):

[csharp]
public List<Vector2> GetPixelsInLine(Vector2 p1, Vector2 p2)
{
List<Vector2> ret = new List<Vector2>();

float deltax = p2.X - p1.X;
float deltay = p2.Y - p1.Y;

float error = 0;
float deltaerr = deltay / deltax;

float y = p1.Y;

for (float x = p1.X; x < p2.X; x++)
{
ret.Add(new Vector2(x, y));

error = error + deltaerr;

if (Math.Abs(error) >= 0.5)
{
y = y + 1;
error = error - 1;
}
}

return ret;
}
[/csharp]

Please help me!

[esalaka]

It... Didn't work?

What do you mean it didn't work?

Also, couldn't you just [url=http://en.wikipedia.org/wiki/Linear_equation#Two-point_form]use the line equation[/url] and round y correctly by doing something like Math.Floor( y + 0.5 )?

[editline]11th November 2010[/editline]

It should be basic algebra but I can't tell you why it doesn't work if it doesn't 'cause I'm gonna go to sleep :v:

[Overv]

[QUOTE=esalaka;25999169]It... Didn't work?

What do you mean it didn't work?

Also, couldn't you just [url=http://en.wikipedia.org/wiki/Linear_equation#Two-point_form]use the line equation[/url] and round y correctly by doing something like Math.Floor( y + 0.5 )?

[editline]11th November 2010[/editline]

It should be basic algebra but I can't tell you why it doesn't work if it doesn't 'cause I'm gonna go to sleep :v:[/QUOTE]

Only works when dx > dy. The Bresenham algorithm switches x and y if that is not the case.

[NovembrDobby]

You could easily get a specific point or set number of points between them by using Vector2.Lerp():

List<Vector2> pointsBetweenPoints(Vector2 p1, Vector2 p2)
{
   List<Vector2> points = new List<Vector2>(100);
   for(float i = 0; i <= 1; i += 0.01)
   {
      points.Add(Vector2.Lerp(p1, p2, i);
   }

   return points;
}

What's this for by the way? If it's to move something in a line there are nicer ways :v:

[Jallen]

Probably line drawing.

Here's my implementation (essentially identical to the wikipedia pseudocode but translated into C++. It works 100%) of bresenhams (which I used in my [B][url="http://jallenbah.co.uk/mousetracker.php"]Mouse Tracker[/url][/B]) with all of the drawing/library specific stuff edited out:
[cpp]void Line(int x0, int y0, int x1, int y1)
{
bool steep = abs(y1 - y0) > abs(x1 - x0);
if(steep)
{
Swap(x0, y0); // swap simply swaps the values of the parameters
Swap(x1, y1);
}
if(x0 > x1)
{
Swap(x0, x1);
Swap(y0, y1);
}
int deltax = x1 - x0;
int deltay = abs(y1 - y0);
float error = 0;
float deltaerr = float(deltay) / deltax;
int ystep;
int y = y0;
if(y0 < y1) ystep = 1; else ystep = -1;
for(int x = x0; x < x1; x++)
{
if(steep)
{
// the pixel at Vector(y,x) is on the line
}
else
{
// the pixel at Vector(x,y) is on the line
}
error = error + deltaerr;
if(error >= 0.5)
{
y = y + ystep;
error = error - 1.0;
}
}

}[/cpp]
Hope this could be of some help.

[esalaka]

[QUOTE=Matthew0505;26015119]The Wikipedia page has an integer version you know. Using floats is useless since you can't split pixels.[/QUOTE]

Actually you can on a LCD screen. It's called subpixel rendering.

[sim642]

[QUOTE=Matthew0505;26027917]Still not something you would do with floats[/QUOTE]

Why not? SFML does that AFAIK.

[Siemens]

[QUOTE=Matthew0505;26039916]Because you wouldn't use an inaccurate float for precision subpixel rendering[/QUOTE]

Subpixel rendering is just fuzzing colours together anyway, so it's hardly accurate to begin with.

[esalaka]

[QUOTE=Siemens;26040651]Subpixel rendering is just fuzzing colours together anyway, so it's hardly accurate to begin with.[/QUOTE]

No.

Subpixel rendering is the process of using the R, G and B components of a pixel on a LCD screen to draw things in triple the horizontal (x-axis) definition by essentially treating each component as a gray pixel. That's why your fonts look so smooth nowadays!

[Siemens]

Last I checked R, G and B are colours bro.

[ShaunOfTheLive]

Yes, but the eye treats an RGB block and a GBR block as both being black, even though one is shifted one-third of a pixel. There is always some colour fringing though.

[Siemens]

[QUOTE=shill le 2nd;26048501]Yes, but the eye treats an RGB block and a GBR block as both being black, even though one is shifted one-third of a pixel. There is always some colour fringing though.[/QUOTE]

I should clarify.

What I meant by 'colour fuzzing' is that the techniques used to achieve subpixel rendering basically treat each pixel's colour components as a unit that's able to be used to gain a greater horizontal resolution.

So there's really nothing wrong with using floats to do subpixel rendering because subpixel rendering isn't the most accurate technique anyway.