SRL Archive - SRL Bugs

(* start math.simba additions *)
type
  TVector = record
    a: Extended;
    b: Extended;
    Magnitude: Extended;
  end;


function fixRad(rad: Extended): Extended;  // by Nava2
var
  math_2pi: Extended;
begin
  math_2pi := 2*pi;
  result := rad;

  while (result >= math_2pi) do
    result := result - math_2pi;

  while (result < 0) do
    result := result + math_2pi;
end;

function AngleBetween(v1, v2: TVector): Extended;
var
  a1, a2: Extended;
begin
  a1 := ArcTan2(v1.b, v1.a);
  a2 := ArcTan2(v2.b, v2.a);

  Result := fixRad(a1 - a2)
end;


procedure Normalise(var vect: TVector);
begin
  vect.a := vect.a / vect.Magnitude;
  vect.b := vect.b / vect.Magnitude;
  vect.Magnitude := 1;
end;

function CreateVector(Origin, vPoint: TPoint; NormaliseV: Boolean): TVector;
begin
  Result.a := vPoint.x - Origin.x;
  Result.b := 0 - (vPoint.y - Origin.y); // inverts y axis as well
  Result.Magnitude := Sqrt(Sqr(Result.a) + Sqr(Result.b));
  if NormaliseV then
    Normalise(Result);
end;

(* end math.simba additions *)

function rs_GetCompassAngleRadians: Extended;
var
  tpa: TPointArray;
  atpa: T2DPointArray;
  i, h, l: Integer;
  VectorArray: Array of TVector;
  RefPoint: TPoint;
  angles: TExtendedArray;
  math_2pi: Extended;
begin
  math_2pi := 2*pi;

  { Colour of 3 E/S/W dots }
  FindColorsTolerance(tpa, 2370604, 529, 9, 558, 39, 0);

  atpa := SplitTPA(tpa, 1);

  h := High(atpa);
  setLength(tpa, h+1);
  l := 0;
  for i := 0 to h do
    { Gets rid of dots under length of 3 and get the MiddleTPA }
    if high(atpa[i]) > 4 then
    begin
      tpa[l] := MiddleTPA(atpa[i]);
      inc(l);
    end;
  
  if l <> 3 then
    Exit;

  setLength(tpa, l);

  FilterPointsDist(tpa, 9, 20, 543, 24);

  SetLength(VectorArray, 3);

  for i := 0 to 2 do
    VectorArray[i] := CreateVector(Point(543, 24), tpa[i], True);

  angles := [AngleBetween(VectorArray[0], VectorArray[1]),
         AngleBetween(VectorArray[1], VectorArray[2]),
         AngleBetween(VectorArray[2], VectorArray[0])];

  for i := 0 to 2 do
  begin
    if angles[i] > pi then
      angles[i] := math_2pi - angles[i];
    { compare angle between, > 100 then it's the south dot }
    if angles[i] > radians(100) then
    begin
      case i of
    0: RefPoint := tpa[2];
    1: RefPoint := tpa[0];
    2: RefPoint := tpa[1];
      end;
      { break the loop, we have reference pt. }
      break;
    end;
  end;


  { Using the south dot to measure compass angle.. }
  Result := ArcTan2(-(RefPoint.Y - 24), RefPoint.X - 543) - Radians(80);
  Result := fixRad(Result - Pi - 0.05*Pi);
end;

function rs_GetCompassAngleDegrees: Extended;
begin
  Result := Degrees(rs_GetCompassAngleRadians());
end;

Excellent. So this version works?

What's the difference between this and the other one you posted. Also, you could have just posted on the same thread, no need for a new one.

Compass was updated in rs, the difference is the colour and filterpointsdist. Wiz said to just post a new thread.

Alright. I'll commit this now.

how do i do this, do i just add it in somewhere? plz help =P

It has been added to SRL, just wait for you SRL to automatically update.

Since the thread title now says [Committed] that means it was added to SRL.
So if you just go to Simaba, click the SRL tab, and click update it will update it for you (but it does take a few minutes before simba recognized there was an update if you see the sane number for local version and remote version when you update, or if you see no update available, then you need to wait a few more minutes before you try updating again.
SRL Updater: Local Version - "834" | Remote Version = "834"

This didn't work for me at all. It returned 0 when I was around 180 degrees which caused SPS to fail.

From: Coh3n
This didn't work for me at all. It returned 0 when I was around 180 degrees which caused SPS to fail.

Ditto.

From: Coh3n
This didn't work for me at all. It returned 0 when I was around 180 degrees which caused SPS to fail.

You didn't test it before committing it?

For the first time ever. The only thing changed was the colour and I assumed it was tested before posted.

That's my mistake.

From: Coh3n
For the first time ever. The only thing changed was the colour and I assumed it was tested before posted.

That's my mistake.

Even the best can make mistakes :)

This is so weird. Perhaps the colours are dynamic. Of course I tested it, I'm not just blindly committing a fix…

From: ReadySteadyGo
This is so weird. Perhaps the colours are dynamic. Of course I tested it, I'm not just blindly committing a fix…

Shush! :fiery:

Definitely doesn't work.. if anything, it made scripts worse.. My script no longer runs to the middle but rather to the very far left on minimap lol. reverted back to normal :P and now it rotates and can't stop.

I say for now we use Fakawi's way of getting to North and remove the angle check (?) in SPS. I'll give it a shot to make sure it works.

Removing the angle check in SPS will work, but it will slow things down because there's a 1-2 second wait in SetAngle no matter what.

Just revert to old one, it works for clickNorth atleast, and makes sps work as long as u dont rotate.

Yeah that'll present itself as a problem later on with scripts that make good use of random camera movement. For the record I disabled SPS's angle check, added Fakawi's North procedure and put in a SetAngle and it's working fine for me so far, but I only call it once upon logging in initially or after a break.

Just another random idea but… could we not do something along the lines of making a rough bitmap of the "N" symbol on the MM and checking if that bitmap is within a coordinate range to tell us if we're facing north or not? We have BitmapRotated, ect… I know it would be limited to north only, but if it worked that should take care of SPS.

Fakawi's needs some failsafes, if anyone's going to commit it ;)

I fixed this. ReadySteadyGo was using a tolerance of 0 which I changed to 5. He also used "if (length <> 3) then" and I switched to "if (length < 3) then". From what I tested, it works, so please let me know when your guys' SRL updates.

From: Coh3n
From what I tested, it works, so please let me know when your guys' SRL updates.

My SLR updated and i'm still running into the same problem, also when im attacking shifters or whatever their called the script constantly right clicks to attack them. It will do that until they die. Also the bot seems to be confused and sometimes is unable to find the enemy and instead right clicks in the center of my screen. This doesnt slow stuff down it just pretty much screams that i'm a bot.

From: funnyman9292
My SLR updated and i'm still running into the same problem, also when im attacking shifters or whatever their called the script constantly right clicks to attack them. It will do that until they die. Also the bot seems to be confused and sometimes is unable to find the enemy and instead right clicks in the center of my screen. This doesnt slow stuff down it just pretty much screams that i'm a bot.

I committed this like 5 minutes ago. Your update was probably from an earlier update. The shifter thing has nothing to do with this, it's something you have to post on the thread from which you downloaded the script.

From: funnyman9292
My SLR updated and i'm still running into the same problem, also when im attacking shifters or whatever their called the script constantly right clicks to attack them. It will do that until they die. Also the bot seems to be confused and sometimes is unable to find the enemy and instead right clicks in the center of my screen. This doesnt slow stuff down it just pretty much screams that i'm a bot.

You are still running into compass rotation/accuracy issues? Also, the rest of your post is detailing problems with Weequ's Pest Controller. Notice the title of this thread "[Committed] rs_CompassAngle fix, again" this thread is about fixing the compass, not Weequ's script. Please post in the script's thread if you are having problems with it.

E: Ninja'd by Coh3n <3

is this the reason for constant spinning in scripts that use compass angle?

been out most of today it looks like the compass changed

yea doesn't seem to work for me either

I think so yes, and it wasn't working all the time. Everything should be fine in the next SRL update.

From: Coh3n
I think so yes, and it wasn't working all the time. Everything should be fine in the next SRL update.

Can you atleast post the fix so I don't need to wait a random hour…?

This has been giving some people out of range errors. I looked it over and got into it. This is my revised compass:

Function rs_GetCompassAngleRadians: Extended;
var
  TPA: TPointArray;
  T, M: TPoint;
  B: TBox;
  S: Extended;
begin
  B := IntToBox(529, 9, 557, 39);
  with B do
    FindColorsSpiralTolerance(T.x, T.y, TPA,65536,  x1, y1, x2, y2, 0);
  if Length(TPA) < 1 then
   Exit;
  M := MiddleBox(B);
  T := MiddleTPA(TPA);
  Result := ArcTan2(-(T.Y - M.Y), T.X - M.X+0.5) + Radians(90);
  Result := fixRad(Result - Pi);
end;

Its only off by 2.6 degrees instead of over 5 when a pixel shift happens.