SRL Archive - SRL Bugs

It seems like MM Dots and Flag are now dynamic colors?
By eye MMdots look the same to me, but the flag looks different.
There are many people talking about it/rumors lets figure it out and get it fixed in this ONE THREAD.

I'm using MSI and the walking is working just fine (so far) so I think flag is okay..?
What about MM dots guys?

Everyone post your finding and fixes here!

FlagPresent didn't work for me. I'm working on a new one.

They are all broken. Confirmed.


~Home

I wonder if they really think they can stop color bots this way?

program FlagTest;
{.include SRl\SRL.scar}

function FlagPresent2: Boolean;
var
  I: Integer;
  TPA: TPointArray;
  Colors: TIntegerArray;
begin
  FindColorsTolerance(TPA, 794856, MMX1, MMY1, MMX2, MMY2, 45);//Color and tolerance needs to be optimized with ACA.
  Colors := GetColors(TPA);
  ClearSameIntegers(Colors);
  for I := 0 to High(Colors) do
  begin
    FindColors(TPA, Colors[I], MMX1, MMY1, MMX2, MMY2);
    if (Length(TPA) = 15) then
    begin
      Result := True;
      Exit;
    end;
  end;
  Result := False;
end;


begin
  if FlagPresent2 then WriteLn('Flag found');;
end.

It looks for the main color in flag. It should always have 15 pixels of the exactly same color. I also checked that Item Dots don't contain any of the same color.

Edit: Perhaps I should add TPABounds check to make sure it's not just finding excatly 15 of some other very similar color?

Edit2: Flag only changes when you reload client.

Edit3: Here is an example of the current flag and how much it can change. Shape stays the same so it shouldn't be a problem.

Can someone confirm if this works?

function FlagPresent3: Boolean;
var
  Bmp, x, y: Integer;
  TPA: TPointArray;
begin
  Bmp := BitmapFromString(8, 9, 'meJxjYCAW/P//n0hBPIoxxf/DAFZBZHGsKnEBAOywLNQ=');
  Result := FindBitmapMaskTolerance(Bmp, x, y, MMX1, MMY1, MMX2, MMY2, 0, 0);
  FreeBitmap(Bmp);
end;

Attached the bitmap I put into String format.

Not sure if 0 tol will work, try switching servers, going up/down ladders and such and see if it still works or not. Thanks antti!

YoHoJo the area of the flag I took to the mask is all covered by the same color so 0 should do fine. I'm not 100% sure what the countour tolerance means though.

It looks like there's still the famous relation between colours (the add/substract one). Maybe we can do something with that?

From: Markus
It looks like there's still the famous relation between colours (the add/substract one). Maybe we can do something with that?

I do not know what the words you are talking mean :duh:

From: Markus
It looks like there's still the famous relation between colours (the add/substract one). Maybe we can do something with that?

I checked that first and couldn't find any relations. You can check it with the bitmap I posted earlier.

From: YoHoJo
I do not know what the words you are talking mean :duh:

I'll explain.

Lets say there is a rock that contains 2 different colors.

When the colors change they would both change as much meaning Color1-Color2 would always be the same. That way it was very easy to make autocolor functions.

weequ, if you are fixing the flag finding, make sure you aren't decreasing functionality. You don't want to look for the red, you want to look for one of the shaft colors. That way, even if you click on the very edge of the minimap, the function will still be able to find the flag.

From: Markus
It looks like there's still the famous relation between colours (the add/substract one). Maybe we can do something with that?

The best way to do it imo. It'll be much better than using bitmaps.

@YoHoJo, I use it to find the onion patch next to the draynor bank. Things are so simple when they do that. Findcolorstolerance() then compare all the colors at each point to see if its a constant relationship.

Perhaps you are right ephemism. I wish we had something like a Mask that supports more than one color.

Check to see if theres a relationship between the bottom of the flag, the next pixel up, then the pixel to the left. It'll probably work. I'd do it now, but I've got a paper to do right now and a dress rehearsal tonight.

From: noidea
Check to see if theres a relationship between the bottom of the flag, the next pixel up, then the pixel to the left. It'll probably work. I'd do it now, but I've got a paper to do right now and a dress rehearsal tonight.

Yeah lol I was just coming to post about that relation. It works. I wonder why jagex made some of the colors not related but others not..? I think it's just plain stupid.

From: weequ
Yeah lol I was just coming to post about that relation. It works. I wonder why jagex made some of the colors not related but others not..?

I cant remember if they request colours from the server or if the client calls Math.random(). Either way, if it had to do it all the time, it would use a lot of resources to do it for every color.

From: noidea
I cant remember if they request colours from the server or if the client calls Math.random(). Either way, if it had to do it all the time, it would use a lot of resources to do it for every color.

As I posted earlier they only change Flag when you reload client.

From: weequ
As I posted earlier they only change Flag when you reload client.

Then it may be something to do with the human eye. The eye may perceive it as ugly if there wasnt a constant contrast? Im sure it has more to do with this.

Yea but they could add a little bit of randomness too if they really wanted to give us more challenges.

Edit: Can someone test this?

function FlagPresent4: Boolean;
var
  I: Integer;
  TPA: TPointArray;
begin
  Freeze;
  FindColors(TPA, 65536, MMX1, MMY1, MMX2, MMY2);
  for I := 0 to High(TPA) do
  begin
    if (GetColor(TPA[I].X-1, TPA[I].Y-1) - GetColor(TPA[I].X, TPA[I].Y-1) = 6381921) then
    begin
      Unfreeze;
      Result := True;
      Exit;
    end;
  end;
  Unfreeze;
  Result := False;
end;

Edit2: Added freeze and unfreeze. Should work better now.

Committed weequ's flag fixes from the other thread.

From: Coh3n
Committed weequ's flag fixes from the other thread.

Awesome, thanks Coh3n!

Dot colors also have those sort of relationships. NPC difference = 4369. Drop difference = 23.

Need team and clan differences.
Use these two pixels:

{*******************************************************************************
function GetMiniMapDotsIn(WhatDot: String; x1, y1, x2, y2: Integer): TPointArray;
By: euphemism, mormonman
Description: Results the dots specified by WhatDot in the box x1, y1, x2, y2.
Usage : 'npc', 'yellow', 'n', 'y' : Yellow Dot;
        'cape', 'blue', 'team', 't', 'b' : Blue Dot;
        'item', 'red', 'i', 'r' : Red Dot;
        'player', 'white', 'p', 'w' : White Dot;
        'friend', 'green', 'f', 'g' : Green Dot;
*******************************************************************************}
function GetMiniMapDotsIn(WhatDot: String; x1, y1, x2, y2: Integer): TPointArray;
var
  I, Dif, Radius, Hi, C: Integer;
  TPA: TPointArray;
begin
  case LowerCase(WhatDot) of
    'npc', 'n', 'yellow', 'y': Dif := 4369;
    'item', 'i', 'red', 'r': Dif := 23;
    'player', 'p', 'white', 'w': Dif := 1907997;
    'friend', 'f', 'green', 'g': Dif := 5376;
    {'team', 't', 'blue', 'b', 'cape': Dif := -1;
    'clan', 'c', 'orange', 'o': Dif := -1;}
  else
    srl_Warn('GetMiniMapDotsIn', '"' + WhatDot + '" is not a valid dot type', warn_AllVersions);
  end;
  Freeze;
  Radius := Max((x2-x1)/2, (y2-y1)/2);
  FindColorsPie(TPA, 65536, 0, 0, 360, 0, radius, x1, y1, x2, y2, (x1+x2)/2, (y1+y2)/2);
  Hi := High(TPA);
  SetLength(Result, Hi + 1);
  if (Hi < 0) then Exit;
  C := 0;
  for I := 0 to Hi do
  begin
    if (GetColor(TPA[I].X-1, TPA[I].Y-1) - GetColor(TPA[I].X, TPA[I].Y-1) = Dif) then
    begin
      Result[C] := Point(TPA[I].x, TPA[i].y - 2);
      Inc(C);
    end;
  end;
  Unfreeze;
  SetLength(Result, C);
end;

How do you get the differences?

Look at my post, I added an image. Get the color at both pixels surrounded by a green box. Subtract one on right from one on the left.

Confirmed this working with npc:

{*******************************************************************************
function GetMiniMapDotsIn(WhatDot: String; x1, y1, x2, y2: Integer): TPointArray;
By: euphemism, mormonman
Description: Results the dots specified by WhatDot in the box x1, y1, x2, y2.
Usage : 'npc', 'yellow', 'n', 'y' : Yellow Dot;
        'cape', 'blue', 'team', 't', 'b' : Blue Dot;
        'item', 'red', 'i', 'r' : Red Dot;
        'player', 'white', 'p', 'w' : White Dot;
        'friend', 'green', 'f', 'g' : Green Dot;
*******************************************************************************}
function GetMiniMapDotsIn(WhatDot: String; x1, y1, x2, y2: Integer): TPointArray;
var
  I, Dif, Radius, Hi, C: Integer;
  TPA: TPointArray;
begin
  case LowerCase(WhatDot) of
    'npc', 'n', 'yellow', 'y': Dif := 4369;
    'item', 'i', 'red', 'r': Dif := 23;
    'player', 'p', 'white', 'w': Dif := 1907997;
    'friend', 'f', 'green', 'g': Dif := 5376;
    {'team', 't', 'blue', 'b', 'cape': Dif := -1;
    'clan', 'c', 'orange', 'o': Dif := -1;}
  else
    srl_Warn('GetMiniMapDotsIn', '"' + WhatDot + '" is not a valid dot type', warn_AllVersions);
  end;
  Freeze;
  Radius := Max((x2-x1)/2, (y2-y1)/2);
  FindColorsPie(TPA, 65536, 0, 0, 360, 0, radius, x1, y1, x2, y2, (x1+x2)/2, (y1+y2)/2);
  Hi := High(TPA);
  SetLength(Result, Hi + 1);
  if (Hi < 0) then Exit;
  C := 0;
  for I := 0 to Hi do
  begin
    if (GetColor(TPA[I].X-1, TPA[I].Y-1) - GetColor(TPA[I].X, TPA[I].Y-1) = Dif) then
    begin
      Result[C] := Point(TPA[I].x, TPA[i].y - 2);
      Inc(C);
    end;
  end;
  Unfreeze;
  SetLength(Result, C);
end;

If someone can confirm the rest I'll commit.

From: mormonman
Look at my post, I added an image. Get the color at both pixels surrounded by a green box. Subtract one on right from one on the left.

Oh, haha. Literally the difference of the two numbers. :p

E: Here it is with team dots working:

function GetMiniMapDotsIn(WhatDot: String; x1, y1, x2, y2: Integer): TPointArray;
var
  I, Dif, Radius, Hi, C: Integer;
  TPA: TPointArray;
begin
  case LowerCase(WhatDot) of
    'npc', 'n', 'yellow', 'y': Dif := 4369;
    'item', 'i', 'red', 'r': Dif := 23;
    'player', 'p', 'white', 'w': Dif := 1907997;
    'friend', 'f', 'green', 'g': Dif := 5376;
    'team', 't', 'blue', 'b', 'cape': Dif := 2171941;
    {'clan', 'c', 'orange', 'o': Dif := -1;}
  else
    srl_Warn('GetMiniMapDotsIn', '"' + WhatDot + '" is not a valid dot type', warn_AllVersions);
  end;
  Freeze;
  Radius := Max((x2-x1)/2, (y2-y1)/2);
  FindColorsPie(TPA, 65536, 0, 0, 360, 0, radius, x1, y1, x2, y2, (x1+x2)/2, (y1+y2)/2);
  Hi := High(TPA);
  SetLength(Result, Hi + 1);
  if (Hi < 0) then Exit;
  C := 0;
  for I := 0 to Hi do
  begin
    if (GetColor(TPA[I].X-1, TPA[I].Y-1) - GetColor(TPA[I].X, TPA[I].Y-1) = Dif) then
    begin
      Result[C] := Point(TPA[I].x, TPA[i].y - 2);
      Inc(C);
    end;
  end;
  Unfreeze;
  SetLength(Result, C);
end; 

E: Committed this for now since it works, and clan is probably never used, but can easily be added.

Pink… really?
Saw it twice when I logged in at two different times today.
Those are pinkish right…?

From: YoHoJo
Pink… really?
Saw it twice when I logged in at two different times today.
Those are pinkish right…?

They are pimps :D. Interestingly you really have to look close to notice they are pink.

They stood out to me instantly, wtfaa!
I've logged in/out many times today doing things, but only twice did they appear pink.

Yes, no close look required. Those are very, very pink. The question is, will the legitimate player-base stand for these ridiculous color changes?

From: euphemism
Yes, no close look required. Those are very, very pink. The question is, will the legitimate player-base stand for these ridiculous color changes?

I doubt it. It looks ridiculous.

Edit2: Flag only changes when you reload client.

I suggest we add a FlagColor variable? Once found for the first time it'll then make it faster to search for the correct color than to search for a color difference…

RSColors.FlagColor, RSColors.PlayerDotColor, RSColors.NPCDotColor, etc..

… And then we could have our functions search for colors as we usually do, going through a FindColorThroughDifference function if Colour = 0

~RM

Flag is working fine for me, and I've reloaded the client several times.

function FindFlagEx(var X, Y: Integer): Boolean;
var
  AX, AY, XX, YY, W, H, Map : Integer;
  MapColors :T2DIntegerArray;
begin
  AX:= MMX2 - MMX1;
  AY:= MMY2 - MMY1;
  Map := BitmapFromString(MMX2 - MMX1, MMY2 - MMY1, '');
  CopyClientToBitmap(Map, MMX1, MMY1, MMX2, MMY2);
  SetTargetBitmap(Map);
  GetBitmapSize(Map, W, H);
  MapColors := GetBitmapAreaColors(Map, 0, 0, W - 1, H - 1);
  for XX := 0 to AX do
    for YY := 0 to AY do
      If MapColors[XX][YY] = 65536 then
         If MapColors[XX - 1][YY - 1] - MapColors[XX][YY - 1] = 6381921 then
         begin
           Result := True;
           X := XX + MMX1;
           Y := YY + MMY1;
         end;
  FreeTarget(Map);
  FreeBitmap(Map);
end; 

Takes 52 MS. To find.
Might be little bit slower, but should be more accurate.


~Home

Almost twice as fast.

function FindFlagEx(var X, Y: Integer): Boolean;
var
  i, Hi, W, H, MapB,
  KMTarget, ITarget : Integer;
  TPA: TPointArray;
begin
  MapB := BitmapFromString(MMX2 - MMX1, MMY2 - MMY1, '');
  CopyClientToBitmap(MapB, MMX1, MMY1, MMX2, MMY2);
  KMTarget := GetKeyMouseTarget;
  ITarget := GetImageTarget;
  SetTargetBitmap(MapB);
  GetBitmapSize(MapB, W, H);
  if FindColors(TPA, 65536, 0, 0, W - 1, H - 1) then
  begin
    Hi := High(TPA);
    for i := 0 to Hi do
      if ((FastGetPixel(MapB, TPA[i].x - 1, TPA[i].y - 1)
            - FastGetPixel(MapB, TPA[i].x, TPA[i].y - 1)) = 6381921) then
      begin
       Result := True;
       X := TPA[i].x + MMX1;
       Y := TPA[i].y + MMY1;
       Break;
      end;
  end;
  FreeTarget(GetImageTarget);
  FreeBitmap(MapB);
  SetKeyMouseTarget(KMTarget);
  SetImageTarget(ITarget);
end;

From: mormonman
Almost twice as fast.

function FindFlagEx(var X, Y: Integer): Boolean;
var
  i, Hi, W, H, MapB,
  KMTarget, ITarget : Integer;
  TPA: TPointArray;
begin
  MapB := BitmapFromString(MMX2 - MMX1, MMY2 - MMY1, '');
  CopyClientToBitmap(MapB, MMX1, MMY1, MMX2, MMY2);
  KMTarget := GetKeyMouseTarget;
  ITarget := GetImageTarget;
  SetTargetBitmap(MapB);
  GetBitmapSize(MapB, W, H);
  if FindColors(TPA, 65536, 0, 0, W - 1, H - 1) then
  begin
    Hi := High(TPA);
    for i := 0 to Hi do
      if ((FastGetPixel(MapB, TPA[i].x - 1, TPA[i].y - 1)
            - FastGetPixel(MapB, TPA[i].x, TPA[i].y - 1)) = 6381921) then
      begin
       Result := True;
       X := TPA[i].x + MMX1;
       Y := TPA[i].y + MMY1;
       Break;
      end;
  end;
  FreeTarget(GetImageTarget);
  FreeBitmap(MapB);
  SetKeyMouseTarget(KMTarget);
  SetImageTarget(ITarget);
end;

Yes true. :)

~Home

From: mormonman
Almost twice as fast.

function FindFlagEx(var X, Y: Integer): Boolean;
var
  i, Hi, W, H, MapB,
  KMTarget, ITarget : Integer;
  TPA: TPointArray;
begin
  MapB := BitmapFromString(MMX2 - MMX1, MMY2 - MMY1, '');
  CopyClientToBitmap(MapB, MMX1, MMY1, MMX2, MMY2);
  KMTarget := GetKeyMouseTarget;
  ITarget := GetImageTarget;
  SetTargetBitmap(MapB);
  GetBitmapSize(MapB, W, H);
  if FindColors(TPA, 65536, 0, 0, W - 1, H - 1) then
  begin
    Hi := High(TPA);
    for i := 0 to Hi do
      if ((FastGetPixel(MapB, TPA[i].x - 1, TPA[i].y - 1)
            - FastGetPixel(MapB, TPA[i].x, TPA[i].y - 1)) = 6381921) then
      begin
       Result := True;
       X := TPA[i].x + MMX1;
       Y := TPA[i].y + MMY1;
       Break;
      end;
  end;
  FreeTarget(GetImageTarget);
  FreeBitmap(MapB);
  SetKeyMouseTarget(KMTarget);
  SetImageTarget(ITarget);
end;

There's also BitmapFromClient, I think. This doesn't look bad and is what I had in mind. :p You could still use Freeze I suppose, but that would load the entire client (in the backend anyway), rather than only a part. So if you're going for speed this is nice. Good work both of you. Now we just need to know it this works well enough. :)

E: http://docs.villavu.com/simba/scriptref/bitmaps.html?#bitmapfromclient

Wow mormonman that looks great! I learned something very usefull from your code. I will probably be using your method to write FindColoredAreaTolerance :)

Also is FastGetPixel faster than GetColor?

It worked very well during my tests.

function FindFlagEx2(var X, Y: Integer): Boolean;
var
  i, Hi, W, H, MapB,
  KMTarget, ITarget : Integer;
  TPA: TPointArray;
begin
  MapB := BitmapFromClient(MMX1, MMY1, MMX2, MMY2);
  KMTarget := GetKeyMouseTarget;
  ITarget := GetImageTarget;
  SetTargetBitmap(MapB);
  GetBitmapSize(MapB, W, H);
  if FindColors(TPA, 65536, 0, 0, W - 1, H - 1) then
  begin
    Hi := High(TPA);
    for i := 0 to Hi do
      if ((FastGetPixel(MapB, TPA[i].x - 1, TPA[i].y - 1)
            - FastGetPixel(MapB, TPA[i].x, TPA[i].y - 1)) = 6381921) then
      begin
       Result := True;
       X := TPA[i].x + MMX1;
       Y := TPA[i].y + MMY1;
       Break;
      end;
  end;
  FreeTarget(GetImageTarget);
  FreeBitmap(MapB);
  SetKeyMouseTarget(KMTarget);
  SetImageTarget(ITarget);
end;

Edit: New dots function too(its 1 ms faster :0 )

function GetMiniMapDotsIn(WhatDot: String; x1, y1, x2, y2: Integer): TPointArray;
var
  MapB, KMTarget, ITarget, W, H,
  I, Dif, Radius, Hi, C: Integer;
  TPA: TPointArray;
begin
  case LowerCase(WhatDot) of
    'npc', 'n', 'yellow', 'y': Dif := 4369;
    'item', 'i', 'red', 'r': Dif := 23;
    'player', 'p', 'white', 'w': Dif := 1907997;
    'friend', 'f', 'green', 'g': Dif := 5376;
    'team', 't', 'blue', 'b', 'cape': Dif := 2171941;
    {'clan', 'c', 'orange', 'o': Dif := -1;}
  else
    srl_Warn('GetMiniMapDotsIn', '"' + WhatDot + '" is not a valid dot type', warn_AllVersions);
  end;
  MapB := BitmapFromClient(x1, y1, x2, y2);
  KMTarget := GetKeyMouseTarget;
  ITarget := GetImageTarget;
  SetTargetBitmap(MapB);
  GetBitmapSize(MapB, W, H);
  C := 0;
  if FindColors(TPA, 65536, 0, 0, W - 1, H - 1) then
  begin
    Hi := High(TPA);
    SetLength(Result, Hi + 1);
    for I := 0 to Hi do
    begin
      if ((FastGetPixel(MapB, TPA[i].x - 1, TPA[i].y - 1)
            - FastGetPixel(MapB, TPA[i].x, TPA[i].y - 1)) = Dif) then
      begin
        Result[C] := Point(TPA[I].x + x1, TPA[i].y + y1 - 2);
        Inc(C);
      end;
    end;
  end;
  FreeTarget(GetImageTarget);
  FreeBitmap(MapB);
  SetKeyMouseTarget(KMTarget);
  SetImageTarget(ITarget);
  SetLength(Result, C);
end; 

From: weequ
Wow mormonman that looks great! I learned something very usefull from your code. I will probably be using your method to write FindColoredAreaTolerance :)

Also is FastGetPixel faster than GetColor?

I used fastgetpixel because its a bitmap.

Edit: Oops on the double post

Have to agree, your FindFlag working the best for me as well.

From: mormonman
I used fastgetpixel because its a bitmap.

Edit: Oops on the double post

I think GetColor should work on the bitmap as well, with the same speed. :)

From: Wizzup?
I think GetColor should work on the bitmap as well, with the same speed. :)

Run this :)

program Default;
{$DEFINE SMART}
{$i srl\srl.scar}

const
{---SMART Setup Constants---}
  WORLD = 35;
  MEMBERS = false;
  SIGNED = true;
  SUPERDETAIL = false;
{---------------------------}

var
  bmp: integer;

begin
  Smart_Members := MEMBERS;
  Smart_Server := WORLD;
  Smart_Signed := SIGNED;

  SetupSRL;

  Bmp := BitmapFromString(100, 65, 'meJzt2lsKgzAQBdCNtH92/xtT1' +
        '1CKYNtEYyKKjxw4H8FSSC8zYyw2j64BALib9ugNXNLr+TGuD98P96' +
        'VDE/qjN3BmC5VjcK2Orgb/N7h0IG3wlarEvzq6opyKohBXKNFZUx8' +
        'NAbo5si2NyYbamTUk5JSN0hp9jwEZJ/nKxeHkXKlW0QlTbp25tH+q' +
        '1RJCkWB2zTWmVCf9xtJHCxJxpS/CsqV/2pXWutwIhCNdgBl0X5ahl' +
        'rxIs5bj1mImImJfppas4EK0IZyDR0X5bMZgL+TxEAAAgDJvqvncXA==');
  SetTargetBitmap(bmp);
  writeln(GetColor(1,1));
  writeln(fastgetpixel(bmp,1,1));
end.

mormonman, I've added your versions of FindFlag and GetMiniMapDots in to my repos. I'm running a test run over night to make sure it all works without runtime errors, and I'll commit them, along with several other bug fixes tomorrow.

E: FindFlag seems to be working, but GetMiniMapDotsIn returns 0 points, no matter which kind of dot I try.

E2: FindFlag seems to work when I test like writeln(booltoStr(flagPresent)));, but when I use it in a loop, like this, it ALWAYS returns true, even when the player has stopped walking.

  mouse(577, 131, 3, 3, true);
  while (flagPresent) do
    wait(50);

Could be something stupid I'm overlooking, it's late. >.<

E: Seems like it always returns true after it's found it once.

From: mormonman
Run this :)

program Default;
{$DEFINE SMART}
{$i srl\srl.scar}

const
{---SMART Setup Constants---}
  WORLD = 35;
  MEMBERS = false;
  SIGNED = true;
  SUPERDETAIL = false;
{---------------------------}

var
  bmp: integer;

begin
  Smart_Members := MEMBERS;
  Smart_Server := WORLD;
  Smart_Signed := SIGNED;

  SetupSRL;

  Bmp := BitmapFromString(100, 65, 'meJzt2lsKgzAQBdCNtH92/xtT1' +
        '1CKYNtEYyKKjxw4H8FSSC8zYyw2j64BALib9ugNXNLr+TGuD98P96' +
        'VDE/qjN3BmC5VjcK2Orgb/N7h0IG3wlarEvzq6opyKohBXKNFZUx8' +
        'NAbo5si2NyYbamTUk5JSN0hp9jwEZJ/nKxeHkXKlW0QlTbp25tH+q' +
        '1RJCkWB2zTWmVCf9xtJHCxJxpS/CsqV/2pXWutwIhCNdgBl0X5ahl' +
        'rxIs5bj1mImImJfppas4EK0IZyDR0X5bMZgL+TxEAAAgDJvqvncXA==');
  SetTargetBitmap(bmp);
  writeln(GetColor(1,1));
  writeln(fastgetpixel(bmp,1,1));
end.

Ah. That's because SMART is overriding the Simba GetColor method. Nothing we can do about that, until SMART adds support for EIOS.

From: Coh3n
mormonman, I've added your versions of FindFlag and GetMiniMapDots in to my repos. I'm running a test run over night to make sure it all works without runtime errors, and I'll commit them, along with several other bug fixes tomorrow.

E: FindFlag seems to be working, but GetMiniMapDotsIn returns 0 points, no matter which kind of dot I try.

E2: FindFlag seems to work when I test like writeln(booltoStr(flagPresent)));, but when I use it in a loop, like this, it ALWAYS returns true, even when the player has stopped walking.

  mouse(577, 131, 3, 3, true);
  while (flagPresent) do
    wait(50);

Could be something stupid I'm overlooking, it's late. >.<

E: Seems like it always returns true after it's found it once.

Yeah what seems to work best for me is FFlag(10) to 20 with mormonman's FindFlag, but it still seems off sometimes.

What it looks like so far is the color doesn't change unless you reload the client. I'm going to try a different approach by getting a temporary flag color and use that for FindFlag afterwards.

Doesn't one point on the flag never change color? If so you could get all the colors Around that spot in a box. Then get rid of all the color that have a duplicate. By get rid of I mean remove completely not clear duplicates that still leaves one of the color in the Array. After that you can take all the Remaining colors and do a count color tolerance for the Correct amount of colors. Then make the Colors of the Flag dynamic.

That's the general Idea Anyway. Hope I explained it properly.

From: BraK
Doesn't one point on the flag never change color? If so you could get all the colors Around that spot in a box. Then get rid of all the color that have a duplicate. By get rid of I mean remove completely not clear duplicates that still leaves one of the color in the Array. After that you can take all the Remaining colors and do a count color tolerance for the Correct amount of colors. Then make the Colors of the Flag dynamic.

That's the general Idea Anyway. Hope I explained it properly.

Ahh, yes good idea. I will try this.

Or, you could grab the color of the flag once, and keep it in memory while we are still in the same session. I believe the flag color is stable for the session as long as you stay logged in, no?
That would keep things simple. You might do a click on MMcentre once to grab the Flag color at script startup, and from that monent on do a simple findcolor like we used to do.

If the flag is on top of everything else, a FindBitmapTolerance should also work, no? (With 0 as transparent colour)

From: WT-Fakawi
Or, you could grab the color of the falg once, and keep it in memory while we are still in the same session. I believe the flag color is stable for the session as long as you stay logged in, no?
That would keep things simple. You might do a click on MMcentre once to grab the Flag color at script startup, and from that monent on do a simple findcolor like we used to do.

That's the end result I was talking about but the color Spectrum they are using currently is so wide you have to get the colors dynamicly during runtime on the start up. What I was discribing was like GetFlagColors to be used in conjunction with FindFlag.

I've looked at it more and come to a conclusion of sorts.

The flag changes color after an initial click. When you first click it its a lighter color very subtle and then shortly after stays a more darker color.

Our current methods just aren't fast enough from grabbing from the screen. Freeze should work but doesn't fully (no clue why). Currently the best FindFlag is mormonman's which I'm going to commit for now.

Edit:
Maybe not, I'm not getting delays with it. /sigh

Has anyone tried a simple findColors with tolerance and some split/sort TPA work? I don't see why that wouldn't work. I'm still convinced that the major change in colour of the flag is a bug in RS (the same with the pink dots on the MM).

From: Coh3n
Has anyone tried a simple findColors with tolerance and some split/sort TPA work? I don't see why that wouldn't work. I'm still convinced that the major change in colour of the flag is a bug in RS (the same with the pink dots on the MM).

I've been trying some things via TPA to findflag. But don't have the correct TPA sort/find functions in Simba to make it faster.

I've even been trying something like this which works in theory but the speed/functions aren't there. Also this errors if it doesn't find colors /sigh

Function FindFlag(var X, Y: Integer): Boolean;
var
  i: integer;
  TPA1, TPA2, FlagTPA, TempTPA, MTPA: TPointArray;
begin
  FlagTPA := [Point(5, 0), Point(4, 1), Point(4, 2), Point(3, 3), Point(4, 3),
              Point(2, 4), Point(3, 4), Point(4, 4), Point(1, 5), Point(2, 5),
              Point(3, 5), Point(4, 5), Point(0, 6), Point(1, 6), Point(2, 6)];
  FindColorsSpiralTolerance(MMCX, MMCY, TPA1, 65536, MMX1, MMY1, MMX2, MMY2, 0);
  FindColorsSpiralTolerance(MMCX, MMCY, TPA2, 14333, MMX1, MMY1, MMX2, MMY2, 40);

  H := High(TPA1);

  for i := 0 to H do
  begin
    TempTPA := CopyTPA(FlagTPA);
    OffsetTPA(TempTPA, Point(TPA[i].x-6, TPA[i].y-14));
    GetSamePointsATPA([TPA2, TempTPA], MTPA);
    if Length(MTPA) = Length(FlagTPA) then
    begin
      Result := true;
      X := TPA[i].x;
      Y := TPA[i].y;
      Exit;
    end;
  end;
end;

Hmm, I feel like we might be over thinking it. What method did we use previously?

This should work better.

function GetMiniMapDotsIn(WhatDot: String; x1, y1, x2, y2: Integer): TPointArray;
var
  MapB, KMTarget, ITarget, W, H,
  I, Dif, Radius, Hi, C: Integer;
  TPA: TPointArray;
  P1, P2: TPoint;
begin
  case LowerCase(WhatDot) of
    'npc', 'n', 'yellow', 'y': Dif := 4369;
    'item', 'i', 'red', 'r': Dif := 23;
    'player', 'p', 'white', 'w': Dif := 1907997;
    'friend', 'f', 'green', 'g': Dif := 5376;
    'team', 't', 'blue', 'b', 'cape': Dif := 2171941;
    {'clan', 'c', 'orange', 'o': Dif := -1;}
  else
    srl_Warn('GetMiniMapDotsIn', '"' + WhatDot + '" is not a valid dot type', warn_AllVersions);
  end;
  MapB := BitmapFromClient(x1, y1, x2, y2);
  KMTarget := GetKeyMouseTarget;
  ITarget := GetImageTarget;
  SetTargetBitmap(MapB);
  GetBitmapSize(MapB, W, H);
  C := 0;
  if FindColors(TPA, 65536, 0, 0, W - 1, H - 1) then
  begin
    Hi := High(TPA);
    SetLength(Result, Hi + 1);
    for I := 0 to Hi do
    begin
      P1 := Point(TPA[i].x + 1, TPA[i].y);
      P2 := Point(TPA[i].x + 2, TPA[i].y);
      if PointInBox(P1, IntToBox(0, 0, W - 1, H - 1))
        and PointInBox(P2, IntToBox(0, 0, W - 1, H - 1)) then
        if ((FastGetPixel(MapB, P1.x, P1.y)
              - FastGetPixel(MapB, P2.x, P2.y)) = Dif) then
        begin
          Result[C] := Point(TPA[I].x + x1 + 2, TPA[i].y + y1 - 1);
          Inc(C);
        end;
    end;
  end;
  FreeTarget(GetImageTarget);
  FreeBitmap(MapB);
  SetKeyMouseTarget(KMTarget);
  SetImageTarget(ITarget);
  SetLength(Result, C);
  RAaSTPA(TPA, 3);
end;

From: Coh3n
I'm still convinced that the major change in colour of the flag is a bug in RS (the same with the pink dots on the MM).

These are obviously not bugs. They are directed at stopping us. You have to realize this. Jagex is actively attacking us.

I'm not saying it like it's anything to be more worried about, but continuing to call these "bugs" is frankly ignorant.

I don't know if this was never fix.. but I think it may be broken AGAIN..

I noticed: When the bot clicks the minimap, flag is bright red.. When character starts moving, flag is dark red.. The same goes for the flag stick.. Is this just happening to me alone? or anyone else?

Character doesn't know that it reached the flag.. I have to use while (ismoving) do wait(1);

From: ggzz
I don't know if this was never fix.. but I think it may be broken AGAIN..

I noticed: When the bot clicks the minimap, flag is bright red.. When character starts moving, flag is dark red.. The same goes for the flag stick.. Is this just happening to me alone? or anyone else?

Character doesn't know that it reached the flag.. I have to use while (ismoving) do wait(1);

No it happens, and yes I knew about the color change and checked it. It still is same color difference though so technically it should still detect it… but randomly it doesn't, what I'm curious about is WHY…

Any suggestions not already tried?

Narcle can you commit my last function? It is incredibly more stable.

Yes I will. There's other stuff I need to commit later tonight as well will do it then.

Edit: Success!
Fixed FindFlag!!!! Added bmpFlag to global call
Fixed FindNPCChatText will now search for more then one word at a time (its not the best fix though)
Added mormonman's GetMiniMapDotsIn
Removed unneeded call in SRL for text shadow

Fastest and most accurate Flag.

function FindFlag(var X, Y: Integer): Boolean;
var
  KMTarget, ITarget, I, Hi, Map, W, H: Integer;
  TPA: TPointArray;
begin
  Result := False;
  Map := BitmapFromString(MMX2 - MMX1, MMY2 - MMY1, '');
  CopyClientToBitmap(Map, MMX1, MMY1, MMX2, MMY2);
  KMTarget := GetKeyMouseTarget;
  ITarget := GetImageTarget;
  SetTargetBitmap(Map);
  GetBitmapSize(Map, W, H);
  FindColors(TPA, 65536, 0, 0, W - 1, H -1);
  FilterPointsDist(TPA, 0, 75.5, W / 2, H / 2);
  Hi := Length(TPA);
  for I := 0 to Hi - 1 do
    if ((FastGetPixel(Map, TPA[I].X - 1, TPA[I].Y - 1) - FastGetPixel(Map, TPA[I].X, TPA[I].Y - 1)) = 6381921) then
    begin
      Result := True;
      X := TPA[I].X + MMX1;
      Y := TPA[I].Y + MMY1;
      Break;
    end;
  FreeTarget(GetImageTarget);
  FreeBitmap(Map);
  SetKeyMouseTarget(KMTarget);
  SetImageTarget(ITarget);
end;  

~Home

From: Home
Fastest and most accurate Flag.
~Home

Testing it atm.. I have to say it finds the flag much faster than before.. Problemo though:

Every time it reaches the flag, it will wait till a complete stop even if I have FFlag(10); Using SPS_WalkPath(TPointArray);



Edit:

What is with the Simba tags.. I've noticed sometimes when editing a post or quoting one, the simba tags get sooo messed up :S

From: ggzz
Testing it atm.. I have to say it finds the flag much faster than before.. Problemo though:

Every time it reaches the flag, it will wait till a complete stop even if I have FFlag(10); Using SPS_WalkPath(TPointArray);



Edit:

What is with the Simba tags.. I've noticed sometimes when editing a post or quoting one, the simba tags get sooo messed up :S

Thanks for your input, I will try re produce problem. (I don't use SPS) I will check out SPS walkpath function.


Edit: Something is wrong with SPS walkPath.




~Home

From: Home
blah

Thank you Home. SRL's function doesn't work properly. Yours however, works just fine.

From: Home
Fastest and most accurate Flag.

function FindFlag(var X, Y: Integer): Boolean;
var
  KMTarget, ITarget, I, Hi, Map, W, H: Integer;
  TPA: TPointArray;
begin
  Result := False;
  Map := BitmapFromString(MMX2 - MMX1, MMY2 - MMY1, '');
  CopyClientToBitmap(Map, MMX1, MMY1, MMX2, MMY2);
  KMTarget := GetKeyMouseTarget;
  ITarget := GetImageTarget;
  SetTargetBitmap(Map);
  GetBitmapSize(Map, W, H);
  FindColors(TPA, 65536, 0, 0, W - 1, H -1);
  FilterPointsDist(TPA, 0, 75.5, W / 2, H / 2);
  Hi := Length(TPA);
  for I := 0 to Hi - 1 do
    if ((FastGetPixel(Map, TPA[I].X - 1, TPA[I].Y - 1) - FastGetPixel(Map, TPA[I].X, TPA[I].Y - 1)) = 6381921) then
    begin
      Result := True;
      X := TPA[I].X + MMX1;
      Y := TPA[I].Y + MMY1;
      Break;
    end;
  FreeTarget(GetImageTarget);
  FreeBitmap(Map);
  SetKeyMouseTarget(KMTarget);
  SetImageTarget(ITarget);
end;  

~Home

Currently works for my scripts.

I figured it out. Its not FindFlag making it wrong. In fact because that function is slower it makes it higher chance to detect it. They make the flag disappear randomly for 100-250ms so fast you don't even see it do it (most of the time).

Tricky little bastards.

From: Narcle
I figured it out. Its not FindFlag making it wrong. In fact because that function is slower it makes it higher chance to detect it. They make the flag disappear randomly for 100-250ms so fast you don't even see it do it (most of the time).

Tricky little bastards.

I'v thought a fix for this.

Yes this is ugly but this works.

FFlag(10);
FFlag(10);

~Home :P

I have it resolved in SRL5.

Flag; Is now not working for me, but only in certain locations.

At Yanille it worked, Now I am trying to use it near Seers, and its not working at all. I used the same procedure in both places, just different colors and positions.

Procedure WalkToSpinner;

Begin
  if not LoggedIn then Exit;
  FindNormalRandoms;
  ClickNorth(True);
  Walking(2717036, 20, Point(576, 59));
  Flag;
  Walking(733019, 20, Point(611, 35));
  Flag;
  Walking(5405324, 20, Point(601, 65));
  Flag;
End;  

From: Narcle
I figured it out. Its not FindFlag making it wrong. In fact because that function is slower it makes it higher chance to detect it. They make the flag disappear randomly for 100-250ms so fast you don't even see it do it (most of the time).

Tricky little bastards.

No it doesn't?

program FlagTracker;
{$i SRL\SRL.scar}

var
  x, y: Integer;
  Found: Boolean;

begin
  ActivateClient;
  Writeln('You now have 2 seconds to place a flag');
  Sleep(2000);
  repeat
    Found := FindColor(x, y, 17378, MMX1, MMY1, MMX2, MMY2);
    MoveMouse(x, y);
    Writeln(Found);
  Until(IsKeyDown(VK_F4));
end.   

Always was spewing out True until I personally saw the flag disappear, which was when it was spewing out False. It is incredibly fast (since it is a static colour with no tolerance comparison), and I deliberately placed the flag in the top-left corner (which is the area first to be found by FindColor).

From: HardRockers
Flag; Is now not working for me, but only in certain locations.

At Yanille it worked, Now I am trying to use it near Seers, and its not working at all. I used the same procedure in both places, just different colors and positions.

Procedure WalkToSpinner;

Begin
  if not LoggedIn then Exit;
  FindNormalRandoms;
  ClickNorth(True);
  Walking(2717036, 20, Point(576, 59));
  Flag;
  Walking(733019, 20, Point(611, 35));
  Flag;
  Walking(5405324, 20, Point(601, 65));
  Flag;
End;  

Recent Simba update messed up FindBitmap making FindFlag messed up.

From: Daniel
No it doesn't?

program FlagTracker;
{$i SRL\SRL.scar}

var
  x, y: Integer;
  Found: Boolean;

begin
  ActivateClient;
  Writeln('You now have 2 seconds to place a flag');
  Sleep(2000);
  repeat
    Found := FindColor(x, y, 17378, MMX1, MMY1, MMX2, MMY2);
    MoveMouse(x, y);
    Writeln(Found);
  Until(IsKeyDown(VK_F4));
end.   

Always was spewing out True until I personally saw the flag disappear, which was when it was spewing out False. It is incredibly fast (since it is a static colour with no tolerance comparison), and I deliberately placed the flag in the top-left corner (which is the area first to be found by FindColor).

Wait what? Are you saying it does or doesn't disappear bc… I have like 50 screen shots to prove otherwise lol.

From: Narcle
Wait what? Are you saying it does or doesn't disappear bc… I have like 50 screen shots to prove otherwise lol.

Nope, it doesn't. It didn't even in Al-Kharid for me, where you said it "definitely" does. :-/