[OSR]Reflection Include

Quote:

---

Originally Posted by Harrier

I'm guessing you got these from rs-hacking? They quite often get the x/y the wrong way around
Edit; you should also credit who you got the hooks from if so

---

Stop guessing. The hooks are from the googlecode for this include. I switched the X and Y's, as it was uploaded switched in two different cases - Camera & Base.

Quote:

---

Originally Posted by Fitta

.

---

Thanks, not sure how that happened! I was tired last night

Quote:

---

Originally Posted by elfyyy

Thanks, not sure how that happened! I was tired last night

---

Hey, I'm thankful for all the work you put in!
Cheers :)

Thanks For the Fix man! <3 Back to finishing This script. <3

Quote:

---

Originally Posted by Rules of Joe

Thanks For the Fix man! <3 Back to finishing This script. <3

---

Out of curiosity, are you working on a script for OSR or RS3?

Walkpath screws up and wont reconise the coords (R_NearTile).
No hate just saying.;)

-Bas.

Quote:

---

Originally Posted by bas

Walkpath screws up and wont reconise the coords (R_NearTile).
No hate just saying.;)

-Bas.

---

Works fine for me

Quote:

---

Originally Posted by Frement

Out of curiosity, are you working on a script for OSR or RS3?

---

I would assume OSR if he was waiting for hooks but ya never know ;)

Quote:

---

Originally Posted by elfyyy

Works fine for me

---

Hmm strange, its randomly working once in X resets.

if R_NearTile(point(3380, 2965), 7) then
begin
CenterToBank;
end;

I've also had this with few other scripts i made that used NearTile.
All seem to bugg out on R_NearTile and walking since yesterday for me.

Quote:

---

Originally Posted by bas

Hmm strange, its randomly working once in X resets.

if R_NearTile(point(3380, 2965), 7) then
begin
CenterToBank;
end;

I've also had this with few other scripts i made that used NearTile.
All seem to bugg out on R_NearTile and walking since yesterday for me.

---

Set the version of your hooks in Hooks.simba to -1 so it updates from the server, then validate that tile is correct.

Quote:

---

Originally Posted by Interfaced

Set the version of your hooks in Hooks.simba to -1 so it updates from the server, then validate that tile is correct.

---

Done that, still not working.
EDIT i need to re-grab the coords??

edit1 seems to work, thanks!

Quote:

---

Originally Posted by bas

Done that, still not working.
EDIT i need to re-grab the coords??

edit1 seems to work, thanks!

---

Glad it was sorted out.

Mostly on OSR, but My friends have me writing privet scripts for them on RS3. As they are Comp'd and dont want to restart and transfer to OSR. -_- Its a battle!

client_base x and y hooks are swapped in rev 60. :)

Quote:

---

Originally Posted by Shatterhand

client_base x and y hooks are swapped in rev 60. :)

---

Yup, already fixed, thanks!

Quote:

---

Originally Posted by elfyyy

Yup, already fixed, thanks!

---

So is camera y and z.

Quote:

---

Originally Posted by Shatterhand

So is camera y and z.

---

Already fixed as well :P Might need to change the hooks file to a lower rev so it downloads the newest hooks!

Quote:

---

Originally Posted by elfyyy

Already fixed as well :P Might need to change the hooks file to a lower rev so it downloads the newest hooks!

---

Could you make a different version counter for this? So the users dont have to change numbers and could get fixes automatically.

Quote:

---

Originally Posted by Shatterhand

Could you make a different version counter for this? So the users dont have to change numbers and could get fixes automatically.

---

Yeah. I normally increase the include rev so it automatically downloads the updated include itself but didn't get around to it. I might add in a sub rev number for the hooks though that's a good idea.

Quote:

---

Originally Posted by elfyyy

Yeah. I normally increase the include rev so it automatically downloads the updated include itself but didn't get around to it. I might add in a sub rev number for the hooks though that's a good idea.

---

Yeh it needs to be sub version, keep the revision numbers. :)

When did they start including multipliers for the field values?? I'm old school I know but I've never seen that.

Is it the same in RS3? Is finding them as simple as iterating the instructions near the field looking for a "fmul" instruction (multiply 2 floats) ?

Quote:

---

Originally Posted by the bank

When did they start including multipliers for the field values?? I'm old school I know but I've never seen that.

Is it the same in RS3? Is finding them as simple as iterating the instructions near the field looking for a "fmul" instruction (multiply 2 floats) ?

---

I believe it was sometime around rt5 that multipliers were introduced. I currently search for imul and lmul then get the constant and field instructions just before them. They only use multipliers for ints/longs as far as I have seen. It's just multiplication overflow.

Quote:

---

Originally Posted by tls

I believe it was sometime around rt5 that multipliers were introduced. I currently search for imul and lmul then get the constant and field instructions just before them. They only use multipliers for ints/longs as far as I have seen. It's just multiplication overflow.

---

Awesome thanks for the reply, the client is structured so differently than last time I was active in this scene. Just to see how many hooks on my own updater might still even function, I decided to try running it. Hit with my first issue, they encrypt the client in an inner-pack. Was able to steal the decryption function from within one of the classes outside of the inner-pack but it was still a pain in the ass.

I'm sure you all already know all this I just wanted to /vent haha

Quote:

---

Originally Posted by the bank

Awesome thanks for the reply, the client is structured so differently than last time I was active in this scene. Just to see how many hooks on my own updater might still even function, I decided to try running it. Hit with my first issue, they encrypt the client in an inner-pack. Was able to steal the decryption function from within one of the classes outside of the inner-pack but it was still a pain in the ass.

I'm sure you all already know all this I just wanted to /vent haha

---

rs-hacking.com/forum/

Quote:

---

Originally Posted by the bank

Awesome thanks for the reply, the client is structured so differently than last time I was active in this scene. Just to see how many hooks on my own updater might still even function, I decided to try running it. Hit with my first issue, they encrypt the client in an inner-pack. Was able to steal the decryption function from within one of the classes outside of the inner-pack but it was still a pain in the ass.

I'm sure you all already know all this I just wanted to /vent haha

---

:O Are you "The Bank" from MoparIsTheBest? This guy: https://www.moparscape.org/smf/index.php?topic=299985.0

The one and only, bored and checking in on my ex-favourite scene...which it appears to be shit :(

Quote:

---

Originally Posted by tls

I believe it was sometime around rt5 that multipliers were introduced. I currently search for imul and lmul then get the constant and field instructions just before them. They only use multipliers for ints/longs as far as I have seen. It's just multiplication overflow.

---

Curious, why do you search that way?

In the example:

Code:

---

static int var = 1

var *= 23214322

---

The bytecode (AFAIK) should just be

Code:

---

getstatic ldc imul putstatic

---

so why don't you search for that? Unless I'm missing something..

Quote:

---

Originally Posted by the bank

Curious, why do you search that way?

In the example:

Code:

---

static int var = 1

var *= 23214322

---

The bytecode (AFAIK) should just be

Code:

---

getstatic ldc imul putstatic

---

so why don't you search for that? Unless I'm missing something..

---

They don't set the variable equal to the overflow value. They use it in parameters and local variables.

Code:

---

somemethod(var * 23214322)

---

so bytecode is

Code:

---

getstatic ldc imul invokevirtual

---

Also, they obfuscate the flow. It can be

Code:

---

ldc getstatic imul invokevirtual

---

Quote:

---

Originally Posted by tls

They don't set the variable equal to the overflow value. They use it in parameters and local variables.

Code:

---

somemethod(var * 23214322)

---

so bytecode is

Code:

---

getstatic ldc imul invokevirtual

---

Also, they obfuscate the flow. It can be

Code:

---

ldc getstatic imul invokevirtual

---

---

Makes perfect sense now that I see what you're talking about. It would be pointless for them to simply set the field to equal the product of the two.

Thanks for clearing that up, I really appreciate you taking the time to reply!

Hooks updated to revision 61 :)

Quote:

---

Originally Posted by elfyyy

Hooks updated to revision 61 :)

---

https://www.youtube.com/watch?v=y6Sxv-sUYtM

Hey guys, just wanted to give you guys some updates on the include and what we have been working on and what to expect going forward. We are working on getting our updater up on a server which will be updated with hooks within a few minutes of rs updating in order to have the most up to date hooks possible. I Can't guarantee that we will do this, just wanted to point it out that it is one of our goals. What I have been working on quite a bit lately is rewriting parts of the include to be more efficient and also added a completely new memory management system. I am doing all of this in a separate github located here, as to not mess up any functionalities of the googlecode's include.

The new memory management stores all global object references from the time it is first called, until termination of the script, making much better use of the already called pointer. All of the other objects will also be stored in the array, until the finishing of the individual function, which again increases the speed quite significantly. In the most extreme case, the old way of object finding, in order to gather all objects in the current loaded region which is 104x104 tiles would take around 12 seconds, using the new memory management it takes around 2.5 seconds. Npc finding used to take around 250ms to find all npc's, it now takes ~30ms.

Using this method also cuts down on potential memory leaks to the point where they aren't even a factor anymore, due to the fact that unless a old object reference is overwritten, it will always be free'd, if not during runtime, then upon termination of the script. Also, the references to the pointers are done using enums instead of integers, which are smaller(1 byte compared to 4), which also improves efficiency.

So until the includes are merged, feel free to test out the include on my github, but ofc note, I am constantly changing stuff, so it may or may not always be stable and what not.

Quote:

---

Originally Posted by elfyyy

Hey guys, just wanted to give you guys some updates on the include and what we have been working on and what to expect going forward. We are working on getting our updater up on a server which will be updated with hooks within a few minutes of rs updating in order to have the most up to date hooks possible. I Can't guarantee that we will do this, just wanted to point it out that it is one of our goals. What I have been working on quite a bit lately is rewriting parts of the include to be more efficient and also added a completely new memory management system. I am doing all of this in a separate github located here, as to not mess up any functionalities of the googlecode's include.

The new memory management stores all global object references from the time it is first called, until termination of the script, making much better use of the already called pointer. All of the other objects will also be stored in the array, until the finishing of the individual function, which again increases the speed quite significantly. In the most extreme case, the old way of object finding, in order to gather all objects in the current loaded region which is 104x104 tiles would take around 12 seconds, using the new memory management it takes around 2.5 seconds. Npc finding used to take around 250ms to find all npc's, it now takes ~30ms.

Using this method also cuts down on potential memory leaks to the point where they aren't even a factor anymore, due to the fact that unless a old object reference is overwritten, it will always be free'd, if not during runtime, then upon termination of the script. Also, the references to the pointers are done using enums instead of integers, which are smaller(1 byte compared to 4), which also improves efficiency.

So until the includes are merged, feel free to test out the include on my github, but ofc note, I am constantly changing stuff, so it may or may not always be stable and what not.

---

Please don't change too many function/procedure names!
Also, loving the work, keep it up :f:

Quote:

---

Originally Posted by Fitta

Please don't change too many function/procedure names!
Also, loving the work, keep it up :f:

---

Yeah ofc not. As of now the only thing I can think that is any different is instead of using constants for object types it will now be enums. So OBJ_GAME for example

Hooks updated to revision 62

Hooks updated to revision 63

Hooks update to revision 64

Simba Code:

function R_WaitChooseOption(Option: String; Time: Integer): Boolean;
var
  T: Integer;
begin
  Result := False;
  T := GetSystemTime + Time;
  while (GetSystemTime < T) do
  begin
    if (R_IsMenuOpen) then
    begin
      Result := R_ChooseOption(Option);
      Exit;
    end;
    Wait(20 + Random(20));
  end;
end;

Urgent!! -> as atm it always return true, fix is above.



EDIT:

Also, seems like it doesn't matter how much I post about fixes, nobody seems to add/tweak them?

Quote:

---

Originally Posted by Fitta

Simba Code:

function R_WaitChooseOption(Option: String; Time: Integer): Boolean;
var
  T: Integer;
begin
  Result := False;
  T := GetSystemTime + Time;
  while (GetSystemTime < T) do
  begin
    if (R_IsMenuOpen) then
    begin
      Result := R_ChooseOption(Option);
      Exit;
    end;
    Wait(20 + Random(20));
  end;
end;

Urgent!! -> as atm it always return true, fix is above.



EDIT:

Also, seems like it doesn't matter how much I post about fixes, nobody seems to add/tweak them?

---

I'm pretty sure that everything that has been is fixed quite promptly....

EDIT: Pushed new update

Excuse me for trying to help you guys..
I've posted quite a lot of urgent stuff and nobody has made a comment about anything, therefor I had to ask. It was nothing personal at all, just wondering. Please calm down.