SRL Archive - Bugs

Known problems.

Post a thread if you've fixed a "Known problem" so we can delete it from this thread!

Developers will look into this when we find time.
Important bugs will be priorities-ed!

• None

DON'T POST ABOUT PROBLEMS HERE.
LINK TO THREADS!

Non important
R_ClickToContinue seems to find it for certain interfaces, but not a wide array of interfaces.

Interface crashes simba when doing for loops

If I do something like this:

for i:=0 to 999 do
begin
  if ValidInterface(i) then writeln (i);
end;

That code in particular might not crash it, but when I started debugging text from interface/child/child components, if I didn't throw in a 10 second wait… simba would crap out with some access violation. If I changed it from i:=0 to 999, down to i:=0 to 99, it wouldn't need a wait.

I'll look into this more

Seems to be like the TGroundItemArray overflowing that crashed simba. Except nothing is being stored. Maybe this time its too many writeln? Why do you need to go from 1 to 999?

As for R_ClickToContinue, maybe try fetching all interfaces using GetValidInterfaces instead of using a list of predetermined interfaces. Then searching them.

From: pyroryan
Seems to be like the TGroundItemArray overflowing that crashed simba. Except nothing is being stored. Maybe this time its too many writeln? Why do you need to go from 1 to 999?

I don't know what the highest possible interface number is, so I set 999 as it ;)
Maybe there is a way to find out what the highest valid interface is? I dont know if we have code for that though.

As for R_ClickToContinue, maybe try fetching all interfaces using GetValidInterfaces instead of using a list of predetermined interfaces. Then searching them.

I think it will need to get updated for that, someone hardcoded constant windows in for it :(

From: cstrike
I don't know what the highest possible interface number is, so I set 999 as it ;)
Maybe there is a way to find out what the highest valid interface is? I dont know if we have code for that though.



I think it will need to get updated for that, someone hardcoded constant windows in for it :(

couldn't you use try..except blocks to figure out where you get the out of range error and that would be the highest interface? I'm not 100% sure but that could be a possibility.

From: Inspirations
couldn't you use try..except blocks to figure out where you get the out of range error and that would be the highest interface? I'm not 100% sure but that could be a possibility.

How would that work? (if you had to verbalize it :)?)

My goal is to find out why using 0 –> 99 is stable, yet 0 –> 999 creates an access violation as soon as the script terminates.

I've not used and try/except's before, so it'd be new ground I'm covering [though Id like to learn!]

var
  i: Integer;
  TIA: TIntegerArray;

begin
  try
    for i := 0 to 10 do
    begin
      SetArrayLength(TIA, i);
      TIA[i] := i;
    end;
  except
    Writeln('uh oh..');
  end;
  Writeln('TIA: 'ToStr(TIA));
  Writeln('now without try/except(/finally(not included in this example)).');
  Wait(5000);
  for i := 0 to 10 do
    begin
      SetArrayLength(TIA, i);
      TIA[i] := i;
    end;
end.

Now thats neat! :)

I'm going to try it and see if I can get around the simba crashing issue

Fatal errors will not be stopped with try/except/finally.

Here's an up to date image of what I get when I do a DumpInterfaces
Apparently my 10 second wait at the end isn't enough to prevent a crash :(


Maybe if someone knows reflection well they could look into
"Freeing Plugins[1]"
Do not know if that has any bearing on the above problem.


Here's the code:

// This code is by Zephyrsfury
function GetChildComponentText(ParentI, ChildI, CompI: Integer): String;
var
   Text, Child, Comp: Integer;
begin
  Child := GetInterface(ParentI, ChildI);
  Comp := SmartGetFieldArrayObject(Child, hook_interface_GetChildren, compI); //Getting the children components of the Child component
  Text := SmartGetFieldObject(Comp, hook_interface_GetText);
  Result := GetJavaString(Text, 255);
  SmartFreeObject(Text);
  SmartFreeObject(Comp);
  SmartFreeObject(Child);
end;

// Do not include quotation marks if you're harvesting the string
procedure DumpAllInterfaces (maxChild, maxComponent : Integer);
var Iin, Ich, Ico : Integer;
         Sch, Sco : String;
  AValidInterface : TIntegerArray;
begin
  AValidInterface := GetValidInterfaces;
  writeln ('Number of interfaces available: '+inttostr(high(AValidInterface)));
  writeln ('-----------------------------------');
  for Iin:=0 to high(AValidInterface) do
  begin
      writeln ('<Interface #'+inttostr(AValidInterface[Iin])+'>');
      for Ich:=0 to maxChild do
      begin
        Sch := GetInterfacetext(AValidInterface[Iin], Ich);
        if Sch <> '' then writeln ('   <Child #'+inttostr(Ich)+'>: "'+Sch+'"');
        for Ico:=0 to maxComponent do
        begin
          Sco := GetChildComponentText (AValidInterface[Iin], Ich, Ico);
          if Sco <> '' then writeln ('      <Component #'+inttostr(Ico)+'>: "'+Sco+'"');
        end;
      end;
  end;
  writeln ('Waiting 10 seconds to prevent program crash'); // doesn't stop the crash
  wait (10000); // doesn't help
  writeln ('Dump All Interfaces completed.');
end;

It seems that the more interfaces you go through, the longer the wait has to be before it crashes. I assume something isn't getting freed?