"if" function is saying that 1 is greater than 2

[Diogetz Barbosa]

Hello, IM really confused right now because my script is saying that 1 is greater than 2... This never happened before

i will put parts of the code to try to explain:

integer ab = (integer)x.x - (integer)postarget.x;  //the result of this is 1
llSay(0,(string)ab); // just to confirm the result, it says: 1
if(ab > 2)
{

llSay(0, "Error");
}

and than it says "Error", dont know why.

if it is my mistake let me know. thanks

[Innula Zenovka]

Hmm.  This behaves as expected (i.e., says "1" but doesn't say "error", which makes me wonder what's happening in your code.

default{    state_entry()    {       integer ab =1;       llOwnerSay((string)ab);       if(ab>2){           llOwnerSay("error");        }    }}

 Can there be a typo in your original code, where you are evaluating the condition, that's not been reproduced here -- that is, you've typed here what you think you've typed in the script, not what's actually there (which I frequently do -- find my script doesn't say what I'm sure I typed).

[Jenny Siddeley]

Like Innula said, I suggest you watch closely your script, I had a similar problem once and then I realized I added a ; at the end of the if, which in your case would indeed make the script act the way you said it does.

[Diogetz Barbosa]

in the original script the x.x and the postarget.x that result on the ab are sent from a sensor in other place, but i already tested if it is correct, but what is bottering me is that the llSay, is saying ab as 1, what proves that it is one : /

dont know if this haves something to do with it, but in the original code there is another if after this one

this one is really the original one this time

 

integer ab = (integer)x.x - (integer)postarget.x;
llSay(0,(string)ab); // it says: 1

if(ab >= 2)
{
if(nx != targetname)    //nx is really different from targetname.
{

nx = "error";
}}

if(ab < 2)
{
if(nx != targetname) //nx is really different from targetname.
{

nx = "clear";
}}

llSay(0,nx);

ab is = 1 , and it says error

[Ciaran Laval]

Are you sure your integer isn't ending up being a float? Try doing 

integer ab = (integer)x.x - (integer)postarget.x;  //the result of this is 1

As a float then cast it as an integer.

[Diogetz Barbosa]

like this?

float ab = (float)(x.x - postarget.x);

if((integer)ab >= 2)
{

 

[Ciaran Laval]

Is postarget a vector? 

[Diogetz Barbosa]

yes, and x too, but im using only the x of the vectors, so it is x.x and postarget.x

[Ciaran Laval]

I think the result of part of a vector is a float, so cast the result as an integer and see if that works.

[Jenny Siddeley]

I'd suggest you copy/paste that last code you wrote on this forum inside your script and see if it works better, because based on that code, it's totally impossible for it to say "error"... unless there is something else about your script that we don't know about.

[Diogetz Barbosa]

i got a better precision on the value doing this, now my result is 1.73, but still, 1.73 is not greater than 2 and still the if says it is

[Diogetz Barbosa]

there is my problem.... its impossible to say error. but it says, the last code i posted is the one im using now, im sure the rest of the script dont matter on this part

[Innula Zenovka]

But if you replace the line 

 integer ab = (integer)x.x - (integer)postarget.x;  

 with 

 integer ab = 1; 

 it works perfectly.   So it's something about the definition of integer ab that's causing the problem, or so it seems to me.

ETA -- this works fine, too!

default{    state_entry()    {        string nx ="a";        string targetname="a";        //integer ab = (integer)x.x - (integer)postarget.x;        integer ab  = (integer)1.73;        llSay(0,(string)ab); // it says: 1        if(ab >= 2)        {            if(nx != targetname)    //nx is really different from targetname.            {                nx = "error";            }        }        if(ab < 2)        {            if(nx != targetname) //nx is really different from targetname.            {                nx = "clear";            }        }        llSay(0,nx);    }}

 What are nx and targetname?  

[Diogetz Barbosa]

i deleted this one and started another script and now it worked

thanks for the help, everyone

[161488303349]

depends on how/where you declare variable nx

pencil test shows 4 conditions (edit: of your code as written)

at start

nx = undefined (possibly "error" ??)

conditions:

1) if (ab >= 2) and (nx != targetname) then nx = "error"

2) if (ab >= 2) and (nx == targetname) then nx = undefined

3) if (ab < 2)  and (nx != targetname) then nx = "clear"

4) if (ab < 2)  and (nx == targetname) then nx = undefined

 

on condition 1) say nx = "error"

on condition 3) say nx = "clear"

on conditions 2) 4) say nx = undefined

 

is condition 4) that can seemingly? give you (1 >= 2 and nx = "error") when

nx = "error" defined at start

or on previous execution condition 1) was true and nx was undefined at start

 

+

eep cant even work out my own logic until I edit about 5 times lol (:

[Eaglewing Foxtrot]

Integers and floats are stored differently. Most program languages will cast the integer to a float before checking, but it seems this didn't. Floats are stored as a group of bits organized based on the sign, the exponent, and the mantissa (what is being raised). This means 1.0 could very well be greaer than 2, if compared as bits rather than the value. Integers are merely stored based on a predictable base 2 system, which we know as biinary. When you have to break it into pieces to get a float, the language has to compensate for that. LSL should've caught that.