ios 8 problem

[ramarro123]

ios 8 problem

[MikeMcl]

How strange. Unfortunately, I can not try it myself at the moment.

What does 1e14 % 1e14 produce in the console?

Please post the result of adding console.dir(a) to your snippet so I can see the c, e and s properties of a.

[ramarro123]

modified the code as follow

<script> var a = new BigNumber(0.5); var b = new BigNumber(0.5); a = a.plus(b); console.log("printing +a"); console.log(+a); console.log("printing console.dir(a)"); console.dir(a); console.log("printing 1e1%1e1"); console.log(1e1 % 1e1); console.log("printing 1e5%1e5"); console.log(1e5 % 1e5); console.log("printing 1e7%1e7"); console.log(1e7 % 1e7); console.log("printing 1e8%1e8"); console.log(1e8 % 1e8); console.log("printing 1e9%1e9"); console.log(1e9 % 1e9); console.log("printing 1e10%1e10"); console.log(1e10 % 1e10); console.log("printing 1e11%1e11"); console.log(1e11 % 1e11); console.log("printing 1e12%1e12"); console.log(1e12 % 1e12); console.log("printing 1e13%1e13"); console.log(1e13 % 1e13); console.log("printing 1e14%1e14"); console.log(1e14 % 1e14); console.log("printing 1e15%1e15"); console.log(1e15 % 1e15); </script>

output: (i have try to 'beautify' the report)
---------------------------------------------------------------- KO (a is NaN)
[Log] printing +a (test.html, line 10)
[Log] NaN (test.html, line 11)
---------------------------------------------------------------- KO (please note c[1])
[Log] printing console.dir(a) (test.html, line 13)
[Log] BigNumber (test.html, line 14)
c: Array[2]
0: 1
1: 2.5333397e-315
length: 2
proto: Array[0]
e: 0
s: 1
proto: BigNumber
---------------------------------------------------------------- OK
[Log] printing 1e1%1e1 (test.html, line 17)
[Log] 0 (test.html, line 18)
---------------------------------------------------------------- OK
[Log] printing 1e5%1e5 (test.html, line 20)
[Log] 0 (test.html, line 21)
---------------------------------------------------------------- OK
[Log] printing 1e7%1e7 (test.html, line 23)
[Log] 0 (test.html, line 24)
---------------------------------------------------------------- OK
[Log] printing 1e8%1e8 (test.html, line 26)
[Log] 0 (test.html, line 27)
---------------------------------------------------------------- OK
[Log] printing 1e9%1e9 (test.html, line 29)
[Log] 0 (test.html, line 30)
---------------------------------------------------------------- KO!!!
[Log] printing 1e10%1e10 (test.html, line 32)
[Log] 2.65249474e-315 (test.html, line 33)
---------------------------------------------------------------- KO!!!
[Log] printing 1e11%1e11 (test.html, line 35)
[Log] 1.9230586856e-314 (test.html, line 36)
---------------------------------------------------------------- KO!!!
[Log] printing 1e12%1e12 (test.html, line 38)
[Log] 1.3428254615e-314 (test.html, line 39)
---------------------------------------------------------------- KO!!!
[Log] printing 1e13%1e13 (test.html, line 41)
[Log] 1.900263809e-314 (test.html, line 42)
---------------------------------------------------------------- KO!!!
[Log] printing 1e14%1e14 (test.html, line 44)
[Log] 2.5333397e-315 (test.html, line 45)
---------------------------------------------------------------- KO!!!
[Log] printing 1e15%1e15 (test.html, line 48)
[Log] 3.166674627e-315 (test.html, line 49)

[MikeMcl]

Thanks.

It looks like the % operator has problems with operands over 32 bits, which seems to be a surprisingly rudimentary failing of the JavaScript implementation.

I am reluctant to apply a fix to work round it here at this stage, but further details/confirmation of this issue would be welcome.

[ramarro123]

if you can share the snippet code with the fix, i will apply it locally eventually.
this issue unfortunatly prevent me to use this lib, as i work in mobile env.

I don't think that my fix is correct, probably it will be not too much fork for you to put out a fix for that, even locally or in a special version

[MikeMcl]

What are the results given for each of the following?

1 % 1e14
2147483647 % 1e14
2147483648 % 1e14
4294967295 % 1e14     
4294967296 % 1e14

[ramarro123]

here we go

[Log] printing 1 % 1e14 (test.html, line 17)
[Log] 1 (test.html, line 18)
[Log] printing 2147483647 % 1e14 (test.html, line 20)
[Log] 2147483647 (test.html, line 21)
[Log] printing 2147483648 % 1e14 (test.html, line 23)
[Log] 2147483648 (test.html, line 24)
[Log] printing 4294967295 % 1e14 (test.html, line 26)
[Log] 4294967295 (test.html, line 27)
[Log] printing 4294967296 % 1e14 (test.html, line 29)
[Log] 4294967296 (test.html, line 30)

[ramarro123]

code modified as follow

console.log("printing 1 % 1e14");
console.log(1 % 1e14);

console.log("printing 2147483647 % 1e14");
console.log(2147483647 % 1e14);

console.log("printing 2147483648 % 1e14");
console.log(2147483648 % 1e14);

console.log("printing 4294967295 % 1e14");
console.log(4294967295 % 1e14);

console.log("printing 4294967296 % 1e14");
console.log(4294967296 % 1e14);

[MikeMcl]

Oh, so it just seems to fail when the first operand equals the second operand.

1e14 % 1e14        // 2.5333397e-315 fail
(1e14+1) % 1e14    // 1 okay
(1e14-1) % 1e14    // 99999999999999 okay  

Is that correct?

[ramarro123]

confirmed!!

[Log] printing 1e14 % 1e14 (test.html, line 17)
[Log] 2.5333397e-315 (test.html, line 18)
[Log] printing (1e14+1) % 1e14 (test.html, line 20)
[Log] 1 (test.html, line 21)
[Log] printing (1e14-1) % 1e14 (test.html, line 23)
[Log] 99999999999999 (test.html, line 24)

[MikeMcl]

Your fix is correct then.

https://developer.apple.com/bug-reporting/