v4.15: Odd "FloatToStr" behaviour (solved in v4.60)

[Dany]

Hi, the "FloatToStr" routine (Conversions lib) in v4.15 does not behave as in the help example:

Code: Select all

  
var ff1, ff2, ff3 : real;
    txt : array[23] of char;
  ...
  ff1 := -374.2;
  ff2 := 123.456789;
  ff3 := 0.000001234;

  FloatToStr(ff1, txt);  // txt is "-374.2"    // actually '-374.20001'
  FloatToStr(ff2, txt);  // txt is "123.4567"  // actually '123.45678'
  FloatToStr(ff3, txt);  // txt is "1.234e-6"  // actually '0.00000' (still so in v4.60)'

As you can see, especially the last one is odd, I did expect the "...e..." notation here.

Also, the longest string (txt) needed was of length 10, why is it specified as 23?

[filip]

Hi,

It will be fixed.
Thank you for reporting this.

Regards,
Filip.

[Dany]

Hi,

It will be fixed.
Thank you for reporting this.

Regards,
Filip.

Solved in v4.60. The help file is adapted to reflect the actual behaviour of "FloatToStr" .
Thanks mE!

[frasca]

Hi,

It will be fixed.
Thank you for reporting this.

Regards,
Filip.

Solved in v4.60. The help file is adapted to reflect the actual behaviour of "FloatToStr" .
Thanks mE!

is a solution to change the help file???
or is better to remove the bug?

[janni]

It wasn't a bug. It was description from mC left in mP help file.

[frasca]

It wasn't a bug. It was description from mC left in mP help file.

Code: Select all

FloatToStr(ff3, txt);  // txt is "1.234e-6"  // actually '0.00000' (still so in v4.60)'

This is my problem... the old version when the number is too small it becomes exponential notation!!! Good idea!! the new version return 0.00000 that is wrong!!

[janni]

I see what you mean, though there are some that would not like seeing engineering notation on their LCD just because calculation result is a little off zero due to rounding problems. Either implementation of FloatToStr may be inconvenient for some applications.
Maybe a request for two versions of FloatToStr in Wish List forum would be in place?

...the old version when the number is too small it becomes exponential notation!!!

I cannot recall - which version of mikroPascal had FloatToStr working this way?

[tihomir.losic]

It wasn't a bug. It was description from mC left in mP help file.

Hello Janni,

I agree with you.
Given floating point number will be truncated to 7 most significant digits before conversion.
For example:

Code: Select all

float ff1 = -374.2;
float ff2 = 123.456789;
float ff3 = 0.000001234;
char txt[15];
...
FloatToStr(ff1, txt);  // txt is "-374.2"
FloatToStr(ff2, txt);  // txt is "123.4567"
FloatToStr(ff3, txt);  // txt is "1.234e-6"

Best regards,

Losic Tihomir