but then you might be getting a genetically modified one, which might be
dangerous (?!).
just joking, I am attaching old fashioned float.h to this message
Boris
P.S. in that IEEE 754, you can see the answer to why and not only what are those
numbers as they are... and also what are the consequences of what they are...
Peter Jay Salzman wrote:
I already know the names of all my constants and what they mean. I wanted a
file that I could open and peek at numerical values. No gcc, no web
browser. Just some float.h that I can open with vim, less, or grep to
quickly get a value in 2 seconds.
Such a float.h used to exist on my hard drive. I guess that changed with
some gcc update.
Telling me to read the IEEE 754 standard is like telling me that I need to
know the DNA sequence for a Fuji apple when all I really want is to eat one.
Peter
On Mon 29 Jan 07, 8:58 AM, Boris Jeremic <[EMAIL PROTECTED]> said:
Then you need to read IEEE 754 (and accompanying documents...):
http://grouper.ieee.org/groups/754/
Boris
Peter Jay Salzman wrote:
On Fri 26 Jan 07, 9:04 AM, Boris Jeremic <[EMAIL PROTECTED]> said:
on my machine (fedora core 5) there are definitions in
./usr/lib/gcc-lib/i386-redhat-linux/3.2.3/include/float.h
/* Difference between 1.0 and the minimum double greater than 1.0 */
#undef DBL_EPSILON
#define DBL_EPSILON 2.2204460492503131e-16
but not in
./usr/lib/gcc/i386-redhat-linux/4.1.0/include/float.h
where it appears as:
/* The difference between 1 and the least value greater than 1 that is
representable in the given floating point type, b**1-p. */
#undef FLT_EPSILON
#undef DBL_EPSILON
#undef LDBL_EPSILON
#define FLT_EPSILON __FLT_EPSILON__
#define DBL_EPSILON __DBL_EPSILON__
#define LDBL_EPSILON __LDBL_EPSILON__
It's in neither place for me.
You can also calculate them yourself:
Yeah, I'm aware of that, but this is not what I wanted.
It's not DBL_EPSILON that I care about (I simplified my question). I
wanted
a file that I could look at and know any of the constants that are of
interest to me. It's not one constant I wanted to know, but all of them,
in
one place.
_______________________________________________
vox-tech mailing list
[email protected]
http://lists.lugod.org/mailman/listinfo/vox-tech
_______________________________________________
vox-tech mailing list
[email protected]
http://lists.lugod.org/mailman/listinfo/vox-tech
