Machine epsilon: conclusion

[Wade Ward]

jacob navia wrote, On 30/06/07 23:42:

As you've been told a million times, don't exaggerate.

In any case, this is about your tutorial which clearly does not explain it
a thousand times.


IMHO your tutorial would be clearer if you did show the guards. One of
many things that you should realise when writing a tutorial is that a
piece of information only presented once will in all probability not be
taken in by a significant portion of your audience. Also some of those
using your tutorial and compiler will probably then look at the header and
be puzzled by the difference.

You should also mention clearly that other values are used in other
implementations.

Who has such a crystal ball? The varieties of C, seen pejoratively, are
like cancer. People who talk about "a" cure for cancer are selling you
something or stumped at the Miss Universe Pageant. Cancer is thousands of
different diseases; the optimists hope that we do better with more of them.

 

[Keith Thompson]

Who has such a crystal ball? The varieties of C, seen pejoratively, are

No crystal ball is needed.

like cancer. People who talk about "a" cure for cancer are selling you
something or stumped at the Miss Universe Pageant. Cancer is thousands of
different diseases; the optimists hope that we do better with more of them.

Would you care to try that again, but making sense this time?

Note that there's a standard for C, which makes dealing with the
variations much easier.

 

[CBFalconer]

.... snip ...


Would you care to try that again, but making sense this time?

Note that there's a standard for C, which makes dealing with the
variations much easier.

Well, if you get busy and write (and enforce) a cancer standard,
things will get much easier for many people.

 

[Jean-Marc Bourguet]

Most Unix extensions are not in the C standard headers, rather they are
provided in additional headers.

And I wonder how many which are in the C standard headers where there
*before* C was standardized.

Yours,

 

[Army1987]

Have you already explained that "a fraction of 1" actually means
"a significand of 1.00000000000000000000000000000000000000000000000000001" ?

Actually that'd be 1 + DBL_EPSILON.
(My subconscious was implicitly using Hanlon's razor, so I felt
that saying "an exponent value of the bias" was meant to make
sense but JN failed at it, rather than that he was trolling.)

 

[Flash Gordon]

Jean-Marc Bourguet wrote, On 01/07/07 09:59:

And I wonder how many which are in the C standard headers where there
*before* C was standardized.

None, since there were not any standard headers then ;-)

 

[those who know me have no need of my name]

in comp.lang.c i read:

Why do you refer to the ISO C99 standard as "ANSI"?

well, ansi did adopt iso c99, so there is also an ansi c99.