Comments on: Float

By: Tony

Tony — Wed, 12 Sep 2007 23:46:51 +0000

Not sure if anyone else caught this, but there’s a little bug where you take the first 23 bits (minus the first) for the mantissa… the first 24 bits are:
0101 1110 1000 1111 1001 0001 , while you use:
0101 1110 1000 1111 1001 001 for the mantissa.

What’s weird is why it still works.

By: Programming » So that’s what floating point numbers look like

Programming » So that’s what floating point numbers look like — Thu, 24 Aug 2006 07:21:03 +0000

[...] s, but NSSlider is floating point. MS does time in integers but Apple does it with doubles.read more | digg story No Comments N [...]

By: ditech

ditech — Sun, 13 Aug 2006 04:14:13 +0000

ditech

ditech
ditech – ditech
Long life is in store for you.
Kiss me, Kate, we will be married o’ Sunday.
– William Shakespeare, “The Taming of the Shrew”

By: Rosetta @ Null-Zero.com » Liczby zmiennoprzecinkowe

Rosetta @ Null-Zero.com » Liczby zmiennoprzecinkowe — Sat, 29 Jul 2006 03:15:43 +0000

[...] 5:15 |

PomijajÄ…c dyskusjÄ™ nad pisowniÄ… tego trudnego wyrazu, tutaj jest doÅ›Ä‡ ciekawy artykuÅ‚ na ten temat, a jeÅ¼eli kogoÅ› nie przeraÅ¼a techno- [...]

By: Anonymous

Anonymous — Wed, 26 Jul 2006 03:39:25 +0000

hello

By: Axman’s Blog » Archive du blog » links for 2006-07-26

Axman’s Blog » Archive du blog » links for 2006-07-26 — Wed, 26 Jul 2006 02:19:47 +0000

[...] ey are handled, what they are, etc. (tags: Computers math Numbers programming reference) ridiculous_fish Â» Blog Archive Â» Float Another useful article about floating point nu [...]

By: ridiculous_fish

ridiculous_fish — Tue, 25 Jul 2006 09:18:00 +0000

“Jernej ” wrote: The article is misinformation, MS does NOT do time in integers but INT64

I’ll let this one speak for itself.

By: Warrenb

Warrenb — Mon, 24 Jul 2006 16:23:11 +0000

The point of the article is to explain how floating point numbers work, not further the agenda of mac vs windows, fanboy.

By: Jernej

Jernej — Mon, 24 Jul 2006 13:18:50 +0000

The article is misinformation, MS does NOT do time in integers but INT64, which is 64 bit and reprisents 18446744073709551616 different states, which translates into:
18446744073709551.616 miliseconds
307445734561825.86026666666666667 minutes
5124095576030.4310044444444444444 days
14038618016.521728779299847792998 years…

data range:
INT64: â€“2^63..2^63â€“1, you get 14038618016 years of date represintation

however, the so called “double” that apple uses:
DOUBLE: 5.0 x 10^â€“324 .. 1.7 x 10^30 – the range it reprisents with sufficient precision matches a 32 bit UINT – range 0..4294967295 , it is no match to precision of int64 that windows and linux use which can reprisent more numbers with FULL precision.

so basicly, you can compare it like this, mac uses doubles which are precision equal to 32 bit INTEGERS, windows and linux use 64 bit INTEGERS, which are more precise.

i suggest everyone, especially the author to read THIS DOCUMENT: http://lua-users.org/wiki/FloatingPoint about FPU and precision.

By: Erick

Erick — Mon, 24 Jul 2006 10:10:38 +0000

And don’t forget that large parts of the display engine in OS X share a lot with the PDF spec, which is, as you might have guessed, based on floats. This might also be the origin of the bottom left origins in Quartz graphics versus upper left origins in QuickDraw graphics.