If you'll recall, I pointed out in the past that there was some evidence that the number was actually 103.20000458.

Well it turns out that the number really is 103.200005. But this number does not fit in 32 bits. When you take that number and truncate it to a single precision floating point value, then print it. it displays as 103.20000458. The display of the latter number is an artifact of 32-bit Poser numerical display and is not to be taken seriously.

You can see this for yourself.

Go to this Java applet, which is a 32-bit floating point binary/hex/decimal conversion tool.

http://www.h-schmidt.net/FloatApplet/IEEE754.html

Type in 103.200005, and also 103.20000458. There is no difference

Now get rid of the integer part and type in .200005, then type .20000458. The latter needs 3 less bits than the former. By not including the integer 103 in this number, you let the tool devote more bits to the fraction and you are able to see how they differ.

---

Renderosity forum reply notifications are wonky. If I read a follow-up in a thread, but I don't myself reply, then notifications no longer happen AT ALL on that thread. So if I seem to be ignoring a question, that's why. (Updated September 23, 2019)