Int or float for iobox

hi, as iobox are basics,
it would be great to set (without using I(spreads) or mapping the iobox) with inspektor on this 3 types:
double
float
int

could be simplier for patchs ;-)

agreed. vvvv internally only uses doubles, so what would you expect to happen when selecting float?

oooops … i was thinking we were in float…
;-)

floats (i.e. doubles) or ints would be a good choice to have

Also, on a side point. When i type in a high accuracy number in vvvv it rounds to 4 decimal places. 32bit floats (single floats) are generally accurate to 7 significant figures. Does vvvv store the accurate value that I entered? or does it store the rounded value?

But Int would be nice… without using a getslice node to force that.

+1 to that!

bump… no text …

vvvv stores the accurate value but usually displays only the first four digits. obvsiously this is somehow stupid at times.

You could always use FormatValue (String) to convert the numbers to a string with more precision. I find myself sometimes putting a multiplication or division after an IOBox (Value Advanced) as a workaround.

If you open the .v4p file in an ascii or xml editor you will also see the full precision.

a nice feature, regardless of the display of the values, would be to copy the full precision number to the clipboard when double-clicking an IOBox and pressing ctrl-c.

** so what would you expect to happen when selecting float?

would be a nice feature to tell iobox how many digits after comma should be displayed… 0 digits would be int then. no need for force int modules then anymore…