logarithmic slider support (feature request)

[3dickulus]

logarithmic slider support (feature request)

[3Dickulus]

requires rewriting FloatWidget(s)

[3Dickulus]

...and adding slider type linear, default, and logarithmic, optional, to the parser, this will probably affect easing curves when applied to a slider.

[3Dickulus]

re:useless slider
with such a large number the slider step multiplier needs to be adjusted, right click on the Zoom slider and enter 1000 or 10000, that will make the arrow keys and page-up/down keys working.
The step value is automagicaly calculated for all other sliders but gets set to 1 for Zoom slider in 2D thus rendering it useless... investigating...

[claudeha]

step multiplier could be relative to current value, eg 1% for fine adjustment, 10% for slightly coarser, etc.

implemented by *= 1.01 and /= 1.01 for up and down

[3Dickulus]

step multiplier is controled by Shift+wheel but doesn't work in the double test frags, only in standard float frags. not sure why? the code mechanism should not be affected by float/double type frags as it's all in the C++ code not GLSL, hmmm, something else to investigate... fixed?

I have a beta logarithmic slider version, enabled via log_ prefix on slider name, but only works halfway, by the time the slider is dragged to just over half the spinner hits max.
just pushed,
using...
#define USE_DOUBLE
#include "MathUtils.frag"
#include "Complex.frag"
#include "Progressive2D-4.frag"

[3Dickulus]

look at ThirdParty/qdoublerangeslider.cpp there is some settings and calculations for logarithmic range sliders but I think it may have been left in an unfinished state, I don't recall exactly where I found that code, an example somewhere? maybe can subclass QSlider and add the log bits?

[claudeha]

Working on it... approach is to add a logarithmic slider bool flag to ComboSlider, and plumb it through from the parser for float vec2 vec3 vec4 slider. Default will remain Linear, syntax will be like:

uniform float Zoom; slider[1e-6,1,1e16] Logarithmic NotLockable

Seems cleaner than having to modify uniform names.