web.4707b008153c00f6bb2c29160@news.povray.org...
> They are in fact just as much of a hinderance to productivity as the 
> mouse.
> When you're a beginner, the mouse may help you by letting you visualize 
> and
> choose options in a menu, but once you've done that 100 times a day,
> there's no good substitute for a keyboard shortcut...

Well, one major reason I switched from POV-Ray to a "regular" 
modelling/rendering package with GUI is that there's so, so much more than 
one can do with a graphical interface. Writing script only goes so far: for 
many things, eye/hand coordination remains the best combination. Of course 
I'm speaking from a practical, not theoretical, point of view.

The cumbersome GUI you describe here is something that basically doesn't 
exist, and is somehow one of those long-standing myths about modelling 
packages in the POV-community (along with the "magical commercial plug-in 
for lazy GUI users" that creates fabulous pics out of the box just by 
pressing a key...).

To make things clear: all modern packages feature 1) customisable interfaces 
that let people create their own keyboard shortcuts and other user-specific 
GUI widgets, 2) various scripting abilities to automate the boring tasks and 
3) various development kits for developers to create plug-ins. So basic 
users get the best of both worlds: GUI for things where the eye is the best 
tool, script for automation. Power users with CG programming knowledge can 
play with the SDKs and develop their own stuff.

> Likewise, visual modellers may have some appeal by letting you choose 
> options and link
> visual boxes with arrows, but soon you'll find it cumbersome and willing 
> to
> just spell pigment { foo } rather than drop a pigment box inside a texture
> box and that inside a material box...

pigment{foo} only works as an example - sort of - because it's basic (and in 
fact a GUI package will let you create it in two clicks: one to create the 
default material, another one to choose the color in a color space). A 
complex, more useful material in POV-Ray requires a lot of scripting + trial 
and error. The ability to define your pigments, surface properties, blurred 
refraction, subsurface scattering etc. with real time visualisation using 
nodes and other GUI features isn't just handy, but a much more efficient way 
to do things, and this is particularly true if there are *** lots *** of 
parameters to play with. Just try to layer materials and apply blend modes 
to each layer without a GUI...

There's indeed some beauty in "writing" images with script. For certain 
usages, it remains a rather unique and flexible tool. There are lots of good 
things about POV-Ray and its scripting-as-modelling ability is one of them, 
but please let's keep things into perspective. Modern 3D software really 
pack a lot of power, and their GUIs are specifically designed with 
productivity and deadlines in mind.

G.

Post a reply to this message