Christoph Hormann wrote:
> LightBeam wrote:
> 
>>
>> It's an excellent news ! I found megapov very impressive but... Where 
>> is the
>> finish_map patch that was included in mlpov 0.83 ? I don't think that 
>> patch
>> was:
>>
>> - buggy
>> - incomplete
>> - published too late to be included
>> - incompatible to POV-Ray 3.6 (I don't really know)
>> - no one found the time to add it (Really ;-)  or in beta 2 ??)
>> - did not appear useful (I don't think so...)
> 
> 
> Well - that one is difficult.  When i had a look at it i found it quite 
> badly designed (or at least unusual).  The syntax was not easy to use 
> and misleading - specifying functions instead of floats without a 
> wrapping function{} is not a good idea (*).   Also i think Mael 
> considered it incomplete - he wanted to add support for anisotropic 
> finishes IIRC.
> 
> In fact 'finish map' is a misleading name - a finish map would be 
> something like:
> 
> texture {
>   pigment { ... }
>   finish {
>     bozo
>     finish_map {
>       [0 Fin_1][1 Fin_2]
>     }
>   }
> }
> 
> (*) if it is nor clear what i mean imagine the following variation of 
> the sample in the MlPOV manual:
> 
> #declare fct = function { pigment { checker color rgb 0 color rgb 1 
> scale .1 } }
> 
> sphere { ...
>   finish {
>     uv_mapping
>     reflection { fct (1) }
>   }
> 
> Should that be an incorrect call to function fct() or a function based 
> reflection minimum and a constant reflection maximum?
> 
> Christoph
> 
I certainly understand the drawbacks with the implimentation.  Just want 
to add that I think it renders real world situations where the 
depressions in a texture have a different finish that the protrusions. 
So I find the functionality useful.  There are workarounds I know but it 
adds power and flexibility I thought.

Post a reply to this message