When there is a nearby particle with a similar color to CSNS's deco color, it creates a spark.
The detection radius is determined by .tmp2 (like other sensors).
CSNS will not detect other CSNS particles.
Precision is set by .temp (defaults to 10 C). Each RGB channel must differ by no more than the .temp (Celcius) for it to count as "similar"
If temperature is negative, detection is inverted: Each RGB channel must differ by AT LEAST -.temp.
CSNS checks the final deco color of itself and other particles (which includes alpha blending), unless .tmp is set to 1, when it will ignore the deco color on other particles.
Properties will be the same as TSNS, except for the color (green maybe?)
Maybe if the TMP is 0, it would detect deco, and otherwise it doesn't (like FRME).
Anyway, I think this would be cool because currently, there is no way to check the color of stuff like QRTZ and CLST.
Also, there should be some way of setting the tolerance, so if the color is "pretty close" it will still count. maybe it could be set using temp.
This would be really useful for things like sorting stuff like qrtz, which is currently impossible