As of now, it appears that fog decay syncing from server to client has a precision of 1/127 (8-bit); any fog decay value that does not match a multiple of 1/127 is rounded to the nearest value when being transferred to the client.

This is quite problematic when trying to achieve precise values below 0.1, especially since that <b>the visual effect of fog decay varies extremely greatly for values between 0 and 0.01</b>.

Here are some of the readings I took:

<pre>

server         client

0        0                   
0.0025   0                   
0.005    0.00787401 (1/127)  
0.0075   0.00787401 (1/127)  
0.01     0.00787401 (1/127)  
0.0125   0.015748 (2/127)    
0.015    0.015748 (2/127)    
0.0175   0.015748 (2/127)    
0.02     0.023622 (3/127)    
0.0225   0.023622 (3/127)    
0.025    0.023622 (3/127)    
0.0275   0.023622 (3/127)    
0.03     0.0314961 (4/127)

</pre>
Please make it so the fog decay is properly transmitted as 32-bit float over the network, not as a rounded 8-bit integer divided by 127 on the other side.
Thanks.