when i run your code i get 0.8.

odd.  are you using the latest version of processing and on what OS?

what happens when you do this:

float a=4.0;
float b=5.0;
float c;
c=a/b;
println(c);

(i think it's good practice to make your floats floats (eg 4.0 no 4) anyway.)

Thank you for your reply.

Your original code at the top works as expected!

It is in the Wiki in a kind of FAQ, too: Why does 2 / 5 = 0 instead of 0.4?

As I said to Tom before:

Indeed. As he said, he found out that c=4/5; so he tried to make a complete example, and perhaps hadn't test it before asking...
In short, if you assign an int to a float, there is an automatic conversion, because there is no risk of loss of precision. The reverse isn't possible without a cast, because we will loose information.

ah ha.  miscommunication.  i was referring to the original original code where Tom wrote c=a/c, not the code where Tom wrote c=4/5.

Dear all

Thank you for all your reply.

PhiLho 's estimatation is right. I have not test the code for c=a/b;
I am sorry for any misleading.

Tom Tong

Thanks for clarifying! That's an honest mistake, I would say, and people can learn from it.

ohhhhhhh. now you tell me (wink).  no harm no foul.  good show.