Richard Burts Mon, 06/11/2007 - 17:57

I also do not know for sure why 16 and 32 were used. But given the nature of binary arithmetic my history as a programmer says that double word binary boundaries scales better than single word binary boundaries. And the choice of 16 is a logical choice for the lower value (giving 65535 as the maximum hello interval). And since the dead interval must be larger, in double word values 32 is the next logical choice (any other choice probably wastes bits in storage of values).

And I note that if you look at many of the early RFCs designing IP operations there are many values that are stored in binary as double words.




This Discussion