hxSUInt is a simple implementation of a sequentially comparative unsigned integer.
SUInt(Sequential Unsigned Integer) provides a common solution for sequence number arithmatic wrapped as an abstract unsigned integer Type.
An SUInt behaves like a typical 32-bit unsigned integer, but allows for reliable comparisons in 32-bit looping sequence space by using operators.
A sequence, in this case, provides an increment capable unsigned integer that retains its reliability to compare a sequence even in the event that the sequence reaches the max UInt(4294967295). For further clarification, consider the following code:
//MAX_UINT represents the maximum value of an unsigned integer.
var s:SUInt = MAX_UINT;
s++;If we print out the value of s, as we expect, s equals 0, however, where this begins to deviate from a typical unsigned integer, is if we make a comparison here such as:
//0 > 4294967295;
var higherSequence:Bool = s > MAX_UINT;In a typical UInt, the value here would print out as false, but since we are doing the comparison in sequential space, the value of higherSequence is actually true. This is important because it allows us to deal with infinite sequences in a finite number space(32-bits). It allows you to keep a reliable sequence order, even when an integer overflow occurs by providing a 32-bit revolving index pool.
Here's a brief demystification at how this actually works under the hood:
Take a 8-bit uint sequence space where the max unsigned integer is 255(8 bits is a little easier to visualize in this case).
We know from the previous run through that 0 > 255 = true, but what is missing from the explanation is how this assumption is made. How we reach this conclusion is actually quite simple. Lets assume s equals 64. Applying the same rules of our sequence space, we can reliably determine that 64 > 193 = true.
We do this by calculating the distance between s and s + MAX_INT and determine that any numbers preceding the sequence s, within that range, is less than s.
A particular common use case for this can be seen in the TCP protocol which relies on this type of sequence number arithmatic to ensure features like reliable packet delivery.