let’s subnet your home network // You SUCK at subnetting // EP 6

A single home network can be made more secure by splitting it into multiple smaller subnets—specifically four networks—using subnetting rather than simply “adding networks” in a straightforward way. The practical payoff is isolation: wireless IoT devices, a DMZ, and user devices no longer share one flat address space, reducing the blast radius if something on one segment gets compromised.

The method starts with the existing subnet mask and the key idea that subnetting works by “stealing” bits from the host portion of the address space to create additional network identifiers. To get four networks, the process determines how many host bits must be converted into network bits. That bit count is derived through binary math: converting the required number of subnets into the smallest power-of-two that can cover it. For four networks, two bits are enough, meaning two host bits are flipped into network bits.

Once those bits are moved, the new subnet mask is rebuilt in binary and converted back to decimal. In the example, the resulting mask becomes 255.255.192.0, which corresponds to CIDR notation /26. That single change is the foundation for everything that follows: it defines both the size of each subnet and how addresses are allocated within them.

With /26 in hand, the next step is calculating the increment—the amount by which the network portion “jumps” from one subnet to the next. The increment equals the value of the last network bit in the subnet mask, which in this case is 64. That increment determines the address ranges for each of the four subnets. The first subnet starts at 192.168.0.0 and ends at 192.168.63.255 (broadcast at .63.255), the second runs 192.168.64.0–192.168.127.255, the third 192.168.128.0–192.168.191.255, and the fourth 192.168.192.0–192.168.255.255.

The episode then ties the subnet mask to host capacity. Because /26 leaves six host bits per subnet, each network contains 2^6 = 64 total addresses. Subtracting the subnet address and broadcast address leaves 62 usable host addresses per subnet. The episode emphasizes that this is not guesswork: the same counting logic applies whenever a subnet mask changes.

Finally, it frames the workflow as a reusable four-step approach: (1) decide how many subnets are needed, (2) determine how many bits to borrow, (3) compute the new subnet mask and CIDR, and (4) use the increment to list ranges and confirm host counts. The takeaway is that even if the address class feels different (Class B or Class A instead of Class C), the mechanics remain the same—only the numbers shift. A homework challenge asks viewers to redo the same network but split it into five subnets, reinforcing the idea that the bit-borrowing and increment logic scales to new requirements.