Before you can read the parent topic, you need to own a handful of symbols and pictures. We build each one from nothing, in an order where every idea rests on the one before it. Nothing here is assumed — if the parent used it, we define it.
Why the topic needs it: hexadecimal and octal are just different-sized menus of digits. Octal's menu has 8 boxes; hexadecimal's has 16 boxes. Everything else is bookkeeping about how many boxes exist.
Look at the figure: four number lines, one per base, each ticking off its own menu. Notice hexadecimal needs letters because it runs out of the usual ten symbols before it fills 16 boxes.
Look at the figure: the digits of 352 (in decimal) each sit on a stair-step. The rightmost step is worth 100=1, the next 101=10, the next 102=100. The number is the total once you multiply each digit by its step-height and add.
Why "powers of the base" and not something else? Because each position is one full menu wider than the position to its right. After you exhaust a menu of size b in the units place, you carry into the next place — so the next place counts in groups of b, the one after in groups of b×b=b2, and so on. That is literally what bi measures.
Why the topic needs it: place value is impossible to state without it. And the reason hex and octal are special is a fact about powers: 8=23 and 16=24. Those exponents (3 and 4) are exactly the "chop sizes" — 3 bits per octal digit, 4 bits per hex digit.
Why the topic needs it: computers store everything as rows of these lights. The parent note's entire purpose is making long rows of lights readable by chopping them into groups.
Look at the figure: the eight lights of a byte are fenced into two groups of 4. Each fence's worth of lights maps straight to one red hex digit underneath. Nothing crosses the fence — that "no carrying between groups" is the whole shortcut.
The parent's "repeated division" trick uses two symbols. Both are about splitting a number into whole part and leftover.
Why the topic needs it: the leftover marbles (Nmodb) are always a valid single digit of base b (it's between 0 and b−1), and it's exactly the units digitd0. Remove it (keep the full cups) and repeat to peel off the next digit. That is why repeated division reproduces the digits — bottom-to-top.
Read it top-down: digits and base define what a numeral is; powers and place value tell you what each position is worth; summation totals them (base→decimal); bits + grouping give the power-of-2 shortcut; floor/remainder give decimal→base. All arrows feed the parent topic.