Foundations — SDLC — waterfall, V-model, iterative, agile

This page assumes nothing. If the parent note used a word or a symbol, we build it here from the ground up, in the order that each idea needs the one before it.

0. What is a "defect"? (the atom everything is about)

The picture to hold in your head: a defect is a crack. If you spot a crack in a wall you just started, you scrape it off. If you spot it after ten more layers of plaster and paint have gone on top, you must chip through all ten layers to reach it. That "layers on top" image is the whole reason SDLC exists.

Why the topic needs this: every model below is measured by how early it lets you catch the crack. No defect ⇒ no reason to prefer one model over another.

1. The five phases (the shared vocabulary)

The classic five, in order:

The picture: a staircase going down. Each step's output (a signed document, a diagram, working code) is the next step's input. Notice we number them $0,1,2,3,4$ — that number is going to become an exponent in a moment, so keep it visible.

Why the topic needs this: Requirements Engineering, Verification and Validation, and Software Testing — Unit Integration System Acceptance are all just phases being examined up close. You can't compare models until you can name their stages.

2. "Feedback" — the second atom

The picture: a loop arrow returning from a later phase to an earlier one. A model with feedback late has one long arrow that only loops back after testing. A model with feedback often has many short arrows looping back constantly.

Why the topic needs this: the parent note says every model "answers ONE question differently: when do we lock decisions, and when do we get feedback?" Feedback is one of the two dials every model turns.

3. Reading the symbols in the cost formula

The parent note drops this line:

$c_p\approx c_0\cdot k^p$

Let us earn every single symbol before trusting it.

What does $k^p$ mean?

$k^p$ is $k$ multiplied by itself $p$ times. This is an exponent. Read it as "$k$ raised to the power $p$":

$k^0=1,k^1=k,k^2=k\cdot k,k^3=k\cdot k\cdot k,\dots$

Plug in $k=10$:

4. Verification vs Validation (two words that look identical)

The picture: verification checks the product against a blueprint; validation checks it against a person's actual wish. You can perfectly verify (match the blueprint exactly) a product nobody wanted (fail validation).

Why the topic needs this: the V-Model is literally the "Verification & Validation model"; every test type (Software Testing — Unit Integration System Acceptance) is one of these two, and they get their own vault note in Verification and Validation.

5. Iteration, increment, sprint, backlog

The picture: iteration is repainting the whole rough sketch clearer; increment is adding a new room to the house. Real projects do both — "iterative and incremental."

Why the topic needs this: the iterative model and every agile method (Scrum, Kanban, Extreme Programming) are built entirely out of these four words plus the feedback loop from §2.

6. Risk (why we do the scary part first)

The picture: risk is a fog bank ahead. Iterating means walking a short way into the fog, checking the ground is solid, then walking a bit more — never sprinting blind. See Risk Management.

How the foundations feed the topic

Read it top-down: the defect and the staircase of phases give us the phase number $p$; $p$ becomes the exponent that shapes the cost curve; the cost curve plus when we get feedback is exactly the trade-off that decides which SDLC model to pick.

Equipment checklist

Cover the right side and answer aloud. If any stumps you, re-read that section before the parent note.