SOFTWARE VERIFICATION & VALIDATION

Learn it. Try it. Keep it.

Turn the study guide into something you can practice. A few cards, a little repetition, one concept at a time.

0 / 56Practiced

0 / 56Level 2+

56Due now

01 / SPACED REPETITION

Make it stick.

Recall before revealing. Again and Almost bring the card back after up to three other cards. Got it schedules a later review.

Up to 10 cards per round. Ratings are self-assessments. Level 2+ means the card has reached at least the second review level.

03 / LEARN BY DOING

Change the input. See why.

Four small experiments for the concepts that are easier to understand when you can manipulate them.

Cover every pair with fewer tests.

Each row is one configuration of binary factors A, B, C. Toggle rows to build a suite.

Included test rows

Challenge: cover all 12 pairwise items with four rows. Pair coverage and full-configuration coverage have different denominators.

Study guide · pages 13–15

All branches. Still missing data flows.

1   input x, y; a = 0; b = 0
I   if x > 10:
2       w = a + 1; b = 4
    else:
3       w = b + 1; a = 4
II  if y > 10:
4       z = a + w
    else:
5       z = b + w

Each path label gives the outcomes of I, then II. A pair such as a: 1 → 4 connects a definition to its use without an intervening redefinition.

Select test paths

Study guide · pages 22–23. This example contains 12 definition–use pairs.

Trace the values through both decisions.

P: if x ≤ 0 or y ≤ 0:
       x = x²; y = y²
   else:
       x = x + 1; y = y + 1
Q: if x < 1 or y < 1:
       x = x + 1; y = y + 1
   else:
       x = x - 1; y = y - 1

Load a witness:

Why is FT impossible?

P=false requires x>0 and y>0. Incrementing both makes each updated value greater than 1. Therefore Q (x<1 or y<1) must be false.

Writing algebra for a path does not prove the path is feasible. Its constraints must also be satisfiable.

Study guide · page 26. These numeric traces illustrate the symbolic path constraints.

Learn it. Try it. Keep it.

Make it stick.

Put your recall to work.

Change the input. See why.

Cover every pair with fewer tests.

Find the boundary that changes the grade.

All branches. Still missing data flows.

Trace the values through both decisions.

Keep the distinctions clear.