These are optional micro-modules outside the required lesson progression. Use them when you want a practical shortcut, safer habit or next-level pattern without changing your required journey.
Turn a common loop-plus-append pattern into compact readable Python when the transformation is simple.
prices = [799, 1499, 2499]
with_gst = [round(price * 1.18, 2) for price in prices]If the comprehension needs multiple conditions or side effects, prefer the ordinary loop.
Many Python bugs come from assuming a value is a list, dict or string before inspecting it.
print(type(payload))
print(repr(payload)[:200])
print(len(payload) if hasattr(payload, "__len__") else "no length")Inspect type, representative value and size before changing logic.
Destructuring makes object-shaped data easier to read and reduces repetitive property access.
const order = { id: 42, total: 1499, status: "paid" };
const { id, total } = order;
console.log(id, total);Avoid destructuring everything by default; keep the local scope small.
Sequential awaits are slower when operations do not depend on one another.
const [profile, orders] = await Promise.all([
loadProfile(),
loadOrders(),
]);Parallelise only independent work; preserve sequencing when one result feeds the next.
unknown forces you to prove what external data contains before using it.
function readTotal(value: unknown) {
if (typeof value !== "number") throw new Error("total must be a number");
return value;
}Use unknown for untrusted input; narrow once, then keep the rest of the function strongly typed.
satisfies checks a value against a type while preserving useful literal information.
type Plan = { name: string; seats: number };
const starter = { name: "Starter", seats: 3 } satisfies Plan;Use satisfies for configuration objects when you want validation without widening every property.
A CTE gives an intermediate result a name, making multi-stage analysis easier to verify.
WITH paid_orders AS (
SELECT customer_id, amount
FROM orders
WHERE status = 'paid'
)
SELECT customer_id, SUM(amount) AS revenue
FROM paid_orders
GROUP BY customer_id;A good CTE name should describe the rows it contains, not the SQL operation used to create them.
Window functions calculate across related rows while keeping row-level detail visible.
SELECT customer_id, amount,
ROW_NUMBER() OVER (PARTITION BY customer_id ORDER BY amount DESC) AS purchase_rank
FROM orders;Think window function when you need an aggregate-like calculation but must keep individual rows.
clamp() can scale text smoothly between phone and desktop sizes without many breakpoints.
h1 {
font-size: clamp(2.25rem, 6vw, 5rem);
line-height: .98;
}Keep the minimum readable and the maximum visually controlled; the middle value handles the fluid range.
Container queries let a component adapt to the space it actually receives rather than the whole viewport.
.card-shell { container-type: inline-size; }
@container (min-width: 36rem) {
.card { grid-template-columns: 1fr 1fr; }
}Use media queries for page-level changes and container queries when reusable components own the decision.
VBA can silently create misspelled variables unless declarations are required.
Option Explicit
Sub BuildReport()
Dim total As Double
total = Range("B2").Value
End SubTurn on Require Variable Declaration in the VBA editor so new modules get Option Explicit automatically.
Repeated worksheet reads are slow; arrays let VBA process many values in memory.
Dim data As Variant
data = Range("A2:C1000").Value
Dim i As Long
For i = 1 To UBound(data, 1)
data(i, 3) = data(i, 1) * data(i, 2)
Next iCross the worksheet boundary fewer times; do the repeated work in memory.
A strong prompt should actively search for evidence that could overturn its preferred answer.
Before finalising, give the strongest counterexample or conflicting evidence.
State whether it changes the recommendation and why.Do not ask only for pros and cons; ask what evidence would actually change the decision.
A second explicit verification pass catches unsupported claims and missing constraints more reliably than one blended instruction.
PASS 1: Produce the best answer.
PASS 2: Audit every material claim against the supplied evidence.
Label unsupported claims and revise them before returning the final answer.Verification works best when the audit has a concrete failure condition, not just “double-check this”.
Native disclosure gives keyboard and semantic behaviour without custom JavaScript.
<details>
<summary>Shipping details</summary>
<p>Ships in 2–3 days.</p>
</details>Reach for native interactive HTML before rebuilding the same behaviour with generic elements.
Custom data attributes can attach simple application metadata to elements without inventing invalid attributes.
<button data-order-id="42">Open order</button>
<script>
// button.dataset.orderId === "42"
</script>Use data-* for lightweight element metadata, not as a replacement for semantic content or a database.
git stash can park incomplete changes temporarily when you must switch context before the work deserves a commit.
git status
git stash push -m "wip: pricing experiment"
git stash listDo not let stashes become an invisible long-term backlog; name them and clear them deliberately.
When a bug appeared somewhere in a long commit range, bisect can systematically narrow the first bad commit.
git bisect start
git bisect bad
git bisect good <known-good-sha>
# test each midpoint, then mark good/badBisect is only as good as your repeatable good/bad test; automate that test when possible.
Search and navigation can make an older request irrelevant; cancellation avoids wasted work and stale UI updates.
const controller = new AbortController();
fetch(url, { signal: controller.signal });
controller.abort();Cancellation complements—not replaces—server timeouts and response ordering safeguards.
Network uncertainty can cause clients to retry a write even when the first request actually succeeded.
POST /payments
Idempotency-Key: 7d9b...
{ "amount": 1499 }Idempotency is a server contract: a client-supplied key helps only if the server stores and reuses the original outcome safely.