Read: Finding and Showing Data Back
Why 'just display the data' is never just displaying the data — the hidden decisions in every list, feed, and detail screen: scope, permissions, sorting, and shape.
The Verb That Feels Passive
Create, Update, and Delete all change something — Read just shows it. That makes Read feel like the safe, simple verb: no risk of data loss, no risk of corruption. It’s the one stakeholders are most likely to wave through in a review without questions.
That instinct is worth resisting. Read decides what a person is allowed to know, and getting that wrong is its own category of serious problem — a customer seeing another customer’s invoice, a junior employee seeing executive salary data, a competitor scraping your entire product catalog through an unprotected search box.
The Four Questions Behind Every Read
Whenever data is displayed on screen, four decisions were made to get it there:
- Scope — out of everything in the system, which records are even candidates to show?
- Permission — of those candidates, which is this user actually allowed to see?
- Shape — which fields are shown, and which are hidden or masked (like a partial credit card number)?
- Order and volume — what order are results shown in, and how many at once?
A “simple list of orders” screen has already answered all four: it scoped to one customer’s orders, checked that the logged-in user is that customer, decided to hide internal cost data, and sorted newest-first with a page size of 20.
Scope vs. Permission: A Critical Distinction
These two are easy to conflate but are different failure points:
| Scope | Permission | |
|---|---|---|
| Question | What data could this screen possibly show? | Is this specific user allowed to see it? |
| Example | ”All orders in the last 90 days" | "…but only orders belonging to this logged-in customer” |
| Failure if wrong | Screen is slow or shows irrelevant data | Screen leaks someone else’s private data |
A scope bug is usually a performance or usability problem. A permission bug is usually a security incident. When reviewing a Read feature, always ask both questions separately — “what’s the query looking at” and “who’s allowed to see the result” — because a spec can get the first right and the second wrong.
A common real-world bug: an engineer writes a perfectly correct query to fetch “order #4521,” and it works — for the customer who placed it. The bug is that the same URL, with the number changed to #4522, shows someone else’s order, because nobody checked whether the logged-in user actually owns that order before returning it. This class of bug (sometimes called an access control or “IDOR” issue) is common precisely because the Read itself works fine — the missing piece is the permission check layered on top of it.
Shape: What Gets Shown, and What Gets Hidden
The same underlying record often needs to be displayed differently depending on who’s looking. A support agent viewing a customer’s account might see the last four digits of a card on file; the customer sees the same thing; a random logged-out visitor should see none of it. Deciding the “shape” of a Read result — which fields show, which are masked, which are omitted entirely — is a per-audience decision, not a single fixed answer.
Order and Volume: Why Pagination Is a Product Decision
“Show all the customer’s orders” sounds simple until a customer has 4,000 orders. Someone has to decide:
- Sorted by what — newest first? Highest value first?
- Shown how many at a time — 20? 100? All of them?
- What happens when there are more results than fit on one screen — does the user click “Next,” scroll to load more, or search/filter instead?
These aren’t implementation details to leave to engineering — they shape whether the feature is actually usable, and they belong in the spec.
Take the “read” screen you identified in Module 1 (a list, feed, or detail view). Answer:
- Scope — what’s the full set of data this screen could show?
- Permission — how does it narrow that down to what you specifically are allowed to see?
- Shape — is anything hidden, masked, or simplified compared to the full record?
- Order and volume — what order is it sorted in, and how much loads at once?
Keep your answers — Module 6’s capstone reuses this same screen.
A feature request says: 'Managers should be able to see a dashboard of their team's performance.' What question exposes the biggest hidden assumption in this request?
Reveal Answer & Explanation
“What happens when a manager’s team changes — do they lose access to a former team member’s data immediately, and do they gain access to a new team member’s history retroactively?”
The request assumes “their team” is a stable, well-defined thing, but scope and permission for Read features are rarely static — org charts change, people transfer, managers get promoted. If the dashboard’s permission check is based on a manager’s current team but a report pulls historical performance data, there’s an implicit decision buried in the request about whether past data follows the person or the point-in-time reporting relationship. That’s exactly the kind of scope-versus-permission question this lesson asks you to separate out before a feature ships.
Up Next
The next lesson covers what happens when these four decisions go wrong in practice — the specific Read bugs that generate support tickets and, occasionally, security incidents.