File Architecture, Naming and Backups

→ A folder structure and naming convention set up on your own files, plus a working backup of what you can't afford to lose

A desktop with 200 files is a decision you keep re-making

You know the state: a desktop or downloads folder buried in hundreds of files — Untitled.docx, final.docx, final_v2.docx, final_REAL_use_this.docx, screenshots from March, a PDF you needed once. Finding anything means scanning the mess and hoping. It costs you real time, and worse, it costs you attention — every hunt is a small drain, and the low-grade stress of "where is that file" is surprisingly constant.

Two boring systems fix it permanently: a place for everything (structure + naming) so you can always find things, and a backup so a dead drive or a wrong click doesn't erase work you can't replace. Both are set-up-once, benefit-forever. And the backup one isn't optional paranoia: drives do fail, laptops do get lost or stolen, and files do get deleted by accident — the only question is whether that happens to be a catastrophe or a shrug, and that's decided before it happens, not after.

What you'll have at the end

A shallow, consistent folder structure you can actually maintain
A naming convention that keeps files findable and self-sorting
A real backup of the things you can't afford to lose

Structure: a place for everything

The goal isn't a perfect taxonomy — it's a consistent one you'll actually keep up. Principles:

Organise by area or project, top level. A handful of clear top folders — e.g. School, Work, Personal, Projects — each with sub-folders by course/project. The exact scheme matters less than having one and using it every time.
Keep it shallow. Deeply nested folders (Stuff/2026/School/Term1/Bio/Week3/...) are as bad as no folders — you can't remember the path and burying things ten levels down hides them. Two or three levels is usually plenty; lean on search for the rest (below).
The desktop and downloads are not storage — they're an inbox. Things land there, then get filed or deleted. A quick weekly sweep (pair it with the LN-02 Sunday review) keeps the mess from returning.
One "archive" folder for done-but-keep stuff gets it out of your active space without deleting it.

Structure + your OS search together is the real system: structure for browsing and putting-away, search for fast retrieval. You don't need the perfect folder if you can find the file — but consistent naming (next) is what makes search work.

Naming: make files findable and self-sorting

A good filename tells you what something is without opening it, and sorts sensibly on its own:

Be descriptive, not doc1 or final. biology-essay-causes-of-extinction beats essay. Future you is a stranger who won't remember what notes2 meant.
Dates as YYYY-MM-DD. This is the single best naming trick: written year-month-day, files sort themselves chronologically by name (because the biggest unit comes first). 2026-07-25-report lines up correctly; 25-7-26-report and July report don't. Use it for anything dated.
Version with numbers, at the end: -v01, -v02, not final, final2, FINAL_real. Numbers sort and never lie. (For anything you version a lot — especially code — a real version-control tool is better; that's TC-06 / Git.)
Avoid spaces-and-symbols trouble: use hyphens or underscores instead of spaces, and skip / \ : * ? which some systems choke on. Lowercase is a safe default.

Consistency is the whole point — a naming rule you follow every time is what makes both browsing and search reliable.

Backups: the 3-2-1 rule

This is the part that isn't optional. The standard, and it's worth following exactly:

3-2-1 —

3 copies of anything you can't afford to lose (the original + two backups),
on 2 different types of storage (e.g. your computer + an external drive, or computer + cloud),
with 1 copy off-site (somewhere physically elsewhere — cloud storage, or a drive kept in another location).

Why each part: 3 copies so one failure isn't fatal; 2 media so a single type failing (a dead drive) doesn't take everything; 1 off-site so a fire, theft, or loss of your home doesn't destroy all copies at once.

The crucial trap: cloud sync is not a backup. Services that mirror a folder (the convenient auto-sync ones) copy your deletions and mistakes too — delete or corrupt a file and it's deleted/corrupted everywhere, instantly. Sync is great for convenience; a real backup is a separate copy that isn't automatically overwritten by your changes (true backup services keep version history and deleted-file recovery — check yours does). Ransomware is the same threat: it encrypts your files and your synced copies.

Automate it. A backup you have to remember to do won't happen. Use a backup tool or service that runs on a schedule, so it's protecting you without your attention. Then — occasionally check it actually works (a backup you've never tested restoring is a hope, not a backup).

What this means for you

A place for everything + a backup are set-up-once, benefit-forever — and the backup decides in advance whether a dead drive is a catastrophe or a shrug.
Structure: by area/project, shallow (2–3 levels), desktop/downloads as an inbox to sweep, plus an archive — paired with search.
Naming: descriptive, dates as YYYY-MM-DD (self-sorting), versions as -v01 at the end, no spaces/odd symbols — consistently.
Backups: 3-2-1 (3 copies, 2 media types, 1 off-site). Sync ≠ backup (it copies your deletions); automate it and test a restore.

Exercise (40 min, verifiable output)

Design a top-level structure (4–6 clear folders by area) and create it. Shallow.
File the current chaos: clear your desktop/downloads into the structure (or an archive); delete the junk.
Adopt a naming convention — descriptive + YYYY-MM-DD for dated files + -v01 versions — and rename a batch of important files to match.
Set up a real backup of what you can't afford to lose, following 3-2-1 as far as you can (at minimum: a second copy on a different medium, ideally one off-site/cloud with version history). Make it automatic.
Test it — restore one file from the backup to confirm it works.

✅ Finish check: a shallow folder structure with your files sorted into it, a naming convention applied to real files (with YYYY-MM-DD dating), and a working, tested backup of your important files following 3-2-1.

Summary card

Findability + backups are set-up-once systems; the backup is decided before disaster, not after.
Structure: by area/project, shallow, desktop/downloads as an inbox to sweep, plus archive — lean on search.
Naming: descriptive, YYYY-MM-DD dates (self-sorting), -v01 versions at the end, no spaces/odd symbols — consistently.
Backups: 3-2-1 — 3 copies, 2 media, 1 off-site.
Sync ≠ backup — it mirrors your deletions/ransomware. Automate a real (versioned) backup and test a restore.

Sources

Standard IT guidance — the 3-2-1 backup rule
Library/archival science — file naming conventions
Nielsen, J. — findability and information architecture

Next lesson: TL-10 — Automation (L3) Related: TC-11 Security Basics · TL-07 Calendar and Task System · TC-06 Git and GitHub · TL-08 Keyboard Shortcuts Path: related — never lose a file again

Mark it when you've got the output in hand.

← All Digital Productivity & Tools lessons