Command palette vs. keyboard shortcuts: what wins in a file manager

You open the file manager. You want to focus the second pane. Two designs:

1. Shortcut: ⌘2. Done.
2. Palette: ⌘⇧P, type focus pane 2, enter.

The palette is more discoverable. The shortcut is faster. Which is right depends on whether you do this 5 times a day or 50.

This is the tension in modern file-manager UI design, and the right answer is both — but with deliberate boundaries.

Where palettes earned their reputation

VS Code's ⌘⇧P is genuinely a great UX. It works because:

1. Action count is enormous: 800+ in vanilla, more with extensions. No human memorizes 800 shortcuts.
2. Long-tail usage: most actions used occasionally. Fuzzy-find is the right random-access mechanism.
3. Names are stable: Format Document doesn't suddenly become Pretty-Print Document. The palette is searchable because the names are durable.

Editor users learned the palette in 2015 with Sublime Text and have demanded it everywhere since.

Why file managers are different

A typical file manager action distribution:

The top row dominates. Anything you do 50 times a day must be one keystroke. A palette here is a regression.

What each does best

Shortcuts

• Single-keystroke high-frequency actions: pane focus, layout switch, tab create/close, navigate up, open, search.
• Muscle-memory consistency: matching macOS conventions (⌘N new, ⌘W close, ⌘T tab) means users don't relearn.
• No visual disruption: no overlay opens, no flash. The work continues.

Palettes

• Discovery for new users: "I bet there's a way to do X." Open palette, type X, find action.
• Long-tail actions: things done monthly, where memorizing a shortcut wastes mental space.
• Disambiguation: when the shortcut would conflict (e.g., ⌘D is bookmark, not duplicate? Use palette).
• Configurability discovery: shortcuts customizable from inside the palette result.

A combined design that works

The pattern many tools converge on:

⌘1–4         focus pane
⌘⌥1–4        switch layout
⌘T / ⌘W      tab create / close
⌘F           search current
⌘⇧F          search recursive
⌘P           quick-open by path
⌘K           command palette (everything else)

Top six lines: hard-coded, top-frequency. Bottom two: search and palette.

The palette in this design is the escape hatch, not the front door. Users can ignore it for months and lose nothing. Users who need it can find any action.

What a "good" file-manager palette includes

If you do build one, four categories of actions:

1. Navigation: Open path..., Go to favorite..., Switch project...
2. Layout: Single pane, 2 horizontal, 2 vertical, 4 panes, Toggle sidebar, Toggle preview
3. Selection actions: Rename, Move to trash, Compress, Get info, Reveal in Finder
4. Admin: Open settings, Reset window state, Open log file

Don't include:

• High-frequency actions already on shortcuts (creates dual-source-of-truth confusion).
• Things that aren't actions (status messages, file lists).

The naming problem

Palette quality lives or dies on action names. Bad:

• "Open"
• "Open recent…"
• "Open file…"
• "Open in default app"

Good:

• "Open selected file"
• "Open recent project"
• "Open file by path"
• "Open with default application"

The principle: say the noun and the modifier. "Open" alone is useless in fuzzy search; the user types the noun ("file", "project", "path") to disambiguate.

What we're shipping in mq-dir

• Cmd+1–4 focus, Cmd+Opt+1–4 layout switch (matches Safari/Chrome muscle memory).
• Cmd+T / Cmd+W per-pane tabs.
• Cmd+F / Cmd+Shift+F local / recursive search.
• Cmd+P / Cmd+K: roadmap. The current alpha leans on shortcuts; palette is the next major UX item.

The order matters. Shortcuts first because the high-frequency loop has to be polished before anything else. Palette second because it depends on shortcuts having stable names.

Where the palette pulls ahead in a file manager

There's one workload where palette beats shortcut even at high frequency: navigating to a remembered folder by name. ⌘O is a generic open; you still have to point at the folder. A palette can fuzzy-match proj launch and jump straight there. This is why every modern file manager that ships a palette puts "open folder by path/name" at its top.

mq-dir's roadmap has this specific palette item: a ⌘P-style quick-open that searches the union of Favorites, recent folders, and indexed paths. It complements the shortcut-driven pane focus instead of replacing it.

Discoverability vs. learnability

The deepest tension: shortcuts have to be learned; palettes are discovered. New users prefer discovery. Power users prefer recall. Most file-manager users sit somewhere in between — week 1 is discovery, month 6 is recall.

Two design moves to support both:

1. Show the shortcut next to the palette result. When the user types "new tab" in the palette and sees New tab — ⌘T, they learn the shortcut by osmosis. Six weeks later they no longer open the palette for that action.
2. Don't deprecate the menu bar. macOS users land on menus first. Ensure every action visible in the menu bar is also in the palette. The palette becomes a faster menu, not a replacement.

Takeaway

A command palette is a great addition; a replacement for shortcuts is a step backward in a file manager. The cost of opening a palette is real — it's a context switch, a small one, but multiplied across 50 actions per day it adds up to a slower tool.

The right design: shortcuts for what you do daily, palette for what you do monthly, plus a single palette-first action (open by name) that earns the keystroke at any frequency. mq-dir lives at the shortcut-first end of the spectrum; that's intentional. The palette comes when it's needed, not before.