liveBook · Manning

Desktop app

As explained in Fast by Default

A desktop application is software installed and executed on a general-purpose desktop or laptop operating system.

A desktop app is installed software expected to feel faster and more capable than browser or mobile alternatives. Its performance directly affects retention: lagging interaction, slow startup, and intrusive resource use prompt reviews, support complaints, version rollbacks, and evaluation of alternatives. section 1.2.3 section 3.2.4

Its critical path is user action → execution → data → result. Native UI or graphics work can combine GPU and CPU processing with large local-file operations or cloud sync; the visible outcome may be a display update, saved file, or export. Delays commonly arise when main-thread work blocks interaction, processing is unbatched, or file I/O is heavy. section 1.3.1 section 4.2

Set measurable baselines around the moments users notice:

  • Launch to a first interactive screen in under 3 seconds; show a responsive window and progress while background loading continues. section 3.2.4
  • Keep keyboard feedback within 50 ms, mouse-click feedback around 60 ms, and resizing and scrolling at 60 fps; avoid synchronous file work and heavy event handlers on the UI thread. section 3.2.4
  • Measure CPU and memory while idle and active, plus open, save, and sync duration for representative files. section 3.2.4

Long-running desktop apps need startup, memory, and cleanup designed as operational concerns. A small recurring leak can accumulate over a workday. Native frameworks such as WPF, AppKit, and GTK offer platform-specific rendering and explicit lifecycle control; cross-platform native frameworks such as Qt share application logic while targeting native code, but interfaces still face different operating-system expectations. section 5.2.3 section 5.2.3

Build size is also a desktop performance concern. Bundled browser engines or runtimes can make installers large before application code is added, increasing download, deployment, and disk-space costs. Analyze artifacts by separating framework payload from application code and assets, then make size ownership part of release work. section 8.4

Table 3.5 Desktop performance metrics
Metric
Measures
Example
Launch time
Time from click to first interactive screen
Application opens and shows content within 3s
Input latency
Delay between user action and visual feedback
Keystroke reflected on screen within 50ms
Resource consumption
CPU and memory usage during idle and active use
App uses under 1% CPU when idle
File operation time
Time to open, save, or sync files
50MB file opens within 3s

Use the metric table to turn startup, interaction, resource, and file-operation expectations into explicit desktop performance budgets. section 3.2.4

How Fast by Default covers it

How deep it goes
Connects user actions to execution, data, and results, then specifies bottlenecks, metrics, architecture choices, and build-size analysis.
Examples
Uses Slack typing lag, Figma rendering, Qt applications, and an oversized Electron installer as performance scenarios.
What it assumes
Assumes familiarity with application execution while defining the user-to-result path and recurring desktop bottlenecks.
What you can do after
Operate/configure it with measurable launch, input, resource, file-operation, lifecycle, and artifact-size performance budgets.
How it connects
Connects desktop apps with critical paths, UI threads, GPU and CPU work, file I/O, cloud sync, memory management, and build artifacts.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Fast by Default. The section links point to the parts of the book it draws on.