THE GITHTML FIELD GUIDE
Understand browser cache versus a saved reading document
A cache is a reuse mechanism, not a universal promise of durable offline storage. A page may reuse some previously fetched resources while still needing others from the network. Judge offline readiness by the actual saved document and a disconnected reading test, not by the fact that the page appeared successfully during an earlier visit.
Separate the storage layers
A document workflow may involve server caches, browser resource caches, application-managed storage, and exported files. These layers serve different purposes. A resource available in one context may not be available in another app or after the application’s state changes.
HTTP caching rules govern whether responses can be reused under particular conditions. They do not by themselves establish what a document reader saves, how long it retains material, or whether it mirrors every linked dependency.
Look for partial success
A cached stylesheet can make a page appear complete while an uncached image is missing. Conversely, the text may be present but a script-driven contents panel may fail. Check the actual information needed for the task rather than judging by visual familiarity.
Also distinguish a list entry from a document body. Seeing a title or thumbnail does not demonstrate that every section and asset can be opened. Use the reading path you will rely on when disconnected.
- Separate library metadata from available document content.
- Reopen the actual reading copy without connectivity.
- Record partial failures and prepare a verified standalone fallback.
Run a realistic check
While connected, prepare the intended documents through the app’s normal workflow. Then disconnect and reopen them, navigating to later sections and required companions. Record what succeeds and what is absent.
Avoid destructive experiments such as clearing application data merely to test a hypothesis. A normal disconnected reopen is a useful first check. If retention behavior remains unclear, consult the app’s documentation or support before relying on the copy for critical work.
Choose a dependable fallback
For a bounded reading task, a complete HTML file with inline CSS and essential content already in the markup reduces dependency uncertainty. Keep provenance and version information inside the file so it remains interpretable later.
Do not turn this recommendation into a guarantee about all apps or storage conditions. Device capacity, user deletion, and application behavior still matter. The practical standard is a known reading artifact, an observed offline check, and a refresh plan proportionate to the importance of the material.
Sources and further reading
AI-assisted writing with source-linked guidance and illustrative examples. Read our editorial approach or report a correction.