Guides

Getting Ready for a Day With No Connection

Install the app, open what you plan to study, download the packs that matter, and know exactly which three things will still refuse to work.

4 min read

Applies to Novus Learn 0.1.0

Illustration: An installed application window beside a dashed cache boundary listing what is stored and what is deliberately excluded.

Installing is not the preparation

Copy link

Installing the site as an app takes one menu item: in Chrome or Edge, choose to install the site as an app; on iPhone or iPad, use Share and Add to Home Screen. There is no store and no download. It is the same site in its own window, with an icon, and the normal website keeps working where installation is unavailable.

That step is worth doing, and it is not the part that determines whether your offline day works. The cache holds public pages this browser has already visited, so the preparation that matters is opening things.

Open what you intend to study, before you need it

Copy link

A page this browser has never loaded cannot be served offline, and no amount of installing changes that. So the useful preparation is boring: open the lessons, the suite pages, the skill pages and the topics you plan to work through, while you still have a connection. Ten minutes of clicking buys a whole day of reading.

For a source-grounded topic there is an extra reason to do it early. Opening a topic builds the project in this browser, and from that moment the guided lesson, the flashcards, the quiz and the Knowledge Canvas all work with no network at all, because they read the project rather than the source.

The same applies to anything you were going to upload. Parsing a file happens on your device, so a document you import while connected and a document you import on a train behave identically, but the model used for speech transcription downloads once and is large. If a lecture recording is part of the plan, transcribe it the night before rather than discovering the download at the point you have no connection to make it with.

Download the packs that actually matter

Copy link

For anything you genuinely need, do not rely on the cache. Download the study guide. A PDF or plain-text pack is an ordinary file on your device and needs nothing to read, including this site, including the browser you downloaded it in.

A practical check before you travel: disconnect, open the exact file you saved, and jump to a heading near the end. That verifies the local copy, which is a different thing from verifying that the download animation finished.

Three things that will still fail

Copy link
  • Opening a new source: a search, a pasted URL, a site index, a paper lookup or a dictionary entry.
  • Starting or submitting a formal practice attempt, which is scored on the server so the answer key never reaches the browser.
  • Any page this browser has never loaded before.
Illustration: The cache holds public pages you have visited. What it refuses to hold is the more interesting half.
The cache holds public pages you have visited. What it refuses to hold is the more interesting half.

What the cache deliberately refuses to hold

Copy link

Private matcher results, practice attempts, local project links, API responses, authenticated requests and downloads are all excluded from the offline cache on purpose. Your device-local data keeps using the browser's own storage rather than being copied into the service worker.

That exclusion list is the reason an installed app here is no more exposed than a tab. A cache that helpfully stored your result pages would be a second copy of private material in a place most people never think to clear.

A checklist for the night before

Copy link
  • Install the app, so you get a window rather than a tab that might be closed to reclaim memory.
  • Open every lesson, suite and skill page you intend to work through.
  • Open each topic you want to study, which builds its project locally.
  • Download the study guide for anything genuinely important, in a format you can read.
  • Disconnect and open one saved file and one cached page, to confirm both routes work.

Why the caching rules are conservative

Copy link

It would be technically easy to cache more. Precaching the whole catalogue would make almost everything work offline, and it would also mean a first visit downloaded several thousand pages that the visitor did not ask for, on a connection that might be metered. Caching result pages would be convenient and would put private material in a second place people forget to clear.

So the rule is that the cache holds what you actually visited, and refuses the private routes entirely. That produces a slightly worse offline experience for somebody who did not prepare and a considerably better one for everybody's privacy and data allowance. It is the same principle as the rest of the product: the useful default is the one that assumes nothing on your behalf.

When the connection comes back

Copy link

Nothing needs syncing, because nothing was queued. Work done offline was written straight to this browser's storage the same way it would have been online, and the next load picks up the newest version of the app.

The one thing worth doing on reconnection is any formal practice attempt you deferred, and any export you meant to take and did not.

Consent version 2026-08-21.1

Cookie preferences