The post argues that Linear feels fast because it behaves less like a classic request-response web app and more like a thick client. It keeps app data in the browser, renders from that local store, applies mutations optimistically, and syncs them to the backend in the background. That cuts perceived latency to near zero on the happy path. Several people pointed out that none of this is novel. It is the same broad pattern seen in local-first apps, client-side prediction in games, Meteor's old Minimongo model, Relay optimistic updates, and IndexedDB-backed web apps.
The useful takeaway from the comments is that the hard part is not making the UI feel instant. The hard part is everything that happens after. Once writes can succeed locally and fail later, you are building a distributed system with retries, conflict resolution, rollback,
schema drift, ordering problems, offline persistence limits, and user messaging that now arrives late and out of context. That is manageable for an issue tracker because most actions are low stakes, conflicts are relatively rare, and some delay is acceptable. It becomes much less acceptable for anything transactional or time-sensitive, where a user must know whether the server has actually committed the change.
A second thread pushed back on the article's framing that "traditional
CRUD" implies 300 millisecond interactions. Many argued that this is an inflated baseline. A well-placed, well-built backend can often get into the tens of milliseconds for a lot of users, which is already close enough to instant for business software. The speed of light still matters for globally distributed teams, but commenters saw the article as overselling client-side syncing as the only route to responsiveness.
Just as notable, many people said Linear itself does not consistently match its reputation. Alongside praise from heavy users who still see it as far better than Jira, there were repeated complaints about slow first loads, high memory and CPU use, hanging sync states, confusing hidden controls, weak discoverability, and even reports of overwritten or apparently lost edits. That left the overall reaction split but sharp: the architecture is interesting and clearly powerful, yet the post glosses over the failure cases, and the real product experience seems much less universally "fast" than its legend suggests.