Overview
This series examines how openGym keeps a training record under a person’s or operator’s control while still supporting guided workouts, editable history, offline use, multi-device sync, imports, backups, security, and optional AI. It follows the record from routine intent to completed evidence and derived recommendations, then tests how revisions, merges, portability, and integrations preserve or blur meaning. The result is a practical ownership model: facts stay inspectable and correctable, interpretations are recomputed, and external access remains bounded. That control is conditional on HTTPS, backups, updates, storage protection, and operator trust.
Learning outcomes
- openGym’s durable workout record remains more authoritative than targets, estimates, recovery views, or AI proposals derived from it.
- Revision checks and field-aware merges reduce silent loss across devices, but openGym does not claim perfect convergence.
- Portability is semantic: history import, plan sharing, full restore, and media transfer preserve different kinds of data.
- Self-hosting provides inspectability and control, but HTTPS, backups, updates, storage protection, and operator trust remain necessary.
Keywords
Preview every chapter · Unlock full text, audio, and video from chapter 4
PlansFAQ
What does openGym ownership actually mean?
For openGym, ownership means keeping the profile’s durable records inspectable, recoverable, movable, and correctable under a person’s or operator’s control. Derived targets and views remain recomputable rather than hidden history, while devices and integrations have separate boundaries. It does not remove host, backup, or administrator trust.
How does openGym keep two devices from overwriting each other?
openGym uses revision-based conditional writes and returns the current server copy when a stale write conflicts. The client merges routines, workouts, weigh-ins, media, and settings according to their meaning, then retries; derived values are recomputed and unfinished sessions stay local. The design reduces silent loss but has limits, including a deletion gap.
What can openGym’s AI integrations access or change?
openGym’s AI Coach receives an allowlisted, consented training projection and can propose bounded routine or week changes; it cannot edit the log or weigh-ins. The local MCP bridge is read-only and exposes selected files. Coach proposals are validated, checked for stale plan edits, approved by the user, and reversible, but model quality and provider retention remain separate judgments.
Is openGym a good fit for self-hosting?
openGym is a good fit when a reader values an inspectable, portable training record and is willing to operate HTTPS, backups, updates, storage, and access policy. Local phone mode reduces server work but gives up server-backed continuity; a self-hosted profile supports shared history while retaining operator responsibilities. It is a weaker fit for anyone who wants managed recovery, scaling, and privacy operations.