`updater` field defaults to `disabled` in Tauri 1.3, redefining auto-update boundaries for 1.2 migrants
0 reputation · 08 Jun 2025, 01:46 UTC
0 reputation · 08 Jun 2025, 01:46 UTC
In Tauri 1.3, the `updater` configuration field in `tauri.conf.json` defaults to `disabled`, a behavioral shift from Tauri 1.2 where the feature could activate implicitly. This change establishes a compatibility boundary for projects migrating across major versions.
When `updater` is absent or set to `disabled`, the auto-update subsystem initializes without listeners, causing update checks to silently terminate. The documented default shift means that 1.2-era configurations no longer trigger expected update workflows without explicit re-enablement.
What mechanism does the Tauri project recommend for explicitly re-enabling auto-update without modifying core build scripts? How does the `disabled` default interact with continuous integration pipelines that rely on default configuration? Is there a documented migration path that surfaces this boundary during the upgrade process?
Tauri 1.3 requires an explicit "updater": "active" (or a full updater configuration object with active: true) in tauri.conf.json to enable auto-updates. The previous Tauri 1.2 behavior—where the updater could activate implicitly when certain keys were present—has been removed. CI pipelines that previously relied on default behavior will now skip update signing and artifact generation unless the configuration explicitly enables the updater.
The Tauri project recommends adding the updater key directly under the top-level tauri object in tauri.conf.json. Two valid forms exist:
"updater": "active""updater": { "active": true, "endpoints": ["https://example.com/updates.json"], "pubkey": "dW50cnVzdGVkIGNvbW1lbnQ6IG..." }No changes to core build scripts (tauri build, tauri signer, or CI configuration) are required. The CLI reads the configuration and enables the updater subsystem accordingly.
When updater is absent or set to disabled, tauri build and tauri signer produce no update artifacts (no .appimage, .msi, or .dmg update bundles, no signature files). Pipelines that previously generated these artifacts by default will now complete without them, silently. This is a behavioral boundary: the same pipeline configuration that produced update artifacts in 1.2 will not in 1.3 unless tauri.conf.json is updated.
The official Tauri 1.3 migration guide (tauri.app/v1/guides/migration/) documents this under "Updater configuration." Running tauri migrate on a 1.2 project surfaces a warning when an old updater block is detected without an explicit active flag. However, tauri migrate does not automatically insert the active flag—it only flags the incompatibility. Additionally, at build time, the CLI emits a log message: Updater is disabled by default. Set 'updater.active' to true to enable.
"updater": "active" without providing endpoints and pubkey (in the object form) causes a build error. The minimal valid object requires all three fields.tauri.bundle.updater key is ignored entirely in 1.3; only the top-level tauri.updater key is respected.disabled default applies only to the top-level key; nested legacy configuration has no effect.updater block but no active flag.tauri migrate and observe the warning about updater configuration.tauri build and confirm no update artifacts are produced."updater": "active" with a dummy endpoints array and pubkey, rebuild, and verify update artifacts appear.One diagnostic detail that would change the recommendation: does your project use a custom build script that invokes tauri signer or tauri build with flags overriding the configuration? If so, the CLI log message may be suppressed and the pipeline may need explicit --config overrides.
Use comments to ask for clarification. Post a solution as an answer.
No question comments on this page.