Why fields differ across phone, Watch, web and attachments

Different carriers are not full replicas; fewer fields do not prove loss or completion

Phone Radar, paired Watch summaries, web drafts and share attachments have different data/permission boundaries. Watch receives the phone’s existing pending decision/acceptance summaries and may truncate them. Web drafts/attachments contain selected text, not a phone-ledger mirror. Start with carrier, source identity and collection, then compare time/fields. Similar appearance is not full synchronization or independent refresh.

Separate from batches and format revisions

Batch comparison inspects two exports from a source; compatibility review inspects producer changes. Here different carriers in one workflow intentionally present different scopes. Missing fields may reflect selection, truncation, absent data or no handoff, rather than sync failure. Do not fabricate values to make screens agree.

Cross-device field and scope matrix

Carrier/field Current scope Check Do not infer
Phone tasks/decisions Ledger versus pending subset; legacy fallback Whether tasks exists and current view/filter scope Few pending items means few ledger tasks
Watch items Paired-phone pending summaries Source time, read failures, reachability and truncation Full ledger or independent source refresh
Watch title/project/next Bridge prefixes up to 160/80/600 characters Return to phone for full text before deciding Every field complete or characters equal bytes
Watch payload Decode limit 50000 bytes and 200 items; bridge can trim further Inspect truncated/more-on-phone signals Guaranteed delivery of 200 items or deletion implied by trimming
Source/capture time sourceDate differs from capturedAt Judge freshness using source time Fresh packaging proves a new successful source read
Target identity workbench/task and protected source/item markers Existing phone checks resolve originals Markers are account authentication or direct Watch-write authority
Web draft Selected text in an unsent destination Actual workbench/page/draft contents Automatic Radar field update or a sent reply
Text attachment Tray text/Markdown, without original audio Actual version and later handoff/upload stage Full ledger/audio replica or automatic cross-device updating

These are source-level payload/prefix limits, not bandwidth, latency or universal pairing-success results. Swift string prefixes are not UTF-8 byte counts. Total payload size can require trimming before 200 items.

A synthetic missing-field comparison

Phone: Three sample tasks; one awaits a decision.
Watch: Only that pending summary appears, with shortened next text.
Web: No handoff selected, so no matching draft exists.
Interpretation: Carrier scopes differ; check complete source on the phone.
Not established: Other tasks completed, Watch lost data or web synchronization failed.

This is not an actual paired-device trace. It calls no Watch, workbench or server. Preserve unknowns/provenance and minimize private material; do not publish task text, feed addresses or protected identity markers.

Review order

Establish the same source/item first, then each carrier’s collection and permission scope. Inspect sourceDate, failure and truncation rather than counting display rows. Continuing on the phone or replying still requires target/busy checks, phone confirmation and server authority; Watch visibility cannot bypass them.

Editing a web/local draft does not automatically update source JSON or staged attachments. This comparison adds no sync repair, field converter or cross-device write tool.

FAQ

Should Watch display every phone tasks field?

No. It carries pending summaries rather than the full source ledger, with possible field/count truncation. Return to the phone for full context.

Does shortened next prove deleted source text?

No. Prefix and total-size limits can shorten presentation without deleting or changing the original.

Does a recent capturedAt establish a refreshed ledger?

No. Packaging and source time differ. Sync can republish an existing phone snapshot without an independent source read.

Do source/item markers permit standalone Watch approval?

No. They protect target checks, not account/server write authority or phone confirmation. This page offers no marker construction/bypass method.

Does Markdown contain original audio and all tasks?

Not as a general statement. Current note sharing is text/Markdown without original audio, not automatically a ledger snapshot.

Does editing a web draft write back to Radar?

No such sync is asserted. Check web text, phone source JSON and attachments separately. No bidirectional field sync/writeback is added.

Guides and source

Radar/Watch · Format compatibility · Originals and handoff · Follow-up evidence minimization · Start · Privacy

Evidence: Watch payload model, phone bridge/trimming, README.