1.1Single frame snapshot install
At the start of a 20 millisecond tick, the destination deserializes a bounded snapshot into a preallocated buffer. It publishes the completed character record before the tick ends.
Generated automatically. Not reviewed by a person. Not legal advice or an opinion.
Questions about how individual parts of claim 1 might be changed. A question may describe something different from what the claim language recites, or a version of it; no view is offered on which.
At the start of a 20 millisecond tick, the destination deserializes a bounded snapshot into a preallocated buffer. It publishes the completed character record before the tick ends.
One entity identifier addresses the character record, which holds position, posture, and equipment selections. The destination reconstructs the entity from that record and uses the equipment fields to select its rendered appearance.
A shared simulation service stores the character under a stable identifier and returns its current state to either region on demand. An atomic update changes the region responsible for simulation writes at the boundary.
After the destination acknowledges preparation, a coordinator assigns the transfer tick. The destination begins updating the character at that tick and temporarily simulates it in an overlap strip that extends into the source region.
The destination applies the final update and acknowledges its sequence number. The coordinator commits the destination as the writer, and the source removes the old record from its simulation table.
Further arrangements of the same subject matter, described here as of the publication date at the foot of this page.
A factory simulator divides its floor into cells and represents each mobile manipulator as a chassis object, an arm object, and a payload object. When the chassis reaches an approach line in its current cell, the adjacent cell builds passive records for the assembly over three simulation frames. The current cell sends changes to chassis pose, joint angles, and payload attachment as they occur, and the adjacent cell applies them to the passive records. When the chassis reaches a transfer line inside the adjacent cell, the simulator makes those records active and places the original records in the passive state.
A write ahead log carries object changes between regions in place of direct object update messages. When the character reaches the first position, the source captures a snapshot and its log offset, and the destination builds passive object records from the snapshot across multiple frames. The destination replays later log entries in sequence by object identifier, buffering entries for records still under construction. At the second position, the source stops writing and publishes its final offset, the destination replays through that offset, and the two regions exchange active and passive roles.
A simulation runs at 60 frames per second and represents each character with 128 object records. At a trigger point 20 meters before the region boundary, the receiving region captures the source snapshot and initializes 16 records per frame for eight frames in object identifier order. During construction, it applies source updates to initialized records and queues updates for the remaining records by identifier and sequence number. At a transfer point 2 meters inside the receiving region, it consumes updates through the source's final sequence number and the regions exchange roles.
Each region runs a rigid body physics solver, and a character contains separate body records linked by joint records. At the approach position, the receiving region creates passive bodies and joints over several frames with contact generation and integration disabled for those bodies. After each physics step, the source sends changes to transforms, linear and angular velocities, and joint state to the passive records. At the transfer position, the source finishes its current step, the receiver applies that step's updates, and the live character enters the receiving solver on the next step while source bodies become passive.
Each region connects character object records to a spatial audio mixer that tracks emitters by object identifier. At the approach position, the destination builds passive records across multiple frames and allocates muted mixer voices for the corresponding emitters. The source sends changes to emitter position, playback state, and sample offset with a shared simulation timestamp, and the destination applies them to its passive records. At the transfer position, the regions exchange roles and the audio mixers crossfade their outputs over 50 milliseconds from the same playback position.
A software controller measures destination frame time and uses a moving average of object creation cost to calculate how many passive records fit in each frame. At the approach position, it opens a copy session and limits the first frame to at most half of the source records so construction spans at least two frames. On subsequent frames it adjusts the batch size to the measured time budget and applies sequenced source updates once each target record exists. At the transfer position, the controller pauses the character if needed while the destination applies outstanding updates, then swaps the active and passive roles.
Both regions start from the same checkpoint and run the same fixed-step simulation for the character. The source streams tick-numbered input commands, and the destination applies them in order using fixed-point arithmetic and matching random seeds.
At the first position, the source captures a consistent snapshot of the character body, mount, and carried object. The destination allocates and fills one object record per tick for three ticks, then applies queued changes to the passive character record.
At the end of each tick with changes, the source sends each modified object identifier with a bit mask and values for the marked fields. The destination applies the packets in sequence to the corresponding passive object records.
The source records each committed object component change in a journal with a tick number, object identifier, and replacement component value. The destination consumes journal entries in order and commits each tick's entries together.
When the character reaches the configured position in the destination region, a coordinator transfers a per-character writer lease at a simulation tick boundary. The source retains its record and applies incoming updates, while the destination uses its record to run the simulation and publish changes.
Passages reproduced word for word from two earlier-published U.S. patent documents, placed beside the language of claim 1. No characterization is offered. Each passage is marked A or B for the document it comes from; ¶ numbers are the paragraph numbers in the USPTO’s published full text of that document.
The lookup table for the first representation and the lookup table for the second representation are used to produce a delta of changes that were made to the first representation. The delta can then be applied to update the second representation.A · US 2009/0187882 A1 · ¶ 6
In accordance with aspects of the disclosure, a publish/subscribe model may be operated by the data store manager 108 (one or more database servers) to provide one or more portions of scene description of a 3D virtual environment to the client(s) 106. Synchronization through the content management system 104 may be incremental with only changes to the scene description being published to subscribers.B · US 2022/0134222 A1 · ¶ 80
An empty cell means only that no passage was reproduced there.
However, changes that are made in either representation are not automatically propagated to the other representation automatically.A · US 2009/0187882 A1 · ¶ 3
A visual representation of a file is synchronized with a textual representation of a file using a same delta propagation that is used to synchronize the textual representation with the visual representation.A · US 2009/0187882 A1 · ¶ 4
Using such a rich description in a system that supports concurrent collaboration and connectivity presents significant challenges to replicating and storing scene elements with fidelity and consistency.B · US 2022/0134222 A1 · ¶ 3
The changes to structural elements may be represented procedurally to preserve the structural consistency of the content across clients while changes to non-structural elements may be represented declaratively to reduce data size.B · US 2022/0134222 A1 · ¶ 6
patent.fyi, US 2026/0355625 A1, claim 1, first published 2026-11-19T13:37:15-05:00, https://patent.fyi/US20260355625/, SHA-256 de7591bc685ff2f410dc988336d03ea64ef175f5707c85e8c86c654874a9d127