Star Citizen Technology Q&A: Social App, Freight Elevators, Desync and AI

Calling a vehicle, meeting a friend and moving cargo should be the uneventful part of a night in Star Citizen. Too often, they become the evening’s main challenge. The 23 October Technology in Focus Q&A is interesting because it spends much of its time on those ordinary interactions: why they break, what needs replacing underneath them, and where another feature has to wait for the foundations to improve.

Across ten community questions, the developers discuss a new Social App, ground vehicles on freight elevators, desync, AI, ship ownership, cockpit visibility and driving. There is plenty to look forward to, but this is an engineering discussion rather than a release announcement. A system being technically possible is not the same as its design being approved, and a design being finished does not put it in the current patch.

view on youtube

Why the same faults keep coming back

Freight elevators, hangars, ASOP terminals and air traffic control sit near the centre of the regression discussion. A fix that works in one straightforward sequence can still fail when multiple players, server recovery or an unexpected interruption enter the picture. The proposed response is deeper root-cause investigation alongside broader tests before changes are submitted.

That includes bringing more multiplayer and recovery scenarios into earlier testing, rather than discovering failures only after a change has already moved further through development. Automated checks and content validation are also part of the effort to reproduce more of the circumstances found on Live.

For players, the important result would be fewer nights lost to a basic interaction failing halfway through a session. The Q&A describes the work intended to get there; it does not demonstrate that recurring elevator or terminal problems have already been eliminated.

Scaleform is nearly gone from the UI—but not entirely

Most of the interface now uses Building Blocks. Remaining pieces discussed include visor elements, Commlink and the Vehicle Manager App, with the latter due to give way to the vehicle loadout system. Removing the remaining Scaleform interface work could also remove a rendering pass and simplify the UI technology stack.

There is an easily missed qualification: even Building Blocks still relies on Scaleform for font rendering. A replacement for that work is complete but has not yet been rolled out. “The interface uses Building Blocks” therefore does not mean Scaleform has already vanished from the game.

Passing a magazine should feel simple

Handing an object directly to another player sounds like a small animation feature. In practice, it asks two independently controlled characters to meet at the right place and time, while deciding who owns the interaction and whether the recipient actually wants it. Preventing someone from simply snatching an item creates another design question.

The answer is encouraging about technical possibility, but it does not announce a finished handover system. It is a useful example of why an action that looks effortless between two scripted characters becomes more complicated when both people can walk away, turn around or try something else.

The Social App is further than a design sketch

The first planned social drop centres on a chat revamp and a Social App to replace Commlink. Regional and location channels are intended to suit larger shards, alongside richer presence information and a revised chat widget. Configurable channels and tabs should give players more control over the conversation they are following.

At the time of the recording, the design is complete, game implementation is about to begin, and UI work is already underway. Players will have to wait for the new app, but the team has moved beyond deciding what it should do.

Backend social work has continued while resources have also been needed for stability. VOIP has received incremental backend changes across recent releases, and 4.10.2 includes more robust face synchronisation. The developers nevertheless acknowledge how awkward calls can be to discover and use, including the lack of proper notifications.

For a group trying to reunite after a death or reconnect, those everyday details matter. The discussion touches on communication while downed or respawning, but that is not a promise that every proposed chat behaviour will arrive in the first drop. Until the new system ships, our VOIP and FOIP guide covers the current tools.

LAMP needs a solution that fits more than fighters

LAMP’s cockpit-visibility effect was easier to apply first to fighters with separately defined exterior canopies. Other ship structures do not expose the same material arrangement in a way the current code can reliably find. Expanding the feature is therefore not simply a matter of switching it on for every cockpit.

The aim is to solve that without sending artists back through more than a hundred vehicles to reauthor each one. The stated direction is broad support, including ground vehicles, but the necessary work still needs planning. There is no delivery date in the answer.

Ground vehicles on freight elevators remain the plan

Being able to bring up a rover inside your own hangar would remove a familiar detour from preparing a ship. The developers still want it, but identify two dependencies: separating vehicle handling from air traffic control, and making the relevant insurance-handling flow available at the freight-elevator terminal.

The insurance point is technical. It does not mean ground vehicles are being assigned a new consumer insurance category. The team wants the claim and retrieval logic to work where the player needs it, rather than sending someone to ASOP for one part of the process and back to a different terminal for the rest.

There is also a blunt reliability condition: freight elevators need to be dependable before another gameplay loop is layered onto them. The attractive promise is still a rover coming up in your hangar. The immediate job is removing the dependencies that would make it another route into existing failures.

A promising physics test does not fix every kind of desync

The recent rigid-body synchronisation work gets a positive assessment, with fallout from the first technical preview still needing fixes. Its initial scope is specific. Ships, vehicles and other entity types have not all moved onto the new approach, so a result for loose physical objects cannot be applied to every moving thing in a shard.

That is why another targeted preview can be more meaningful than a sweeping claim that desync is solved. A cargo object, a character and a ship do not all present the same networking problem. Each needs evidence that its own behaviour has improved.

A much longer written explanation is also being prepared. The developers describe the draft as roughly 45 pages and nearly finished during the recording, but the video does not establish a publication date. It would be premature to treat its release as part of a patch schedule.

Can players sell or delete ships bought in-game?

The engineering answer is that the backend can support such operations. Whether and how players should be able to sell or delete ships remains a design and creative decision. No usable sell-ship button or delivery date is announced.

The underlying change is an entitlement backend introduced in 4.8. It tracks in-game acquisitions as well as platform purchases and subscriber items; stolen ships are excluded from that ownership description. Insurance configuration is also tracked. Rather than rebuilding the ownership picture from the end state of the previous patch, these entitlements persist across patches.

That should not be stretched into a guarantee against wipes or bugs. It explains how ownership is represented. Wallet and ledger work provides adjacent transaction machinery, and a future sale could combine those systems. The developers also discuss deletion in terms of a possible zero-value transaction, with recoverability raised as a design possibility rather than a promised feature.

For anyone with a crowded ASOP list, the useful news is that the backend is becoming capable of more flexible ownership management. The frustrating part is that technical readiness alone does not answer when players will get the controls.

Why better server performance is only part of better AI

Server meshing has improved server frame rates, but AI decisions still take time to process. Streaming many entities into a large area can create stalls during initialisation; parallelising that work reaches deep into the engine. More performance helps, yet not every poor NPC decision is simply low server FPS in disguise.

Squadron 42’s behaviour work offers another route forward. Porting it to the Persistent Universe touches many systems, so it cannot be dropped in unchanged. The discussion covers perception, target selection, limits on how many NPCs may perform an action, and how combat roles are allocated.

More capable enemies could also reduce the need to fill a fight with so many bodies. NPCs that take cover, reposition and coordinate effectively can create pressure without relying solely on numbers. That is an appealing prospect for both combat and performance, but the Q&A supplies no PU delivery date.

Driving has three different foundations to reconcile

Ground vehicles currently span three kinds of rig: older Crytek-based implementations, a hybrid approach and fully simulated setups. The Mule and Nova are discussed as examples of the older and hybrid approaches respectively. That makes a universal handling adjustment less straightforward than changing one shared vehicle model.

The direction is to consolidate around the strongest approach, then improve details such as the drivetrain, braking, engine behaviour and suspension. Rotational limits, collision damping and responses to terrain also feature in the discussion. Better terrain collision representations from Planet Tech V5 could help even before every vehicle has moved to the same rig.

For all the technical detail, the desired outcome is wonderfully ordinary: a vehicle that behaves predictably when it meets rough ground. Across the episode, the best prospects have that same quality. A chat that keeps the group together, a terminal that retrieves what you asked for, and an enemy that responds convincingly would make the universe feel better long before another spectacular trailer.