Golgot's Gossip 2024

https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3959#post-10429351

- Apparently Rich Tyrer is very keen on the fantasy game and has been pushing for it for a long time.


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3948#post-10426310

- Our friend reckons a bunch of Turbulent devs are heading out the door now that hiring is picking up elsewhere in the industry generally. Outside of artists a lot of the original dev hires have either left the building, or are about to...


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3930#post-10418834

- SQ42 devs who supported SQ42 aspects in 3.23 then went back to SQ42.
- Every dev they know who plays SC tends to rage-quit. Normally due to being unable to complete missions etc.
- Some teams are struggling to find experienced devs to replace leavers. Sizeable gaps are filled with juniors.
- A 10 week push to create a new gameplay prototype is underway, to be presented at Citizencon. It is not going well and has involved churn.
- In their experience overtime was not tracked or reimbursed (in lieu etc). Most overtime occurred prior to presentations / Citcon.
- A new light system is coming. That means redoing most of the lights...
- Pyro is in a quantum state of: Is going to ship soon / can't ship because the content is poor or lacking. There is a push on to rectify this.
- Spawn Closets are actually 3 different entities. An area box of the proper size; which you link to a spawn closet, which then needs a spawn closet manager. They all have their own baffling settings. Devs just try all extremes to see differences and are never really sure if they see anything different. But then the most important settings are not in the editor, but in dataforge, the XML editor which controls all entities. There you you set tags to find the closets and then tags for AI that fit your needs. You also specify the numbers there. But it's also not very clear or functional. You have max spawns, max concurrent spawns, max spawns per minute and and a whole lot more, but it doesn't seem to work well. Because clearly nobody set it to infinite and yet that always happens...


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3860#post-10386002

Ok so our friend has been playing through an 'SQ42 demo' branch that appeared :)
Here are their thoughts:

- The current 'demo' build contains the whole game.
- It is playable, but barely. It doesn't crash like previous builds but is more bugged than the average alpha.
- It's heavy on cutscenes, especially in the first levels. Some of them are up to 15 minutes long with little gameplay before the next one.
- Internal gameplay showcases have been very empty. This partial playthrough aligned with that: Vast walking simulator areas etc.
- Players who enjoy flying in SC will probably enjoy it more in SQ42, with 'real' missions and linked objectives.
- The environments are much better in SQ42 than SC as they are crafted for the linear experience.
- You think there's a lot of known actors? Think again, there's even more than that. Almost every character was a familiar actor.
- The first level ends on an enormous opening credits scene. Including all the big name actors. And Sandi.
- There are a lot of missing voice lines and audio. Mocap work is apparently ongoing.
- Usually gameplay is done first, but just like the SC pipeline, the art here is mostly done while gameplay elements sometimes feel added afterwards. Many were still placeholder and whitebox with the surrounding art being complete.
- The story is pretty generic and the game itself is nothing innovative. It feels like playing a dated game with a fresh coat of paint. But some fans will love that since It's the 'same'.
- The AI in combat is a little bit better than SC and the social locations like the Idris feels more alive than SC, but it's still the same AI. They still stand on chairs, they bug out etc. It's a long way from other games like Cyberpunk for instance.
- They added collectibles to find in the game like trading cards, comics etc. It's standard in most games but it's missing from SC so it was nice to find that in there, rewarding exploration.
- The spreading fire tech is decent and works well with the extinguisher, but it was also annoying. In a level surrounded by water fire spread over the objective and main path and they couldn't proceed. Tractor beaming burning stuff into the water doesn't extinguish it. Hopefully it will all be fixed.
- You can thoroughly wash your hands in the bathroom. It's not gameplay, it's just a useless animation. It's the kind of stuff that gets cut in any other game, especially a delayed one like this.
- The game lacks tutorials and practise with key features. Like trying to use the inventory / collect ammo in the middle of your first firefight, while tackling enemies who can one-shot you.
- They stopped playing because they were annoyed by the game-breaking bugs, and also because they got bored. But they get bored in SC too in fairness.
- At an educated guess it's about a 30 hour campaign. That's being generous in case some of the levels they haven't played are particularly long.
- One level that was in whitebox 4 years ago is still not finished. Gameplay elements missing, buggy, with placeholder voice lines etc.
- Framerate is poor on a decent machine. A 4090 would likely still struggle with it. Worse than some pre-alpha console builds by other studios that they've played.
- The save system is in a poor state. Loading and checkpoints break or crash the game.
- The majority of the objectives they experienced were bugged, requiring a relaunch or skipping via debug menu. Scripted events failing etc.


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3845#post-10379059

- 'Too broken to test' is something that devs experience too. Even with tools to reach the correct location crashes and other bugs make reaching the exact spot / replicating the bug impossible at times.
- Certain directors push for unreliable tech to be scaled rapidly (100+ outposts). Sometimes they will join at the review stage and throw in last minute additions which rely on poorly performing features like AI.
- Additions can often break planned or existing features due to poor communication. For example PES persistence broke puzzles as key aspects wouldn't reset anymore but retained their state. Ship and player corpse persistence also impacted playability.
- Citcon can draw teams away from their usual work. One location team had to dedicate a lot of their attention to a fake Pyro map previously. Currently the Audio team has been working on Citcon content since Q1. (The Audio team is small and frequently has a backlog of work for other teams).
- Erin drinks 3-figure bottles of wine on the company dime. (Think near Concierge tier).
- Todd Papy was pushed, he didn't jump.
- Montreal is bleeding devs. They had plans for a UI team, but their UI Lead has left already.
- The programmer working on upcoming cargo additions left recently.


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3796#post-10348934

- There was a company-wide town hall recently. It was focused on the next 5 years. They discussed the coolness of their tech and IPs, excitement for release, and their intent to release more games/projects down the line.
- They said Sony & Microsoft are excited to see what comes out of this 5 year period. (The source reads this as: SQ42 port attempts have not gone well to date).
- Devs have been paid in ships to work at Citcon in the past, and QA have been paid in ships for overtime work at points.
- To go to the Citcon after-party at the studio last year they would have to pay like fans though.
- Instancing is planned for more than just the hangars. Whole levels / missions are expected to be instanced. This is still at the 'high level planning' stage.
- The unfindable 'missing person' bug is due to the way nested Object Containers load sequentially. If the mission entities spawn before all of the nested OCs have finished loading they just fall into the abyss.
- RASTAR requires every object to be an OC. IE a piece of cover has to be exported as an OC, then found in an unsearchable list of say 700+ OCs, then placed. OCs are heavy to load. RASTAR crashes a lot during this process and locations built with it have severe streaming issues.
- NPCs cannot avoid hazards on planets. They just walk into them. It's impossible to mark or exclude areas from the nav mesh. (There's no way they could avoid 'dynamic fire' as things stand, for example).
- It's not unusual for art to be the starting point for mission design, with desired features which don't yet exist getting added as art first.
- Tech designers sometimes work for months on Proof of Concepts, testing what works and what doesn't, only for that to get binned and the process to start again from an art base.
- Building interiors have been rebooted 4 or 5 times.
- Work can end up in the bin when someone leaves the company and a new voice decides 'it would be cooler if...'.
- The recent 15 settlements were green-lit on the back of quick art test in RASTAR.
- Previously CR asked for artists to make clothing interiors, just in case they are placed/folded in a way which might reveal it.


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3789#post-10345061

- The planet surfaces are shaders, not meshes. The terrain is recreated every server tick. Terrain modifiers are done in real time rather than using sculpted meshes which are manually baked in.
- This has some benefits for real time base building / altering the terrain. But there are downsides...
- It's far from optimized, causes entities to fall through the surface, and has issues with mismatches between appearance and collision.
- Regions can't be designated / locked off. Changes to global terrain settings lead to settlements / derelicts having the ground change beneath them etc.
- The AI nav mesh also can't be sub-divided, so AI can walk off across the entire planet...
- RASTAR's core objective was to allow previews of placements on terrain and automatic adaptation to future terrain changes. It never really worked on those fronts.
- Spawn closets were buggy on deployment, but RASTAR added extra issues. The settlement spawn closets are extra flakey due to this.
- A 'kill volume' to clean up NPCs under planetary surfaces has been requested many times.
- Art Director Ian Leyland emphasised that SC is a 'live service game' and that buggy deployments have to be accepted and can't be dwelled on. Devs just have to move on.
- Devs have been told 'server meshing is coming' for multiple years in a row, but the missions/levels deployed at the time were not designed to account for higher player counts. The attitude was: 'We'll cross that bridge when we come to it'.
- The inverse kinematics used for the unified avatar have various drawbacks in a PvP / MMO game. Realtime calculations can go awry, need to be networked, and tend to make gameplay less snappy. (There's also an argument that they don't provide gameplay metrics on the animation side, and so don't bring any standardization / structure to the gameplay design).


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3784#post-10342595

- The 'Star Engine' branding is primarily to attract new investors. It is not intended to be used by other studios.
- The engine itself is still essentially Cryengine. Folders, branches, server data, file extensions, the structure inside the engine etc. Even the parent folder is still called Cryengine ;)
- The first land sales are considered a mistake, due to the dimensions used.
- In the new system the first land claim will be cheap but then prices will rise steeply. They want to steer players towards building enormous bases.
- CIG did very well out of the first golden ticket event. They will continue them as long as they're making money.
- CIG have a small skunkworks team working on a fantasy MMO. It is intended to broaden CIG's portfolio after SQ42 & SC launch. CR sees CIG as a successful name in the gaming world and is planning accordingly.
- Tony Z is working on an unnamed project. Possibly the same one.
- Building Blocks is very awkward to use, comes with no documentation, and like many of CIG's in-house tools, lacks even an undo function.
- Fans who join CIG are most commonly found in QA. There are fans at various tiers of the organization though, and they tend towards being uncritical of project deliveries / state / progress. They are just happy to be working on these projects etc.
- It has been known for experienced devs to join, see the state of the project, and leave as soon as possible. (Waiting ~2yrs, or until they've secured a solid new role elsewhere, is more usual though).
- The UK branch seems to have a tendency to promote inexperienced devs to senior roles. Where they then tutor inexperienced devs in turn.


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3777#post-10339796

- The primary sources of revenue are intended to be ship sales + a subscription.
- Subscriptions will decrease all timers and time sinks. CIG say this is not P2W.
- Claim timers on big ships are intended to be very long. Hours.
- Idris size ships will be craftable. This will take orgs months to achieve.
- The Idris is not expected to have an AI crew by 1.0.
- CIG still talk about Server Meshing supporting 10,000 players at the same time. And it happening "soon".
- 4.0 is planned for after Citcon.
- It's now known internally as the Base Building patch, not the Pyro patch.
- RASTAR was supposed be for BB, but after 3 years work it has been cancelled.
- BB would require the new planet tech, server support, tools.
- The dev is only aware of BB art and design concepts being done to date.
- CIG keep adding things to 4.0. Including Nyx.
- The content density in Stanton is considered a mistake and unsustainable. The other solar systems will have one landing zone.


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/page-3775#post-10339625

- CIG recently announced internally that they are freezing raises, promotions and bonuses, and limiting new hires, for a year.
- The SC 1.0 plan is considered unachievable by everyone they know.
- Their first impression was: It looked like early planning for a whole new game, with 2 years to deliver.
- They confirm that the deadline is 2025.
- It is still just a fraction of what has been publicly announced for SC.
- The proposed SC 1.0 campaign is supposed to be 30 hours long, with 150 missions.
- 1.0 is supposed to be like an updated WoW with an SC flavour: A campaign, dungeons, raids, guilds, a player economy etc
- The ship list is still unclear. Not all concepts will be released, or to gold standard, that much is known.
- Communication between management and devs is increasingly firewalled.
- It has never been great though. Devs learn what other teams are doing from the internet more than internal communication.
- Roberts got particularly annoyed at some ship leaks.
- Devs often don't know when their code is going to ship. Sometimes they learn for the first time when it appears in the PTU.
- The engine requires a lot of maintenance. Dozens of issues a day, minimum, have led to a massive rolling backlog over recent years.
- New hires who have trouble fixing anything promptly leave again.


https://forums.frontier.co.uk/threads/star-citizen-discussion-thread-v12.548510/post-10337385

- They say that Chris had seemed bored and uninterested during dev presentations. He just approves or vetos decisions even if he knows nothing about them. If a solution is investigated and presented he may just go 'Nah I want this instead', regardless of any further issues his approach may bring.
- In more recent years he built a private theatre to review recordings instead of having live presentations from
- Some directors have pushed for playtests periodically. But mainly during this period a review was a CR viewing a 10 minute video and saying 'Ok, ship it' etc.
- The source feels that things have improved on this front, but probably in a 'too little too late' kind of way.

- They say Slipgate Ironworks were working on a Squadron 42 port.
- (Although they note that SI have been through some harsh layoffs and plans may have changed).