Welcome to Star Citizen Base

Register now to gain access to all of our features. Once registered and logged in, you will be able to contribute to this site by submitting your own content or replying to existing content. You'll be able to customize your profile, receive reputation points as a reward for submitting content, while also communicating with other members via your own private inbox, plus much more! This message will be removed once you have signed in.

FoxChard

Gamestar Article

Posted (edited)

IEDIT: the reddit translation is taking a little flak on RSI because it may not be complete-just excerpts. So it may have some errors, then again it might not. We'll know more soon.

Gamestar is back with another interview. They talk about 3.0, SQ42, AI, and Vertical Slice. There is a lot of good information in here and the authors don't shy away from talking about test builds crashing while they played them. 

I did not do the translation and crossover, these are from Redditor kruben95. yonasismad is responsible for the images.

https://www.reddit.com/r/starcitizen/comments/6nau6m/gamestar_titelstory_summary/

https://www.reddit.com/r/starcitizen/comments/6nc6j6/gamestar_the_technique_behind_star_citizen/

https://imgur.com/gallery/9iwKw

WALL OF TEXT BELOW for those who hate reddit.

Demo, with some in-depth look at systems (AI, Planets...)

Gamestar visited Foundry 42 Frankfurt to play a preview of Alpha 3.0. In the beginning text, the author asks the question, why F42 my have delayed Squadron 42. More about that later.

The Demo starts at GrimHex. The Player wakes up and uses the new mobi-glass to equip himself with some clothes.

They walk outside to the landing pads where a Gladius waits

They get introduced to the new use-System and item 2.0, where they can control many things more detailed (entering the ship, HUDS...)

With item 2.0 CIG can make every object in the world interactive by giving it properties.

With the new Star Map via the mobi-glass, they now can fly to POIs instead of the old marker on the hud.

Todd Pappy on the answer after the Strechtgoal Levski: "When C.R. makes a stretch goal, that means it has to be in the game in the given time. Often it depends on how important it is for C.R. Let's say, it's very important for him. The NPC Miles Eckhart, too."

Via the Star Map they jump to the edge of the atmosphere of Delamar and are then flying down manual. During the atmospheric entry, flames start appearing around the hull. This depends on your speed and the size of your ship.

A new flight control system shows the way to a landing pad on the highly detailed Levski landing zone, where are now also garages, where you can get your ground vehicles.

It is planned, that the ship will despawn after landing and some time has passed, to make space for other players.

Normally you would now go to a kiosk to let the ship repair and refuel because this would take some time.

In the Demo they have problems with the elevators and the physics grid, so they had to use console commands to go down.

The Details in Levski are really impressive.

As they approach Miles, they noticed a change in his behaviour. Miles notices us, checks us with a short look and looks down again on his mobi-glass. This seemed very natural.

This behaviour got they already to see when they visited Bob Reininger who is in Frankfurt for final implementations of the NPC's for 3.0. In General, all features from other studios come to Frankfurt to get implemented in the 3.0 build.

Gamestar got a deeper look at NPC behaviour and got shown the new features. In comparison to last years Gamescom demo, they made a huge jump in the details of movement. Everything seems very natural and the facial expression is incredible and in combination with the voice over nearly real. They are not done yet with NPC's, all the time are coming new features and bugs get eliminated. They want to make NPCs in 3.0 as perfect as they can.

The quest-givers and subsumption in 3.0 are test objects because, in the future, every NPC will use this technology in Star Citizen and Squadron 42.

Subsumption is more a system than a script.

Between 7 and 14 missions will be in the game. How many it will be, depend on how many problems they have to face until 3.0 But they will add them via updates.

 

The mission from Miles is finding a missing person. Last known place is Yela. A signal is coming from a possible crash site. The job is obvious: Fly to the position and find out what happened.

They go back to the landing pad and fly up out of the atmosphere, where they choose the new signal as a destination and start quantum travel.

Everything they do is seamless. There is nowhere a Loading screen.

Now they fly down to a caterpillar, a wreck.

this kind of wrecks will be partially be procedurally generated.

Some of them will be taken over by NPCs, but not in 3.0 yet. Some of them will fight you, some of them are friendly and want to trade.

They leave the Gladius and go on foot explore the caterpillars interior.

In one room they find the corpse of the missing person. (Sadly the interaction system in the demo fails) But now 2 enemies are coming from behind and start to fire at you.

To kill the enemies they use an anti-vehicle weapon. A lot of power, a lot of help.

Pascal Müller (Senior Environment Artist) and Michel Kooper (Lead Environment Artist) show them the Leveleditor (PlanEd) in detail. They confess that it sometimes might be a little boring to edit for such a long time a moon, that consists only out of rocks and dust. But Gamestar admits that they achieve with the tool some really interesting locations. Objects groups are automatically changed and look each time different.

They try to make each moon recognisable by their surface, geographic details and lightning.

Now the people from Gamestar are allowed to play 3.0 by themselves.

Shortly after they used the new nox bike, the game crashed.

After a restart, they try it again and everything works fine and they cruise around. The controls aren't perfect yet, at least for the gamepad.

Driving one round around a moon with a Dragonfly will cost you >3 hours.

The dimensions of a solar system becomes very clear when they looked at SolEd (The solar system Editor). Seeing the proportions to each other is stunning. That became clear when Sascha shows the Stanton system in SolEd and kept scrolling out and out and out. These dimensions alone are fascinating.

Even people who closely follow the development might not imagine how big this universe is.

In 3.0 will be a interesting cross-play-system. 2 Players will get different quests and sometimes the same job, maybe even with different goals and this might bring 2 players to the same location where both of them don't know what the other player want. It might be very interesting to see how the people react.

But this might also be a problem because the reputation system is coming with 3.1

With a console command, they teleport to an outpost, which might be part of a mission in 3.0

Currently everything in them is not useable, but this will change, because item 2.0 makes it possible (don't know if already in 3.0).

Not everything that will be in 3.0 could they test. (trading, insurance and stamina).

Many apps and improvements on systems are missing in the demo.

Question for Todd Parry: Can you fill this huge space with interesting stuff? Todd: "I don't know. But it is my goal. I am as long pessimistic until I see results".

He is very optimistic with planets and outposts, but the next challenge is to make it a living environment.

They work piece by piece and systems are coming together which will help them to make the next step. Most of the time it takes longer than C.R. and the developers want, but they need this time to have a foundation to build on.

Many people think, that C.R. should have forced some parts a bit more and don't be the perfectionist. But with 3.0 the editor thinks, that they may not have done something huge (content wise), but something that will be the base for a game that can be played for years.

Where is Squadron 42?

The displeasure after the cancelled demo for SQ42 was huge, but everyone who knows CIG for a bit longer knows why it wasn't shown. At first, there was the new planet tech. SQ42 is C.R.'s baby. And everything has to be perfect. Gamestar speculates, that C.R. want to have the planet tech also in SQ42 (!speculation!). An additional reason was the work on the AI. Until last year, they worked on Kythera and CryAI, but C.R. didn't want two systems. Brian Chambers: the foundation has to be perfect for such an important part of the game. They had to make half a step backwards and build it right. Both systems got merged together into the subsumption AI. Todd about SQ42: "It's going". Brian: "I would like to show and say so much more". He sees every week all these cinematics and epic space-fights and then he always thinks: "what are procedural planets?! Look at this!". But they can't they anything about that. All the animations and their quality are the same in SQ42 and SC. Brian: "It needs more time, but they think it was the right decision. We make good progress, but the core elements they develop for 3.0 are also needed for SQ42. So the work at the NPCs are necessary for SQ42 to achieve such high-level animations, facial expressions and behaviour. They really want a high quality for the characters. Everything must be better or at the same quality as everything in other comparable games. And I am not happy until we reached this goal. For Star Citizen, I show features when they are not ready and polished, but for SQ42, I don't want to do this. If I show it, it has to be like Activision or Bethesda is unveiling a new game at E3, only with the difference that everything we show is ingame." According to C.R., no work got wasted from last years Vertical Slice of SQ42. The team decided to not show any work in progress last year. He wants to show the finished system without any compromise solution and workarounds. With the current status is C.R. very happy:"When the people see it, they will understand and they will be very happy about it". It's done when it's done. When we can see SQ42 didn't want C.R. us to tell.

Off-Text

In front of the studio are loud noises to hear. The city is building a Subway in front of the building, which has to be on the surface because the urban planners missed some how that they can't go under a bridge nearby. With organisation problems is Brian Chambers (Development Director) familiar. Brian worked on Effects for Star Trek Voyager, was animation director for Red Dead Redemption and some GTA's are influenced by him. Then he worked at the WWE brand and then joined Crytek as a Senior Producer at Ryse: Son of Rome. It is relatively quiet although it is a plan office. The atmosphere is relaxed and concentrated. In front of 2 or 3 monitors, the developers are writing code, thinking about problems or making skype calls. They get every day many applications, but they are very careful in hiring people because they have to fit. The mentality is very important, you can't just have good knowledge. Star Citizen wouldn't be possible without this outstanding team.

An often asked question to C.R.: "Do you believe, that you can do it?" On the one hand because of the sheer size and on the other hand because of the technological challenge.

Chris and the team are confident, that they can achieve this.

The passion and the know-how is really impressive

Every goal is with current technology achievable, says C.R.

But not every goal or idea will be implemented in the way it is planned because they realise, that it isn't fun or a good way to do it this way. But they have no problem to go back and start new to make a better implementation.

A huge amount of time was invested in building up this team and the foundation (making changes to the engine etc.)

Marco Corbetta and Carsten Wenzel (ex-Crytek members, now for 2 years at CIG) make clear, that there is no CryEngine anymore. If you compare the engines 1:1 you have around 50% of CryEngine code and 50% of new engine code. But if you look at the actually used code by SC, you have around 10% CryEngine code and 90% new engine code.

2 things resulted in a huge refactor: The sheer size of the universe and the interaction possibilities.

To make a hole solar system with huge planets possible, you need to make it possible, that the engine can handle these coordinates. Such precision in such a huge world is only possible with 64Bit precision.

This change took them about a year.

The huge amount of interaction possibilities with NPCs and objects required a re-write of the Job-Managements-System, so that the code is not blocking the main thread. The logic now makes full use of all cores, because they need many refreshes for the many entities they will have in the game.

A weakness of the CryEngine was, that not everything can be an entity. They changed that and now every entity can have different components. A ship has the flight-control-system as a component. Then it has a life-support component in form of different rooms with oxygen. There was no component-system in the CryEngine.

This is called the item 2.0 system. This lets the team develop and test each component independently.

According to Chris, this is a very modern principle in programming that is expandable-friendly, performance-friendly and makes it easier to synchronise the game world over the network.

The first plan was to make SC like in other MMO's with capsuled instances with not many players.

Currently there are 24 players on one server, which will probably not change in 3.0)

The rework of the engine also affects the network. There will be many servers that can communicate with each other that will enable instances with thousands of players.

A physical server contains currently 32 cores that make 8 virtual machines possible for dealing with data. This is because currently, the CryEngine can't do much with more than 4 cores.

And another problem is, that more cores don't mean more players in a linear equation. This will change in the future.

When they get 24 players with 4 cores and the scaling with CPU relatively linear with the player count is, they can get far more than 100 players on one server.

And when they have the server network and the seamless transition between them, every player practically plays in the same instance of the world.

A huge step towards this is the Batch-updating. With this technology, one batch together from all core processed. This makes Synchronising much easier. Especially physic is with this method much better scalable.

Because this technology still needs some time, it is important to continue to rework the CryEngine to allow for much more, like a huge amount of NPCs.

The plan is, to have 90% of the population to be NPCs.

All these demands from the game (huge universe, planets, ships physics, physics grids, effects, etc.) are not easy to satisfy.

The Demo they played run on an i7 5930K, Nvidia GTX 980 and 32GB RAM with about 30 Frames.

They have still problems with elevators, the ragdoll-system and server crashed.

What Hardware will we need: Corbetta and Wenzel: It should run on modern hardware but it will scale very good with much better hardware. They really want to use all the power of your machine.

To be aware of performance, they don't want to have the best hardware in their PCs and they have a huge variety of components in the developer PCs to notice problems fast.

Optimisations for the new Ryzen processors are scheduled to be worked on.

They don't want have the reputation like Crysis (running gag: "Can it run Crysis?").

The minimum specs for the PC are these: 4 core CPU, 2GB GPU, 8GB RAM. Recommended is an SSD.

Graphic settings will give you many controls over the performance demand.

Currently there is only one quality setting for the overall quality

There probably won't be more than 20 options for the graphic settings because it would need too much testing.

But there will be enough options to control the performance.

Will there be a downgrade? No, because they don't have to develop for consoles. They can develop for future hardware and people can turn back settings by themselves. And current High-end hardware will be in 2 years only upper middle-class hardware and SC will hopefully be played for many years.

It is also important, that this is a crowdfunding game and they have every few month a new public and working build and get feedback.

C.R. is very happy, that so many people understand that they take the time to do it the right way.

Internal tests are done with up to 70 people. After that come 2 waves of community testing before the release. (Evocati with around 1000 player and the public test Universe (PTU) with up to 20.000 players.

Only a version that got through all tests comes online for everyone.

The multiplayer part is the biggest challenge for them according to C.R. It is one thing to test everything alone in the editor and testing it over the internet from all over the world with people in different locations in the game world.

It is a challenge to keep track of so many things in the game world and being smart by knowing what to keep track of or to sync over the internet with the clients.

Huge problem was to synchronize physic over the internet because they have different gravitations and gravitation-directions.

They are not done with many things for 3.0, so more delays can be expected.

They have huge ambitions for everything although it is only in an alpha state.

Question after technical highlights: planets and sheer size of the game, but the answer to this question will always change as they try to add new things and improve the game.

Wenzel says: Don't quote me on that, but we try to achieve Crysis-Niveau for our planets and moons.

Duel of the interfaces: DirectX vs. Vulkan

The game makes use of a Low-Level Interface like Direct X 11. Because Direct X 12 is Windows 10 only, they will switch to Vulkan for newer API features. This will be done in time for the commercial release. They maybe even plan to completely drop even Direct X 11 support.

Vulkan will make a way better core utilisation possible, which leads to more drawcalls.

CryEngine already plans in the near feature to support Vulkan 5.4, but CIG can't take these updates because the code is too different. But they have a similar plan like Crytek, but they first have to complete the underlying re-write of many components to work on this.

The current problems are more in the game code than on the graphics API.

Big, bigger, 64 Bit

The biggest challenge was the 64 Bit implementation.

This is something that hasn't the CryEngine neither the Lumberyard engine.

64 Bit makes it possible to work with much bigger numbers (264 = 18.446.744.073.709.551.616 Bit instead of 232 = 4.294.967.296 Bit)

With 32 Bit, it wouldn't be possible to make such huge solar systems.

For many years CPUs support 64 Bit and even Windows XP got an update for 64 Bit, but today's games are mostly running on 32 Bit.

For comparison: The game world of The Witcher 3 can be placed in one crater of a moon.

Everything procedural, or what?

Someday, Star Citizen should contain around 100 star systems with over 300 planets and moons.

Every planet and moon should you be able to walk on.

For release C.R. is aiming for 5 to 10 star systems.

To accomplish that, they made the procedural planet tech.

This is just a tool for the developers to get a foundation to work on

The planets don't get procedurally generated as you fly by.

This works by overdraw the planet with many biomes and layers. Like with a specific type of forest, stone or in the future also with water (maybe lakes and rivers? We already saw ocean).

C.R. calls it "painting with a bigger brush".

But every planet and moon will be revisited from developers to make POIs (Point of interests) and landing zones.

Where with all the data

when travelling through the game world, your PC has to have to correct data ready to get displayed.

Having a whole star system with real planets is a technical challenge. They have to stream the data very efficient.

To solve this problem, CIG developed the "Object Container Streaming".

An object container contains all data necessary for displaying this container. This could be a whole planet, a space station or just a room in a space station.

SC will now identify which object may be needed and loads this data in time. A hierarchy ensures, that the most important data is at first available.

More cores will support a better streaming because the main thread will be relieved.

To support the technique hardware wise, it is recommended to have an SSD and a huge amount of RAM.

Currently are 4-5 special areas on planets and moons.

As soon as Object Container Streamins is implemented it will be much more.

According to C.R., the implementation will be done until the end of 2017.

 

 

Edited by FoxChard
Disclaimer
Donut, VoA and Riley Egret like this

Share this post


Link to post
Share on other sites

Makes sense why Chris wouldn't want them to show off Squadron 42, there is no closing Pandora's Box. Once it is shown in whatever state it's in will be what it is remembered as for years to come.

You don't want to have spent years of your life in blood, sweat, tears, headaches and sleepless nights only to have your magnum opus turn into gif fodder for Reddit, and the impact of the reveal to go about as well as the Shockmaster.

Spoiler

 

 

Share this post


Link to post
Share on other sites

I think they may have taken a lot of crap from people, but the big public reveal of SQ42 is what will make or break it. If Gamestar is right and they were going to need to fake or fudge it to get it working properly, they were right to postpone.

I hope that we'll get it soon-maybe Citcon since that event historically attempts to cover SQ42?

Share this post


Link to post
Share on other sites
37 minutes ago, FoxChard said:

*snip*

Or Gamescom, where there will be far more than the Choir being preached to.

Share this post


Link to post
Share on other sites

Was about to bring up this article myself just now after happening upon it, really interesting read.

Fintz likes this

Share this post


Link to post
Share on other sites

I'll edit this into the OP, the reddit translation is taking a little flak on RSI because it may not be complete-just excerpts. So it may have some errors, then again it might not. We'll know more soon.

Share this post


Link to post
Share on other sites
10 hours ago, J. Coren said:

Makes sense why Chris wouldn't want them to show off Squadron 42, there is no closing Pandora's Box. Once it is shown in whatever state it's in will be what it is remembered as for years to come.

You don't want to have spent years of your life in blood, sweat, tears, headaches and sleepless nights only to have your magnum opus turn into gif fodder for Reddit, and the impact of the reveal to go about as well as the Shockmaster.

Did someone say gif fodder??

 

Spoiler

giphy.gif

 

Agreed though. Don't want to see Squadron 42 become the butt of the internet's jokes for a couple weeks. 

FoxChard and Riley Egret like this

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now

  • Similar Content

    • By Donut
      Source: http://www.gamestar.de/spiele/star-citizen/artikel/star_citizen,48820,3307804.html
      (Translated)
      THE WAITING GAME
      Four years ago, Star Citizen occurred with a Kickstarter campaign to conquer. $ 141 million later, we can look together with project chief Chris Roberts back on turbulent times and fathom why patience is still a virtue.
      To climb a high mountain, even using the most modern technology is not a walk. Fitness, good planning, an iron will and a lot of patience are essential for the climber. The development of the mega project Star Citizen has become a similar challenge: After the project was initially a not necessarily small but manageable survey, it has now grown to eight thousand, both in terms of volume as well as the technical challenges.
      Bookmakers Chris Roberts might but in 2017 in front of the summit: If all goes to plan, episode one of the single-player campaign is Squadron 42 delivered and the update 3.0 is the first to see a complete game from the multiplayer universe Star Citizen. We spoke with Chris Roberts in an interview at length about the challenges of development, have coaxed him details of technical solutions, drawn information on emissions and Housing from the nose and of course asked about the state of affairs concerning Squadron 42nd
      We can look back with him to four years of development and look a bit into the future. We explain why Star Citizen is a real puzzle and why we believe that the wait could really pay off in the end.
      A dream takes off
      After a long break from the game development and a thoroughly successful foray into film production ( "Lord of War," "Lucky Number Slevin"), the Wing-Commander-father Chris Roberts anno 2011 decides to return to his roots.
      He wants a game after Minecraft model develop: produce an alpha version, sell them and use the proceeds for gradual improvements. As engine selects the CryEngine 3, the prototype for its new space game he can develop from freelancers and friendly studios. The cost it pays out of pocket.
      Originally Roberts wanted to win with the prototype the usual investors for the project. With the advent of Kickstarter but his enthusiasm begins for crowdfunding, ie the financing through many small contributions from private supporters. He first tried it on a website that breaks down promptly after the announcement of Star Citizen in October 2012 under the onslaught.
      Shortly thereafter a Kickstarter campaign built from the ground: After 30 days, Roberts has over two million dollars taken by the Kickstarter source and again four million on its website. If the match can be developed without investors perhaps? About any additional objectives (so-called stretch goals) comes in more money, at USD 22 million announced Roberts complete independence from any investors.
      From 65 million will be no further Stretch Goals more awarded, the feature list is long enough. At present, and after about four years of development more than 140 million dollars have been collected. But that does not mean that the project Star Citizen always went like clockwork.
      Austin, we have problems
      Was initially a manageable project with classic space dogfights, and a single-player campaign (Squadron 42) planned, the steady stream of money will soon generate a rapidly growing extent. Chris Roberts: "When we took more and more money, we said: Hey, we now have the ability to do it the way we really want to do it. The challenge was to get everything together to create a reasonable workflow. "That should be more difficult than thought. Cloud Imperium Games must establish from scratch a complete studio structure. Next to the studio in Austin come 2013 Locations Santa Monica and Manchester (United Kingdom) to do so. In addition, Roberts relies on contract studios as Behaviour Interactive ( WET , 2009), IllFonic (see box) and Moon Collider (Kythera-KI).
      The multi-pronged development of single-player campaign, multiplayer universe and the live operation of the playable modules (hangar and Arena Commander) requires far more specialists than are present. We are looking for highly experienced software developers who are familiar with the CryEngine and write tools for designers. But which are then few and far between, which causes delays in operation. Often the required tools are simply not available in time. Only with the decline of Crytek UK relaxes the situation on the personnel front: After Crytek in April 2014 can no longer pay salaries, engages Cloud Imperium Games there from a number CryEngine specialists. End of 2014 CIG already employs around 180 staff. However, pushing once other structural problems in the foreground.
      IllFonic worked since 2013 with the development of Star Marine, the first-person shooter module for Star Citizen. As their work with levels that were built directly in CIG, should be merged, a catastrophe occurs: IllFonics assets have the wrong scale and do not fit into the CIG-Level!
      "Although it looked as if it were almost ready, but did not work the last 20 percent at the end, and we had to unravel it all over again and start from the beginning," explained Roberts. The throws back the entire development. CIG draws conclusions and begins to unite most of the elements of the development under their own roof. This includes the shooter module and the AI that at Moon Collider was in work to date and is now further developed in the new Frankfurt studio.
      Additionally begun better to delegate powers and responsibilities. Foundry 42, the CIG Centre in Manchester, is developed in the Squadron 42, serves as a role model. Chris Roberts' brother Erin and some of his colleagues had previously worked for years at the lego game and knew how efficient studio structure works. Their knowledge is gradually applied to all studios of Cloud Imperium Games.
      At the same time the shortage of skilled labor decreases slowly: "We've got some really great people, for example, the Frankfurt studio is obviously very good for us have been. There we had a lot of people who were familiar with the engine and have contributed much to the planet technology and other things, "explained Roberts. "We now have a really strong team, which is at least as good as any team in the games industry."
      Extensive Engine Changes
      Having a good team is one thing, the appropriate technical basis the other. Roberts was and is scolded by media and critics repeatedly for his choice of CryEngine: She was not meant for multiplayer player of this magnitude, so the frequently voiced criticism.
      Basically, that's not wrong. The originally planned Star Citizen version had a much smaller scale and significantly fewer features. However, with the financial encouragement by the fans grew the possibilities many times over - and thus the demands on the engine. This makes extensive revisions to the CryEngine necessary.
      One of the biggest restructuring on the CryEngine is the conversion to 64-bit double-precision, culminating with the release of Update 2.0 end, 2015. Until then, the CryEngine runs with 32-bit precision, which only a few square kilometers allows big maps.
      "Most engines work with 32-bit," explains Roberts 2015 compared to the British magazine PC Games Network. "This works well for a first-person shooter or an Overlap shooter where you have only a few square kilometers of areas. But we are in space, we are thousands, millions kilometers. "
      This precise travel within such gigantic maps is possible, the engine must be adjusted to 64-bit. In addition to this construction site and the network code is newly reissued (the work it continues to this day). Around 50 percent of the engine had been previously adapted to individual needs, gave the Frankfurt studio boss Brian Chambers in an interview at the Gamescom 2016 Protocol.
      Although this work required a lot of time and effort, but results are already visible today. Already in the persistent world of current Star Citizen-Alpha (around the planet Crusader), players can explore an impressive 400 quadrillion cubic kilometers Space (official figure). Of course, the majority of "only" empty space, but the technology behind it seems to work fine - apart from some serious server lags.
      With the complete Stanton-star systems in the Alpha 3.0 the card size should even grow. But all these basic work costs much more time than originally planned. And that is reflected especially in the public perception down - no player like delays.
      Gaming expectations
      Despite a largely open development, which is accompanied by a detailed monthly reports from the studios and weekly video formats, not tearing the partial unobjective from criticism. Non-compliance with deadlines and the development time can be found again and again in the crossfire.
      In the original Kickstarter campaign it was then: "After twelve months (which would have been starting from campaign statements the end of 2013) we will allow the early supporters to play the multiplayer Space-Combat-Alpha and other 20 to 22 months (ie the end of 2015) they are the Star Citizen Beta play [...] "And do not forget. Squadron 42 should also be delivered already the end of 2014 to the supporters. The Arena Commander, so the multiplayer Space Combat module appears, in June 2014, six months after the original target date. Already at this point it is clear that the originally mentioned dates can be reached in no way realistic, because the millions of dollars raining for some time in a weekly cycle on CIGS accounts and allow much more features than originally planned. Roberts is considering shortly after the release of the Arena commander to refrain from further Stretch Goals and provides the public with reaching the 46-million-dollar mark for grabs.
      Some 35,000 supporters from voting, 55 percent are for more Stretch Goals, 26 percent opposed and 20 percent other it does not matter. The desire of supporters there are correspondingly more so, in some cases very complex objectives as detailed AI activities and improved modularity for spaceships. Only when the 65-million-dollar mark end of 2014 draws Roberts a definitive line under the Stretch Goals. Had Roberts against the supporters might have to make clear that will significantly extend the waiting time for a finished Star Citizen through more content? "If I go back and would not change a thing, then, that I would say much more clearly: The more Stretch Goals and features are in it, the more complicated it is, the longer it will take," Roberts shows insightful.
      "Looking back, I would have time to much more energetic point out," The boss can develop it but even not go fast enough. "I'm a bit like our Supported and a little impatient," he says. "I wish we had a few things much further.
      "It might like to go a little faster, but we have a great team, and when I look around, I see people who often work longer because they are with heart and soul into it. So if it takes longer, it is not because that is not working hard, but in the development process of a project with this scope and complexity. "
      Dates called Roberts Although no longer as free from the liver away like a year ago. But now and then he is still (much more carefully formulated) data in views that do not work in the end and the impatience of some supporters fueling yet - as the review of the 2016 shows.
      Price of Progress
      The many small and large restructuring of 2014 and 2015 have an effect. The end of 2015 published CIG the first big update for Alpha. With version 2.0 Crusader comes into play, a huge map with various stations, the first missions and basic shooter mechanics that work even in the new EVA mode (Extra-Vehicular Activity, Activities in zero gravity). The Multi Crew feature shown only in August is also attended and players can at service stations carry out repairs and replenish ammunition.
      Update 2.0 is at that time the largest and most important date update the evolution of Star Citizen. It lifts the previously available only in single modules existing game to the level of a true alpha version with many basic features that come together in a small (not persistent) part of the Universe.
      The persistence, so the server-side storage (purchased with the new Alpha-currency) objects and marine and player states will be integrated in June 2016, version 2.4, which represents a further technological milestone. Outwardly this is not a very headline-grabbing thing for the development itself but extremely important: the back-end functionality is complete, the universe starts for players finally continuously to exist and no longer begins with each new login from the beginning.
      A big PR coup succeeds Roberts with the presentation of the procedural planet at Gamescom. In it he shows the approach to a planet, landing both on the surface and in a new landing zone and, based on an impressive, complete story mission. There are gun battles in zero gravity, vehicle hunts over the surface of the moon, and briefly is the interactivity of objects to see (a cargo box).
      Planets and their exploration were originally intended only for the period after release. But the Frankfurt studio has made extreme progress in the technology - so far that it on the CitizenCon are few weeks later another impressive presentation of procedural planet, including weather effects and a giant sandworm. All these things make 2016 more than 36 million dollars in funds for supporters financially most successful year for CIG.
      No Squadron 42
      Victims of this positive development is Squadron 42. The entire 2016 passes without there to see something new on the single-player campaign. On the CitizenCon an almost one-hour demo should be shown - shortly before the event but will be deleted . The reason is CIG to problems with the new AI and animations.
      "We want the crew pursues normal duties on a vessel and you can interact with them," Roberts tells us. "That's the AI page. But now we need to ensure that the behavior is associated with smooth animations, for example, if someone goes to a table, sits down, eats, gets up and goes away.
      There should be no change choppy, but a liquid movement pattern. But that will take longer than planned, and is one of the reasons why we have the demo not shown on CitizenCon. We're trying to achieve just the right level of detail, and that is definitely a big challenge. "
      Roberts suggests after CitizenCon that the demo would eventually refilled later. But even the latest live stream in 2016 goes by without news about Squadron 42. The impatience of many fans makes many, partly unobjective articles on Internet air. What is Roberts to when it massively hails criticism?
      In this project, things go very fast, even if it does not appear outwardly as if it would go ahead quickly. One constantly has the feeling: We need to finish getting this thing, we need that raushauen, people waiting on it. The community is awesome, but you already feel that they have a huge appetite for everything they can get. And if times a while nothing comes, then they are a bit grumpy. "
      Roberts adds:" People say, 'I want to have it now, I do not care if it is not working properly' And if you do them then. would show or give, they say: 'Hey, that works not at all, which does not look good "But apart from that it annoys me sometimes, I think that we have a very passionate, caring community that. provides us with valuable feedback. "
      Details need time
      Besides AI, the desired level of detail is another reason for shifts, even if the team is making good progress, as Roberts states. "Our goal is that you have while walking around on the Idris or in interactions with the crew, the quality of a cut scene. And there are, for example, problems with the lighting. We want to achieve a cinematic lighting and therefore we must highlight and shadow - and there are quite alone on the Idris thousands - adjust to achieve the right effect
      Another point is Object Container streaming, "Roberts says. »Squadron 42 takes place in a complete, open the solar system, in which you can travel freely between the planets. But you can not have all the data at once in memory, but you need so-called containers containing certain areas. "
      The streaming is also run always in the background, so that the player does not notice it, if a new field (or a new object container) is loaded into memory. "However, we need this technology not only for Squadron 42, but also for Update 3.0."
      Ever seems Update 3.0 and the associated features to have had a significant impact on the displacement of the single-player campaign at 2017. While the story of Squadron 42 with more than 1,250 pages of dialogue text already completed and the motion capture of high-profile actresses cast (including Gary Oldman, Mark Hamill, Gillian Anderson) are turned off, it is not merely the fine work that can last for anything longer.
      Technical advances such as the procedural planets are in fact also play a role. If you consider that the first major demonstration of the planetary art takes place only in August 2016, one can imagine that the implementation is in the single player campaign is not too long in labor.
      And then there's Item 2.0, a system that Roberts explained in our interview in connection with Update 3.0 (see box). This system will 42 raise the interactivity in Star Citizen and Squadron in to a whole new level.
      Quo vadis, Star Citizen?
      With the update 3.0 is to perhaps the greatest milestone in the history of development of the project. This Star Citizen would in fact be a full-fledged game, have implemented all the basics and provide enough content so that players can employ in the universe long first time (see box on the planned content of 3.0).
      On the CitizenCon 2016 Roberts makes this update again one of his now infamous date statements - even if vague: At that time there is, CIG would try 3.0 still bring out the end of 2016th Ultimately, they provide at this time (disrespectful words) "only" the release of Update 2.6 with Star Marine (see box).
      On the question of the status of Update 3.0 grins Roberts and raises both hands defensively, "I will no timetable or an assessment for an appointment rausgeben, but there is still much to do. For 3.0-Star Citizen is something like a complete game with all the important corners. "
      Then he goes into detail:" The main ingredients are all in work, but there are still a lot of minor things that need to be made, for example, air traffic controls over landing zones. There are only a certain number of landing zones and it can not land a thousand people at once. Therefore, to a meaningful system to be written, like a real airport. . Such things are not necessarily difficult, but a programmer needs for maybe three or four weeks, "
      Even things like boarding and security talks on Roberts:" At the moment, each a door open to a spaceship. With Item 2.0, you can close the doors of your spaceship. Then, when someone wants in, he must chop or break the door. "
      "So there is still this or that detail, and a multitude of other little things that all must be brought together," Roberts concludes. That does not sound like a release in the near future.
      "We've looked at 3.0 and said. We need that and that and that and then we found: Damn, that's more than has so many complete game. Therefore, we develop a detailed plan for all tasks and subtasks. If that is done, we will share this plan with the community. This is expected to be the case at some point in January, depending on when the production team the information gets from the project managers. "
      Thus, the time until then completely goes by without new content, there should be between updates, for example, improve the performance. Among other things, it is planned to increase the number of players who are adapting to a server in Crusader. Most of the work on performance and net code is published only with 3.0.
      The biggest challenge
      Because so goes according Roberts also perhaps the greatest challenge in the whole process along: "Probably the network setup and the network code are the biggest challenge, because the CryEngine is not really designed for a multiplayer game.
      In addition, it is very difficult to find good network programmers in the games area. Meanwhile, we have a good team, but for a long time we had a few people who have worked on it. And then added that we make a game that has a level of detail and accuracy such as Crysis, but as a multiplayer game and a much larger scale. "
      The importance that the CIG attaches a stable and powerful network that can be good at surprising Engine conversion to Lumberyard (see box) can be read, which has the connection to the global server system Amazons integrated directly.
      Roberts & Co. It is not enough to use traditional technical ways and improve. During the optimization of the network codes rather part of normal daily life in the development and maintenance of multiplayer games, CIG is constantly looking for ways to further develop the technology.
      The physical grid in Grid technology, the multi-crew mechanics makes it all possible (whereby, for example, a player in a spaceship stands quietly on the spot, while the ship itself in space flying wild maneuvers), is a good example.
      Item 2.0 is another example of how Roberts explains in detail: "Among other things we are working on a kind of entities Planner and -Updater. Actually Item 2.0 is more an Entity 2.0. Entity is in game development is a collective term for any object in the game, it was a spaceship, a player or a weapon. In the new implementation, which is introduced with Item 2.0, these entities have their own components. You take just one entity and packst various components in, for example, a physics or graphics or radar component. "
      The entity spacecraft can thus for example, a physics component are attached, allowing gravity inside the ship. "So we have rewritten the engine based on the components, which you take individual functions're stuck on an entity and thus determine what this entity can. And that is updated quite different: Some components are updated every few minutes, others second.
      Thus, the outputting of information is much more efficient. In the old version, each entity has been updated in each frame, which is totally inefficient. And therefore, we have revised the basic systems, which now coincides more with modern engine development. For these changes, we focus on 3.0. Some improvements can be found being observed at 2.6, but the majority is planned for 3.0. "
      Lots of space, lots of content?
      In addition to improving performance, this system allows especially even more opportunities for developers to fill the gigantic worlds that are to open up in the Star Citizen universe. Even the Homestead demo of the CitizenCon impressed us with a huge planet, with almost unlimited amount of space. Each audience shot involuntarily the question through my head: How can this massive room, these many planned giant planets are filled with meaningful content?
      The creation of a complete planet to the designers, if all tools are completely finished, cost no more than a week's work. "The goal is to have templates for specific ecosystems, such as mountain ranges or deserts. From this range of templates, the artist can then a planetary environment "painting", for example, as Tatooine or Hoth.
      Based on this, we work alongside the major landing areas like Area 18 ArcCorp of modular sets of outposts, which can be composed differently from the artists depending on the environment, such as a settlement, there a few farms. Based on these sets the area is then automatically populated, unless the artist overrides the manual. "
      Part of the content and quests is generated from the respective ecosystem. The emissions system also includes procedural influences, for example, certain resources and, based on a specific freight line. "Then pirates may appear that in turn make escort for cargo required and so on. There will be a kind of complete set of rules between AI and players, making it permanently are ways to make money and to do some stuff. "
      In addition, there should be on all planets and some stations special missions that are offered depending on the player's reputation and availability of Quest. Such orders are made composite by designers blocks and should be clearly distinguishable from the things that make the player normally.
      "The idea is that you run around and all that are doing what you normally do, for example, be. And if things go well, certain issues are eventually available, something like Super missions. The do not you ever do or more but succession thereof. There are special missions, specific features, in addition to the normal activities with other players or the AI. "
      Home, Sweet Home
      Presented from the order to constantly have motivational content before and become long-term commitment to the game? Roberts enough that - surprise! - not. And that is why Star Citizen will sooner or later offer a complete sandbox, including housing. Goods initially maximum apartments planned in cities or in stations, the new technology around Item 2.0 and the entities system makes a lot more possible.
      . Chris Roberts: "There will be the opportunity for players to build their own homes or outposts" How is that possible, it leads immediately afterwards technically made "freight - ie crates or boxes, which are made for example in the cargo hold of a Freelancer - is stored in a persistent database.
      »The same technique is used when a player discards important items at a location on a planet. You can go away and come back later and the items will resurface because they are stored in the online database. For us there is no difference between a rifle, a box, a room or home - these are all items in the same item system ".
      Item 2.0 is to allow not only a more efficient flow of information on the technical side and higher interactivity on the gameplay side ie, the system thinks much larger: "One of the plans is to allow players with their ships to fly somewhere and build a home , For example, to portray a small power plant, and then perhaps to protect a radar jammer, so it is not detected.
      "Then, the power plant is connected to a turret, so it creates its own small base. When Tony [Zurovec, responsible for the persistent universe in Star Citizen] talked about farming it was, in principle, exactly that, somewhere to have an outpost and there to plant things and to harvest. "
      Of course there will be limitations, who does what where and how much must build. "Finally, not every player his own Megacity pull" quips Roberts. "But I can imagine organizations somewhere build a small base, perhaps near some resources that break them down or sell me. And then listen to another organization of and attacks them with space ships and land vehicles. "That sounds a bit like the EVE-online dynamic that always brings forth by dominated by players systems and stations major conflicts, involved in some thousands of players are. In this way sandbox contents to be inserted, which do not require emissions but just happen. "Once all the parts are developed and introduced for the players will be able to create their own content. That's one of the rules in the development of Star Citizen that the systems are flexible enough to allow such things.
      Of course, this is also one of the reasons why it takes longer, since such systems must be built in a certain way. But ultimately I think about the game and the game is better in the long run. Because we give players a sandbox and say: Hey, you always wanted in a science fiction universe to live? Here it is!"
      A big cauldron boils slowly
      With this we are at the core of this patience game that Star Citizen called: It is not the game that 2012 was touted in a Kickstarter campaign. Had it remained with the few million dollars from October 2012, then Star Citizen would probably already finished. However, we would then get only the things that would have been possible with the traditional technique.
      About 1.7 million supporters have the financial framework, now with $ 140 million but such reamed that Roberts "ballpark" Star Citizen simply no longer comes into question. Meanwhile, from a technical summit become, the less intended, after all nothing more than to lie absolutely the best space game ever. Even if Roberts does not explicitly say, you can tell him with every word, with every gesture. There's someone here with enormous passion. Someone who only the best is good enough.
      One may accuse Roberts megalomania, however, speak his previous technical success for him. For more and more playing on safety games industry that rarely even take a risk or something truly groundbreaking new venture, the project is certainly much needed breath of fresh air.
      Whether it really is as good in the end, as the Roberts would like, we will find out all probability even, perhaps even this year. However, as with a rise in the unknown regions of a high mountain, we a significant degree will it still have to be patient.
    • By Schatten
      ...by german gaming magazine Gamestar (in english). Please don't make fun of our accent.
      http://www.gamestar.de/videos/chris-roberts-im-interview-wie-stehts-um-star-citizen-und-squadron-42,90207.html
    • By VoA
      An interesting article on Scale (size of the sandbox)..... as it relates to Star Citizen.  See link below or spoiler...
      Space To Ground: Matters of Scale in Level Design
       
    • By VoA
      Massively Overthinking: The Star Citizen money machine   <<-- See link or Spoiler below...