Reply To: Star Citizen – Musings
So last week, German (It has to be them, because US media mostly don’t give a shit anymore because they know what’s coming) magazine, GameStar, had an interview with CIG whereby they claimed to have “played” the upcoming Alpha 3.0. You know, just like they have all these past years even though basically nothing they’ve written, has actually been released yet. And they did the same thing, on the same subject, almost a year ago – again to another German magazine. You should see all the ridiculous claims and promises in that one.
Anyway, since it’s the usual Shillizen nonsense, especially with the upcoming GamesCom (Aug 22-26) coming up in Germany, most of us just laughed. You’d be surprised how much funnier this Star Citizen crap is when you’re reading a German to English translation. Not to mention the irony of backers having to read critical info from the media. This after having to date donated almost $155M to the project. It’s hilarious.
Aside from my usual Twitter trolling for lols, I mostly ignored, and discounted it as the usual rubbish that only desperate backers would pay any attention to. The backers in the Reddit threads (1, 2) were mostly aghast and/or pensive, for the most part. When I finally got to read a proper translation over the weekend, I realized just how right I was. It’s all the usual pandering bullshit, with zero accountability for the fact that the project, after 6 years and $155M (backer money only), is nowhere as complete today, as it was back in 2015. Without any bias, I say that with the utmost sincerity. Below are all the major milestone releases. Here are all the patch releases. Also in April 2015, they revised the patch numbering scheme.
- 3.0 (Moons) is planned for Aug 2017
- 2.6 (Star Marine) // Dec 2016
- 2.0 (Persistent Universe + Multi-Crew) // Dec 2015
- 1.2 (ArcCorp Social Module) // Aug 2015
- 1.0 (Arena Commander) // Dec 2014
- 0.x (Hangar Module) // Aug 2013
So now it’s looking a lot like, with half the year gone, and the last 2.6.3 patch having been released back in April, that this year’s biggest update is going to be 3.0 (Moons!). And according to Chris Roberts at GamesCom 2016, it was supposedly due (On Nov 2, 2016, I wrote an article which cited sources had indicated that 3.0 didn’t even exist) back in Dec 2016; but backers got a consolation prize in the form of the immediately forgettable (seriously, nobody is playing it) Star Marine. And when you look at the sheer amount of work left to do, it’s easy to see how insurmountable their task it.
But enough of that. I wanted to talk about some specific items in the article that caught my attention, and which have also been the subject of much talk and controversy.
SQUADRON 42 DELAYED – AGAIN
Yeah, shocking and completely unexpected.
You do know that since Summer 2015 I’ve been saying this, right? And that since SQ42 relies on ALL the tech for Star Citizen, that there is no way in hell that game ever comes out without that tech being in place. They denied it in 2015. It didn’t release. They denied it in 2016 – even went to the media and said the rumors were rubbish. It didn’t release. It’s not coming out in 2017 folks. So stop talking about it.
MISSION COUNT
They are planning for 7 – 14 missions, depending on how things go with the 3.0 release. This is interesting because, like with 2.0x, in which they created some space missions which quickly became repetitive, that’s basically where they’re going with this too.
SPACE <–> PLANET TRANSITION
The article says that via the starmap, they jumped to the outside of Delamar (moon), then flew down through the atmosphere to the surface below.
If you have been following my writings about this (1, 2), then at this point you can safely utter the words “Derek Smart was right” because since last year when Chris Roberts was touting procedural planets and all that rubbish, I had said that due to the engine, they probably couldn’t do whole planets, let alone procedural ones, or features promised such as orbiting planets, atmospheric day & night effects etc. Instead, they would have to create these surface areas as they would a standard “level”, using a combination of procedural (terrain and asset generation), and hand-crafted areas (derelicts, landing bases). And in the end they would have to access them the same way that Elite Dangerous does. In fact, what you would end up with is basically another entity object similar to their base in space with a landing platform, but created as a moon. Note that fps on planets was a $20M stretch goal, and they got a $1M spending bonus for procedural tech R&D when they hit $41M.
As a game developer and designer, I really have no problems with this because you have to work with what you have. And that’s the problem with making promises that you have to keep down the road. However, when you fund your own game, along the way, you can add and remove anything you want – with impunity and without consequences. Even in Early Access. But with crowd-funding, as the FTC and State AGs have said in all the cases they’ve pursued, if you make a promise, you have to keep it. No exceptions. You can’t take money to deliver a Gold box, then deliver a brass box, and say that’s the end of that, you’ve delivered.
Anyway, if they pull off this Minimal Viable Product of planetary access in 3.0, and assuming they can overcome the performance issues I’ve been writing about, that should probably keep some backers happy. Until they run through the content in one sitting; realize it’s all repetitive and shallow, and not really a “game”. Then we’re back in the lol trenches again.
“An artist can crank out five or six moons in a week for you,” Roberts told us, emphasizing that “once you’ve got your building blocks, somethings will be quicker. There isn’t going to be a matter where we hit a magic number and, ‘boof,’ here comes a planet” –Chris Roberts on procedural planets in Sept 2016.
Landing on a moon and base in Elite Dangerous
CRYENGINE USE
Marco Corbetta and Carsten Wenzel saying that the StarEngine currently has 10% CryEngine and 90% of their code, is interesting. It makes absolutely no sense, seeing as they claimed to have switched to LumberYard in about a day. That was back in Dec 2016 when 2.6 was released.
LumberYard is based on CryEngine 3.x, and Amazon didn’t make any fundamental (in dev speak) changes to it, other than bug fixes, various improvements, as well as adding AWS support and some supporting features for it. It’s all right there in their changelog. If what these guys are saying is true, and I don’t doubt that it is, then it completely confirms my theory that they didn’t “switch” to LumberYard at all. Instead, they merged the parts (e.g. as of this writing, they still can’t get the LumberYard implementation of Render-To-Texture to work) they needed, replaced Google Compute with AWS as is required by Amazon, and continued on from there.
From what I have heard, this move could be due to their original CryTek engine license which may either have royalties, or some sort of “units sold” threshold, like most licenses from years ago used to have. Which also begs the question of how they are going to get around the issue of the competing (LumberYard) engine clause, which, anyone who has seen a legacy CryTek engine license contract, probably knows is one of the bullet points.
In addition to the above, some have speculated that this switch is probably due to CryTek’s financial state, their ability to provide on-going support for CryEngine etc. Well uhm, if you’re only using 10% of CryEngine, what do you need CryTek support for?
“I talked to publishers about doing a fourth Elite game, but some things happened. Publishers were skeptical of space games in general because of the financial failure of Freelancer, an early 2000’s game. It was delayed. It’s a nice game, but in that period, they were just incredibly skeptical.
—
When we first greenlit Elite: Dangerous, there were no other major space games since Freelancer. Now, there are dozens. So, I think we’ve succeeded. We’ve brought the genre back to life. And we’ve proven there’s quite a lot of demand for this sort of game. Yes, it’s niche, but it’s quite a big niche. And we’ve got Chris Roberts coming along now, and so many other games that look interesting. No Man’s Sky, even.” – David Braben, Rolling Stones, 2016
SERVER CAPACITY & CLIENT SYSTEM REQUIREMENTS
They’re still talking about 24 clients on a server, though nobody seems to want to mention that it’s a theoretical limit and which only works in the social module. Any game instance running more than 8 clients, is a horrid experience. So with the much touted networking core improvement – which isn’t going to make much difference anyway – moved out of 3.0 and now defered into 3.1, with the added moons in the instance, it makes sense now why they are having performance issues with 3.0.
Listen, lets face it, with their current engine and architecture, they’re never going to be able to build the MMO they touted. Assuming they survive even a year past the 3.0 launch, there is no way they’re going to get from where they are (an instanced session based game) now, to an MMO (instanced or otherwise).
Not to mention the fact that, just like Elite Dangerous, the game’s architecture means that if they can’t make enough money to be paying huge AWS bills for their cloud servers, they’re going to have serious issues. It’s almost as if they should have been thinking about private servers – as promised – years ago so that when (not if, they’re fucked – completely) the project collapses, at least backers can continue to play the game thereafter.
They also claimed to have played the 3.0 demo on a 32GB machine with an i7-5930K CPU and Nvidia GTX 980 GPU. They claimed 30 fps, without specifically saying whether that was in space or on the Delamar moon, nor how many people were in the demo. But who cares about such details, right? Right now, there are backers with beefier systems having horrid performance with 2.6.3. But somehow for 3.0, they’ve come up with some magic Juju that’s going to make an upcoming build with a major content update, run on those same systems at the same or faster performance rate. I can’t wait to play it.
A TOTAL OF 5 – 10 SYSTEMS AT LAUNCH
Chris Roberts (@14:14) talking about 100 star systems – Sept 2015
This one has been the cause of much discussion since the article was published.
Considering that at $6M stretch goal they promised 100 systems, with 7 (1 exclusive to pre-launch backers) more promised between $18M and $40M stretch goals, and 500+ planets, with even the game’s universe description still touting 100 systems at launch, this one was hilarious to say the least.
Hey, at least they seem to have confirmed what I’ve been saying that the world isn’t procedurally generated at all. Which, now that it’s clear, I can see how they are now talking about 5 – 10 systems at launch because, given the tools we’ve seen, and the sheer amount of work required to build each of these “levels” with points of interest, landing zones etc, it would take another decade or more for them to build the world they promised. Shockingly, a world this size, was built by a single person over two decades ago, using a combination of scripts and procedural techniques. And it contains entire planets and moons, complete with various eco-systems, climate zones, planetary day/night transitions, orbiting & rotating planets and moons etc. Ah yeah, good times.
A lot of backers forget that though they were in the stretch goals, CIG announced procedurally generated planets back in February 2014. They showed an R&D demo during the holiday stream in 2015. Then in July 2016, again in a German magazine, they announced it as coming in 2.7 (the patch that became 3.0).
Anyway, this sounded so ridiculous to me, that I held of commenting on it before hearing back from my sources. Excerpt says it all.
“Nobody here reads what chris and erin tell the media. We still have to deal with scripted and directed shows most don’t want to be in. So if he told G* we’re planning that number at launch, I personally don’t know about it. It could be one of those things he just prattles on about or a lost in translation thing. Did they say he said that, or was it coming from a designer they met with? We don’t even have a launch date for either game, so how can that person know how many systems we would have at launch? Did you see the chatter on Spectrum related to Levski not being in 3.0 and a designer who should know but didn’t? It’s like the loan, none of us knew until we read about it online. I didn’t know about it until I got the chatter about Ortwin’s official statement on Spectrum. I saw your other email about an MVP. We don’t have that. Nobody is working toward that. We are just working on what we have to and doing the best that we can. I will check around about that other thing* and get back you, but I don’t know what they are planning now because 3.0 isn’t in any state to be released in Aug. Keep watching the schedule for the pattern I mentioned in my other email.”
* That was a prior exchange related to whether or not they would be releasing 3.0 as-is because of GamesCom, or if they would delay it again past Aug 25th in order to address the missing features and performance issues. If you haven’t been keeping up with the dev schedules, you should read the analysis of the 07-14-17 one in which a bunch of things were delayed, but the target release date never changed. Yeah, they’ve invented a time dilation machine.