I look forward to the Lumberyard Engine Celebration ship bundle.
I’m assuming this question wasn’t serious, but I would recommend waiting until 3.0.
I laughed out loud. Well done.
Wow this is mental. There is some hot bullshit going on over at CIG. The holiday livestream would have been a perfect opportunity to talk about this to their community but instead, new ships and janky FPS multiplayer.
They have banged on for years how much modification they have done to CryEngine, such that they (or the community) have even coined it’s own name - StarEngine, with commentary that even official CryEngine patches and features now take time to be worked into SC.
Now bam, new engine! From out of nowhere. No announcement, no forewarning, no transparency, just a announcement following the Lumberyard logo making an appearance on the splash screen?!
There is speculation that perhaps they are/will still use StarEngine, but just take the networking and cloud services modules from Lumberyard with a subsequent switch from GCE to AWS. But won’t that that at least mean the last year or so of work on supposed re-writing of the netcode are now wasted? Didn’t one of the recent patches just roll out some much anticipated netcode changes? Is that their own work, or Lumberyard?
I don’t believe for a hot second that this has been or will be a painless ‘like for like’ switch. Either everything that has been delayed for the last 6 months or so is a result of this change being kicked off a while ago which has been kept quiet until now, or the project is due for future pain and delay as they switch and uncover all the gotchas.
I think it speaks volumes that this kind of change isn’t so much announced, as it is discovered as part of a patch two weeks after a major streaming event. If they had legitimately made a painless change, they would have screamed it from the rooftops in a live-stream as evidence of their progress/capability. On the other hand, if the switch is so painless as to avoid a passing mention until the logo is on the splash screen, how much truth is there to the heavy customisations they have had to do to get the engine to meet their groundbreaking requirements?
Either way, somethings smells about this announcement and I am rather surprised their forum has not exploded.
No, it doesn’t mean that.
Most of what they did would have nothing really to do with the low level stuff that would be included in the core engine.
Again, the shift to lumberyard almost certainly just means incorporation of some of amazon’s cloud infrastructure junk.
What? They have spent over a year talking at various times of their move to rewrite netcode services from scratch (aka StarNet). Initially intended for 2.6, then 3.0! That is some pretty low level engine stuff and if they abandon that now for the Double Helix netcode in Lumberjack, that is wasted investment.
It’s of course unclear right now if that will be the case or whether they will persist with whatever they have developed. But if they plan on switching to AWS services, it stands to reason they might well adopt Lumberjack netcode which is presumably optimised for it and for which Amazon will maintain support.
Not necessarily the wrong decision if that’s the case, mind you (netcode is a challenge and at the end of they day switching to something that is proven to work well and is licensed rather than built and maintained may well be the better financial path), but it raises questions about long term direction and how money is spent at CIG. Every time they do something like this (ship re-modelling, animation re-rigging, flight mode changes, engine changes), it redirects time and resources away from content development (actual game mechanics - mining, trading, etc). And they have plenty of that other stuff to do - even to get SQ42 out the door.
Certainly without specific design documentation I can’t offer a concrete argument here, but I can tell you just from looking at what they’ve implemented that the majority of the changes take place at a level of abstraction which is well above the basic netcode included with the engine.
The framing system they’ve developed is not something which is really included with the engine at all, and it’s what they’ve spent the majority of their time on. The low level network implementation at the bottom could be swapped out, while preserving all of that work… and that’s even if they bothered to swap it out. Moving to lumberyard wouldn’t necessitate such a swap anyway.
Again, without detailed documentation anything is possible, but I think it’s silly to assume that based on what we’ve seen that this would constitute some massive shift. What you’re likely to see is that CIG takes the stuff Amazon brings to the game with lumberyard, and rolls it into their own stuff, which will likely play fairly well sitting on top of lumberyard since it’s the same core engine anyway.
I mean, yeah… they re-wrote a bunch of stuff from the CryEngine… and that re-written stuff will continue to exist, replacing/adding to stuff in lumberyard.
The stuff that’s offered by amazon in this regard is super low level compared to the framing system CIG’s made… it’s essentially a replacement of a low level API at that point. It would likely require some rejiggering of things, but not some kind of total reworking of all the higher level stuff they made, assuming some degree of competence in how they designed it.
Updated: I’ve had a response from CIG director of communications, David Swofford, to say that the relationship between CIG and Amazon is that of them being a regular licensee of Amazon’s technology. The reason for the announcement today was that it was turned on with the release of 2.6. He also confirmed that all the work CIG had done to expand the CryEngine has been transitioned to the new engine.
I literally thought this was some sort of joke…it had to be. There is a very short Wikipedia page on it. It went into a preview/open beta in November. It has one named game on the page using it…guess which one.
Man I hope there is a good book written on Star Citizen someday. Could be the greatest gaming story ever told. In a giant slow trainwreck sort of way.
i wasn’t joking but i realize now there are other finished games i could spend my money on.
This will be accurate for many years.
I don’t think the switch from CryEngine to Lumberyard is that big of a deal. The latter is a fork of the former - and if CIG started doing this a while ago already, the gap probably wasn’t that big. It means work for sure, but it’s certainly not as messy as that one project I know where a company switched from Unity to Lumberyard.
The whole “SC switches game engines” headlines thus are a bit misleading and make people think that CIG had to go back to the drawing board technology-wise or something.
Hmmm is Lumberyard mature enough for the switch? I was doing a quick evaluation on it a few months ago and walked away thinking they have some ways to go, got turned off mainly by documentation (or lack there of).
I agree this is not much of a piece of news. Both engines are very similar and the switch probably was not a huge drain of resources.
Probably they are keeping their own net code and just switching to Lumberjack cloud service for user and server administration, with the low level networking untouched.
Also, Lumberjack has more chances of being updated and maintained than CryEngine at this point.
As for documentation, CryEngine was never really mature either. There’s a reason so few projects use it.
I would have thought they would have learned from experience not to do the same.
I’m also assuming that CIG is getting some money out of this either in direct cash or through a very sweet deal on AWS services in general. Amazon certainly was keen on having some sort of flagship title in the Lumberyard portfolio. And it’s not like they simply switched and mentioned it in a minor note - nope, with a press release singing the praise of Lumberyard and Amazon in which the CryEngine does not get mentioned with a single word. Pure advertising an typical of such arrangements.
[quote]Star Citizen and Squadron 42 Utilize Amazon Lumberyard Game Engine
Highly-anticipated space sim games will take advantage of Lumberyard’s deep integration with Amazon Web Services (AWS) and Twitch
Los Angeles, December 23, 2016 — Cloud Imperium™ Games (CIG) announced today the company is using the Amazon Lumberyard game engine to create its ground-breaking space sim games, Star Citizen™ and Squadron 42™. Both games are currently in development and are backed by a record-breaking $139 million crowd funded effort.
Amazon Lumberyard is a free AAA, cross-platform, 3D game engine that empowers game developers to create the highest-quality experiences, connect their games to the vast compute and storage of the AWS Cloud, and engage fans on Twitch.
“We’ve been working with Amazon for more than a year, as we have been looking for a technology leader to partner with for the long term future of Star Citizen and Squadron 42,” said Chris Roberts, CIG’s CEO and creative director. “Lumberyard provides ground breaking technology features for online games, including deep back-end cloud integration on AWS and its social component with Twitch that enables us to easily and instantly connect to millions of global gamers. Because we share a common technical vision, it has been a very smooth and easy transition to Lumberyard. In fact, we are excited to announce that our just released 2.6 Alpha update for Star Citizen is running on Lumberyard and AWS.”
“Star Citizen and Squadron 42 are incredibly ambitious projects which are only possible with great engine technology paired with the transformative power of the cloud. We love how CIG’s bold vision has already inspired a massive community, and we’re thrilled to see what they create with Lumberyard, AWS, and the Twitch community,” said Dan Winters, head of business development for Amazon Games. “We’re excited that they’ve chosen Lumberyard and AWS to provide the performance and scalability they need to bring their games to a massive audience.”
Added Roberts, “We are delighted to be working with a partner with the strength, vision, and resources of Amazon Web Services. We are looking forward to developing our relationship with AWS and the Lumberyard community in the future.”
Star Citizen is 100% crowd funded and was officially announced on October 10, 2012. The money raised pays for the development of the game including the Roberts Space Industries platform where fans and backers can interact with the team, view multiple weekly webcasts about Star Citizen, learn about the story behind the game, read constant updates on the game’s progress and much more. Star Citizen is recognized by Guinness World Records as the top crowd funded project and game in the world. People interested in backing the project can do so at http://robertsspaceindustries.com/pledge.[/quote]
I’m waiting for the announcement in 2018 when Chris Roberts says he has to redo all the motion capture for a new platform.
The commercial ships will have Amazon paint job and free delivery anywhere in the universe (by space drones).
And wow a new source of revenue selling corporate paint jobs! Will be like the Jetsons with space billboards all over the place!
New trailer for Space Marine in 2.6:
Not the part of the game I’m interested in. But I have to say the character and weapon models as well as the environments are pretty cool.
Wendelius
When I worked at Auran on Fury we had some massive performance issues in UE3. The problem stemmed that when we originally started using UE3 it only had optimizations for the 360 (due to how young it was at the time we adopted it, which was after the GOW that used it). By the time Epic had done a lot of work to make it work well on PCs we had already done enough modifications to the engine that backporting any of Epic’s updates would have been extremely time consuming and non-trivial.
And I am pretty sure that all the updates to the engine our programmers did was no where near the complexity CIG had to do (things like the 64-bit update for example).
EDIT: Point being, even though the engines started from the same base doesn’t mean it’s trivial to backport changes from one fork to another. It can be very very hard depending how much the forks diverged, and CIG made it sound like they spent a lot of money to do massive divergence in their fork.