The Star Citizen Thread v5

Status
Thread Closed: Not open for further replies.
Thanks very much for the explanation. I have suffered that in Battlefield 3, where you think you are behind cover but still get shot. Think they fixed it eventually.
I did watch some of that Guinness world record on youtube and also watch Ross Scott (he of Freeman's Mind fame) play it once a month. I seem to get more enjoyment watching people play this stuff than doing it myself. Day Z is another example. Thanks again.

To stay on topic. Go melt all your ships.

I will say, the experience of working as part of a team to get stuff accomplished in PS2 is awesome. The downside: there ain't much to accomplish besides taking a series of bases over and over and over and over and over again.
 
I will say, the experience of working as part of a team to get stuff accomplished in PS2 is awesome. The downside: there ain't much to accomplish besides taking a series of bases over and over and over and over and over again.

Yeah but it's how you do it that makes it fun. Bit like chess I suppose. Same board, same pieces every time you play.
 
As a former PS2 player of a few years, the correct answer is "not well."

When you get into the larger battles, several hundred players all fighting, then you get issues of infantry or vehicles not rendering or popping randomly into and out of existence. You get hitreg crap going on. We had a "RecordSmash" battle where the PS2 devs unlocked the player cap on a server so that we could get the Guinness record for the most players in an online battle. That was an absolute mess. In the main battle it was just people spamming grenades/rez grenades.

The main issue with PS2 was "clientside" where each player would have the instance on his machine, including the locations of other players and all weapons fired or movement made would be calculated from your computer and uploaded to the server, and then sent on to other players. So in a fight, from your perspective, you might make it behind cover but on an enemy's screen it still looks like you are in the open. So you get shot and killed even though there's no way a bullet could have hit you. For me this is the biggest issue with mmos like SC and PS2. There needs to be an excellent networking foundation that isn't cludged together (ahem, I love ED, but c'mon, really?? fix your network code!). In PS2s case it was even worse, because they released the game before it was fully optimized. So the mixture of lag, frame drops, warping and so on made a lot of people's first experiences in the game very poor. I think this has mostly doomed it to being a low-pop game when it could have been very good and more popular.

So, anyway, the whole network angle is the part that I am interested in the most. If SC or Dual Universe can get something working that minimizes "clientside" and makes for a mostly bug free mmo fps space experience, I will be amazed.

PS2 uses a "bubble" system, meaning that you get sent position updates for entities within a certain area around you, but not beyond that. PS2 does include a priority system (a engineer turret is visible even from farther away, whereas a single infantry soldier is not) and they scale that bubble depending on the amount of players. In very big battles, you can run into situations where troops and tanks just "pop in" as they get close to your bubble.

And of course, with something as massive as PS2 you simply cannot do everything serverside. The best you can do is verify post-factum serverside and punish potential cheaters, but PS2 is not doing a great job with that. They are doing a great job with prediction and lag compensation however, especially considering the scale of things.

Still it's a great game and one of only three online games I know that can handle more than 250 players in one battlefield. (The other two were Planetside 1 and Joint Operations)

Incidentally, another game that used a similar system (for some weird reason) was Falcon 4.0 - and indeed that system was responsible for a lot of hassle during development, because it's not easy.
 
From that video my overall feeling is we'll eventually see that SQ42 presentation, looks like it could be impressive, but to actually crunch a complete game? Can't see that happening for 2017, those guys were loosing "eight weeks" sleep trying to put that one level presentation directed at known hardware together, one linear pre-selected route of defined "game-play". I'm thinking they'll kill themselves trying to put a release out.

No, that thing is not ready for 2017 release it seems, even Chris himself wasn't sure during CitizenCon presentation. Yeah, it is actually very good question - how much game is it there? I assume there's lot of assets, mocaps, but how does it make complete game? SQ42 slides Citizencon - again, information provided by CIG itself, so they are not hiding - show 70% of gameplay stuff NOT completed. "Basic things"? What basic things? Pathfinding is a basic thing. It is not there yet. Flying is not ready yet.

As for whole gizzillions players in one instance thing I already said - someone with indepth knowledge smirks at such claims. Smart network code can do much, sure. Major issue is when you optimize thing, you sacrifice a lot of fidelity. Also server costs go up and up. And this is while chasing a pipe dream of "1000 players per instance", when having 50 - 60 perfectly working would be nice intro.
 
Last edited:
Star Citizen is now my favorite reality TV show.

They should do a spin off called Derek Smarts development hell, where Derek goes into game dev companies and yells at people. Like Gordon Ramsay but with less food and more swearing.

Edited. Today at 12:33 PM Reason: suggested catchphrase not allowed past swear filter

I wanna see this show - after the watershed

I loved Robert's explanation of how the super-instancing was going to work - I reckon even a non-codey person could think it through and smell the litter needed changing
 
Sure Derek I notice that they avoid to talk about netcode&networking on Citizen Con and very quietly they mention that "soon"we will be able to play 4vs4 and 12Vs12 in AC and Star Marine which is almost no improvement from the vanilla CE netcode.....but here they are on new episode of Reverse the Verse and the man himself was explaining this magical new tech and how SC is going to be able to handle this"crazy"numbers of players per instance....So I don´t know I need it to ask.....that really sounds all great&perfect almost as a dream but kind of possible in the future then yet it was never done before.....well who knows....I wish that this goes"live"not just for the sake of SC but for the sake of all future gaming......

Out of interest, is anyone here familiar with Mechwarrior Online? Because as far as I can see, it runs on Cryengine 3 and has supported 12v12 for a few years. I don't recall Piranha testiculating wildly over netcode, but it not a game I've picked up so maybe I'm missing something.
 
In that ATV video the guy says:

"it's not a tech demo, we're showing off what they system we built is going to support".

I'm actually not sure what a tech demo is if that isn't exactly what one is.

Call me old fashioned, but isn't a demo when you show a part of a product that your trying to sell? If your just showing the components of the technology your using to make the product, that's the perfect definition of a tech demo.

Unless anyone else can correct me on what a tech demo is, I'm baffled that they think it's something else.

Looked like a whole lot of hard work and a real shame that sort of crunch isn't being directed at the actual product directly rather than trying to appease citizencon, those guys looked exhausted.

From that video my overall feeling is we'll eventually see that SQ42 presentation, looks like it could be impressive, but to actually crunch a complete game? Can't see that happening for 2017, those guys were loosing "eight weeks" sleep trying to put that one level presentation directed at known hardware together, one linear pre-selected route of defined "game-play". I'm thinking they'll kill themselves trying to put a release out.

The issue they don't have endless amount of money to keep delaying things, they need to start earn cash from other sources that JPEGs sales and soon if DS is right.

If they are still grey boxing things out in SQ42 then the thing doesn't stand a chance being release in 2017, games of this complexity usually need about half a year of debugging and polishing once all of the assets and technical foundations are complete.

I think 2018 Easter Holidays is likely the time they will release chapter 1 and even that may eventually be scaled down from what they are current envisioning.
 
Last edited:
PS2 uses a "bubble" system, meaning that you get sent position updates for entities within a certain area around you, but not beyond that. PS2 does include a priority system (a engineer turret is visible even from farther away, whereas a single infantry soldier is not) and they scale that bubble depending on the amount of players. In very big battles, you can run into situations where troops and tanks just "pop in" as they get close to your bubble.

And of course, with something as massive as PS2 you simply cannot do everything serverside. The best you can do is verify post-factum serverside and punish potential cheaters, but PS2 is not doing a great job with that. They are doing a great job with prediction and lag compensation however, especially considering the scale of things.

Still it's a great game and one of only three online games I know that can handle more than 250 players in one battlefield. (The other two were Planetside 1 and Joint Operations)

Incidentally, another game that used a similar system (for some weird reason) was Falcon 4.0 - and indeed that system was responsible for a lot of hassle during development, because it's not easy.

The bubble system work fairly well in PS2 because you can hide things behind buildings and the terrain.

Not sure you can really do that in a space environment.
 

Slopey

Volunteer Moderator
Out of interest, is anyone here familiar with Mechwarrior Online? Because as far as I can see, it runs on Cryengine 3 and has supported 12v12 for a few years. I don't recall Piranha testiculating wildly over netcode, but it not a game I've picked up so maybe I'm missing something.

MWO is a bad choice for comparison. They dropped 12v12 previously and only reintroduced it after many many months after the networking simply couldn't cope with it. The lack of more than one or two endlessly repetitive 4v4 game modes killed the multiplayer side of it immeadiately. Matchmaking was a joke, and you were limited to 4 player sets, which even when 12v12 came back in, you were on the wrong side. Then they implemented paid matches (using in game currency). It was horrendous.

PGI were simply awful for many many years, farmed mech sales for money, broke several red lines of their own (3pv etc), and most active players left the game (myself and the clan I played with included) several years ago.
 
MWO is a bad choice for comparison. They dropped 12v12 previously and only reintroduced it after many many months after the networking simply couldn't cope with it. The lack of more than one or two endlessly repetitive 4v4 game modes killed the multiplayer side of it immeadiately. Matchmaking was a joke, and you were limited to 4 player sets, which even when 12v12 came back in, you were on the wrong side. Then they implemented paid matches (using in game currency). It was horrendous.

PGI were simply awful for many many years, farmed mech sales for money, broke several red lines of their own (3pv etc), and most active players left the game (myself and the clan I played with included) several years ago.

That'll be what I'm missing, then. :) It's almost as if Cryengine 3 isn't the best choice for this sort of game...
 
That'll be what I'm missing, then. :) It's almost as if Cryengine 3 isn't the best choice for this sort of game...

It is good for single player graphically impressive games, but not much beyond that. There were several attempts to make open games with CryEngine that fell hard, and multiplayer was incredibly limited with it.
 
Has anyone seen this tweet from Scott Manley https://twitter.com/DJSnM/status/787342856469876736 ? Star citizen backers actually buying adspace to get referral credits for the game, I guess spamming gaming websites wasn't generating enough income lol.

Give it another few months and the codes will be turning up in your email...

"Refactor your gaming life with this simple trick! You'll be amazed at the fidelity!"

[ugh]
 
Last edited:
Generally I think the amount of drama they spread over the internet speaks of nothing but insecurity - these aren't the actions of investors confident in their returns for sure. If anything it does wonders for proving that those closest to the game are very much aware of what trouble it's in
 

dsmart

Banned
No, that thing is not ready for 2017 release it seems, even Chris himself wasn't sure during CitizenCon presentation. Yeah, it is actually very good question - how much game is it there? I assume there's lot of assets, mocaps, but how does it make complete game? SQ42 slides Citizencon - again, information provided by CIG itself, so they are not hiding - show 70% of gameplay stuff NOT completed. "Basic things"? What basic things? Pathfinding is a basic thing. It is not there yet. Flying is not ready yet.

As for whole gizzillions players in one instance thing I already said - someone with indepth knowledge smirks at such claims. Smart network code can do much, sure. Major issue is when you optimize thing, you sacrifice a lot of fidelity. Also server costs go up and up. And this is while chasing a pipe dream of "1000 players per instance", when having 50 - 60 perfectly working would be nice intro.

Well, back when I was tweeting that this is what was going on, those guys were crying foul and saying that I was making stuff up.

http://www.kotaku.co.uk/2016/10/15/heres-why-we-didnt-see-star-citizens-campaign-at-citizencon
 
Ah, the "you said they'd run out of money and they didn't! lol!" argument. As if the fact that those same people had to dump tens of millions of extra dollars into it to keep it afloat had nothing to do with it.
 
Last edited:
MWO is a bad choice for comparison. They dropped 12v12 previously and only reintroduced it after many many months after the networking simply couldn't cope with it. The lack of more than one or two endlessly repetitive 4v4 game modes killed the multiplayer side of it immeadiately. Matchmaking was a joke, and you were limited to 4 player sets, which even when 12v12 came back in, you were on the wrong side. Then they implemented paid matches (using in game currency). It was horrendous.

PGI were simply awful for many many years, farmed mech sales for money, broke several red lines of their own (3pv etc), and most active players left the game (myself and the clan I played with included) several years ago.

Not to mention that PGI is totally hooked on selling mech packs for $$$ to the detriment of game development. The whole "community warfare" fiasco is case-in-point. They promised a persistent universe in which players and their outfits could capture planets, earn rewards, etc etc. What they gave is a very shallow 12v12 or 4v4 mode. If your outfit manages to secure a planet nothing really changes for anyone. There's no point.

Also, in terms of netcode and combat, when two mechs make contact often they will warp right through each other. They also have had to speed cap mechs to about 200km/h as anything higher than that warps around and is unkillable.

Okay, I swear that MWO and PS2 are the two games I know the most about I won't try to act like an expert with anything else :)
 
Status
Thread Closed: Not open for further replies.
Back
Top Bottom