|
Bubbacub posted:James Webb Space Telescope: 10 years late and $9 billion over budget compared to the original schedule At least the telescope will exist in some fashion.
|
#
?
Oct 9, 2019 21:27
|
|
C-Bangin' it 
|
|
| # ? Jun 5, 2024 15:26 |
|
|
So? It's bigger than the Hubble, isn't it? Like, what do you want? a fair comparison? Hopefully this is not a snipe e: gently caress yeah! taxxe is on Colostomy
|
|
#
?
Oct 9, 2019 21:27
|
|
|
https://www.twitch.tv/videos/492369699?t=07m11s they cant even say that richard garriott is involved with the company.
|
|
#
?
Oct 9, 2019 21:30
|
|
|
Dark Off posted:https://www.twitch.tv/videos/492369699?t=07m11s Lol yeah this is just Garriot washing his hands of that mess
|
|
#
?
Oct 9, 2019 21:48
|
|
|
Dark Off posted:https://www.twitch.tv/videos/492369699?t=07m11s Why do almost all MMO's look the same?
|
|
#
?
Oct 9, 2019 21:53
|
|
|
colonelwest posted:
There have been many peak Star Citizen moments, but for me one of the greatest will be the "let's buy the code and run private instances" attempts and petitions. Bonus points for the fact that once CIG goes tits-up, nobody will "acquire" the code base because it's worthless and broken, so it will just disappear into the air along with the dismantled SAN arrays when the material assets are auctioned off or destroyed. There will be nobody to petition. The humor score will triple every second when backers attempt to contact Crytek for the code base CIG first started on so they can replicate "StarEngine"'s development themselves.
|
|
#
?
Oct 9, 2019 21:53
|
|
|
skeletors_condom posted:Why do almost all MMO's look the same? Because they suck.
|
|
#
?
Oct 9, 2019 21:56
|
|
|
Colostomy Bag posted:That's over an ocean? With a reflection from 1995 video games? The ocean is there to give a sense of scale.
|
|
#
?
Oct 9, 2019 22:21
|
|
|
Zazz Razzamatazz posted:The ocean is there to give a sense of scale. I don't know. How many Skyrims is it?
|
|
#
?
Oct 9, 2019 22:28
|
|
|
Looks like Google is taking a page from Crobo's playback with Stadia: https://www.pcgamesn.com/stadia/negative-latency-prediction "negative latency" - WTF is that? skeletors_condom fucked around with this message at 23:37 on Oct 9, 2019 |
|
#
?
Oct 9, 2019 23:26
|
|
|
!raffle
|
|
#
?
Oct 9, 2019 23:31
|
|
|
skeletors_condom posted:Looks like Google is talking a page from Crobo's playback with Stadia: It does the move before you even think about it
|
|
#
?
Oct 9, 2019 23:32
|
|
|
Colostomy Bag posted:At least the telescope will exist in some fashion. It might blow up on the pad or get hopelessly tangled up in the deployment stage, but yes, it will have some kind of physical presence.
|
|
#
?
Oct 9, 2019 23:47
|
|
|
trucutru posted:It does the move before you even think about it Holy poo poo that's literally what they're saying We’ve received a heads-up (thanks!) that negative latency, powered by a datacentre’s worth of compute silicon, may offer future cloud gaming systems flexibility to anticipate the likely action of a user, and ensure a speedy response ready for that potential eventuality. In what universe does that make sense or even mean
|
|
#
?
Oct 9, 2019 23:47
|
|
|
trucutru posted:It does the move before you even think about it lol you weren't kidding. holy poo poo CIG can use this tech to make boxes clip through the floor before you even try to interact with them
|
|
#
?
Oct 9, 2019 23:48
|
|
|
Rotten Red Rod posted:Holy poo poo that's literally what they're saying It's like OnLive, except this time it won't suck because now we have Magical Bullshit™ technology There's no way this will turn out to be garbage and get shitcanned like so many other Google projects. This one is the real deal you guys
|
|
#
?
Oct 9, 2019 23:50
|
|
|
Rotten Red Rod posted:Holy poo poo that's literally what they're saying Sounds great until you factor in that there will need to be a check after every player action checking whether or not the code guessed correctly, and if it didn't, it will have to recalculate to compensate for the move the player actually made, which is even loving slower than just doing what the player is doing. Having a conversation with someone who tries to finish every one of your sentences is not faster than having a conversation with someone who shuts the gently caress up and listens when it's your turn to talk.
|
|
#
?
Oct 9, 2019 23:50
|
|
|
Looking forward to the incredible gaming future where I experience crippling lag if I attempt to play in a way the developers didn't expressly anticipate. This should be big in China.
|
|
#
?
Oct 9, 2019 23:52
|
|
|
May as well just mash buttons on an attract screen.
|
|
#
?
Oct 9, 2019 23:56
|
|
|
Look just loving upload me into the internet already gently caress sakes
|
|
#
?
Oct 10, 2019 00:32
|
|
|
commando in tophat posted:With the fidelity waiting simulator trains being such a big deal I would expect they at least checked if two trains fit the station. I mean, how does this even happen? It is literally on rails The trains have doors ?
|
|
#
?
Oct 10, 2019 00:39
|
|
|
trucutru posted:It does the move before you even think about it Google invents the aimbot (again).
|
|
#
?
Oct 10, 2019 00:41
|
|
|
Scruffpuff posted:Looking forward to the incredible gaming future where I experience crippling lag if I attempt to play in a way the developers didn't expressly anticipate. This should be big in China. I can't wait for a situation where it guesses every possible action, at every turn, other than the one you're doing. Talk about efficiency! Croberts has done it again.
|
|
#
?
Oct 10, 2019 00:47
|
|
|
This letter is a comedyquote:However, the directors are aiming to minimize such macro risks whilst focusing in upon the micro challenges of the projects they are undertaking and are confident that the current business model will continue to operate efficiently and effectively. Investment into the development and publishing activities of the Company will help maintain its competitive position within the industry and continue the significant progress made in the game development, evidenced by its regular and sustained release strategy maintained throughout 2018 and continuing into 2019 and to date. Competitive position in the industry ?? Position of hopeful yet to deliver a product ? Significant progress made in development evidenced by 2019 releases ? The only wisp of truth in the he whole thing is the company is confident that is current business model will operate effectively.
|
|
#
?
Oct 10, 2019 00:55
|
|
|
skeletors_condom posted:Looks like Google is taking a page from Crobo's playback with Stadia: So... it predicts what button press you are most likely to do and... does that prematurely? If it's still waiting to check what you actually do, it won't really help much. If it's the wrong prediction, does it roll back the whole game simulation? Or does it run two simultaneous copies of the game and switches to the backup one if expected input is not received input? This sounds completely bonkers 
|
|
#
?
Oct 10, 2019 01:33
|
|
|
Subsumption AI is   https://gfycat.com/threadbarejoyouscaecilian From FailureToReport's sober playthrough of 3.7 https://www.youtube.com/watch?v=m9_JCaQLgZ4
|
|
#
?
Oct 10, 2019 01:48
|
|
|
Don't a lot of games deliberately build delays into inputs to make netcode smoother? Maybe negative latency exploits that gap.
|
|
#
?
Oct 10, 2019 01:48
|
|
|
Bofast posted:So... it predicts what button press you are most likely to do and... does that prematurely? If it's still waiting to check what you actually do, it won't really help much. It's bullshit at every level. It doesn't even pass the smallest effort at scrutiny. Say you're playing Pac-Man - the game will predict which way you'll go to make it faster! So let's say if there's ghosts coming from down, up, and left, it predicts you'll turn right. It's the most logical move, the other 3 will result in death. But no matter what it predicts, here we go: 1) It takes more CPU cycles to run a prediction algorithm than it takes to just translate your input directly 2) If you go right, it still has to take your input (he went right) and compare it to what the AI chose for you (he should go right) - that's twice as much effort as just saying (he went right) 3) If you go in any other direction, it STILL takes your input (he went down), compares it to what the AI chose (he should go right), determine you did something unexpected, and take your loving input anyway. I'm convinced it's 100% for the Star Citizen type minds out there. It's complete bullshit. If I lived in any reality except the parody one we're currently inhabiting it'd call it parody.
|
|
#
?
Oct 10, 2019 01:53
|
|
|
TheDeadlyShoe posted:Don't a lot of games deliberately build delays into inputs to make netcode smoother? Maybe negative latency exploits that gap. In theory, it could, because local CPU translation of the move you're most likely to make is inherently faster than waiting for netcode to deliver that move. The result is the illusion of a smoother game. Of course the flipside of this is that if you move in a way other than what's predicted, it's a waste of CPU cycles, the prediction is discarded, and you move at the rate you would have anyway, and the same exact amount of data was sent over netcode. So you're trading CPU cycles on the server for the illusion of very slightly faster gameplay but only if you move in the exact way the algorithm is predicting. They did this in 1983 - it was called Dragon's Lair.
|
|
#
?
Oct 10, 2019 01:55
|
|
|
Scruffpuff posted:In theory, it could, because local CPU translation of the move you're most likely to make is inherently faster than waiting for netcode to deliver that move. The result is the illusion of a smoother game. Of course the flipside of this is that if you move in a way other than what's predicted, it's a waste of CPU cycles, the prediction is discarded, and you move at the rate you would have anyway, and the same exact amount of data was sent over netcode. I think that was cutting edge for the last time Roberts did any programming - when did Stryker's Run come out, again?
|
|
#
?
Oct 10, 2019 01:57
|
|
|
Scruffpuff posted:It's bullshit at every level. It doesn't even pass the smallest effort at scrutiny. Say you're playing Pac-Man - the game will predict which way you'll go to make it faster! So let's say if there's ghosts coming from down, up, and left, it predicts you'll turn right. It's the most logical move, the other 3 will result in death. Not to mention that this approach is unusable on complex strategy games (builders, tycoons).
|
|
#
?
Oct 10, 2019 01:59
|
|
|
Scruffpuff posted:It's bullshit at every level. It doesn't even pass the smallest effort at scrutiny. Say you're playing Pac-Man - the game will predict which way you'll go to make it faster! So let's say if there's ghosts coming from down, up, and left, it predicts you'll turn right. It's the most logical move, the other 3 will result in death. I have not read the article, but off the top of my head, perhaps it predictively renders results of future actions to display, and discards them if incorrect. That could help in simpler games to reduce perceived lag.
|
|
#
?
Oct 10, 2019 02:01
|
|
|
Experimental Skin posted:I have not read the article, but off the top of my head, perhaps it predictively renders results of future actions to display, and discards them if incorrect. That could help in simpler games to reduce perceived lag. That could work, sort of like Nic Cage in Next, but they don't even know what kind of theoretical game would have that many outcomes that the current generation of physics engines couldn't handle completely on the fly. There are countless games out there with AI and collapsible structures reacting to an errant grenade throw with perfect emergent results. This technology is a theoretical solution for a problem that does not exist. It's peak Silicon Valley.
|
|
#
?
Oct 10, 2019 02:03
|
|
|
A new spaceship that mines game outcomes using predictive algorithms that you can sell for space bux.
|
|
#
?
Oct 10, 2019 02:05
|
|
|
Call Your Grandma posted:A new spaceship that mines game outcomes using predictive algorithms that you can sell for space bux. It's actually the perfect technology for Star Citizen to embrace. I present to you, courtesy of your idea, The Warmind. This capital ship not only houses hangars and can scramble fighters near-instantly, but it houses one of the most complex battle predictive algorithms known in Space Future. Send your fighters to intercept the medical frigate before the enemy even realizes they're deploying it. Aim your countermeasures with perfect accuracy at missiles the enemy hasn't even launched yet. Everything that transpires will happen according to your design. You have foreseen it. You command the Warmind. The Star Citizen Warmind: "Things happen before it even sees them." Scruffpuff fucked around with this message at 02:13 on Oct 10, 2019 |
|
#
?
Oct 10, 2019 02:10
|
|
|
Experimental Skin posted:Google invents the aimbot (again). It's going to bring a brand new meaning to "Psychic Dragon Punch". Soon, people will be configuring their Stadia to their favorite level of "yomi" (I just realized that I might be a nerd) -Stadia A: I am going to interpret your flailing on the joystick as a flash kick, dear player. -Stadia B: I knew that stadia A was going to flash kick so I will parry -Stadia A: Actually, I knew you knew I was going to flash kick so I altered the timing -Stadia B: Gotcha! I knew you knew I knew so... Meanwhile, the two "players" continue to rub their ballsack over the buttons, which has been found to be the most agreeable input for the stadia overlords.
|
|
#
?
Oct 10, 2019 02:34
|
|
|
Scruffpuff posted:It's bullshit at every level. It doesn't even pass the smallest effort at scrutiny. Say you're playing Pac-Man - the game will predict which way you'll go to make it faster! So let's say if there's ghosts coming from down, up, and left, it predicts you'll turn right. It's the most logical move, the other 3 will result in death. I wrote a bit about the Battlefield series and instead will say this won't help with the typical FPS game. It's instead all about the feedback loop. Latency is concealed by making MP games client-authoritative, but when you have a high-latency player, they end up doing crazy stuff like teleporting across the room to execute a one-hit knife kill, which you can't react to because on your end you're seeing his position 175 ms ago. If you project future movement, the feedback loop makes it fall apart. Let's say there's a doorway up ahead. I can dash in, I can hold at the entrance, or I can dash across it. The other guy is camping the doorway from the inside. If the game predicts I dash across and I don't, how do you handle that guy firing at my projected self? If I get shot anyway, I'm mad at the game. If his shots don't register, he's mad at the game and I get an easy kill since his accuracy is skewed. The best approach is to change gameplay--don't have instakill melee weapons or guns with a TTK lower than typical latency.
|
|
#
?
Oct 10, 2019 02:41
|
|
|
Bofast posted:Or does it run two simultaneous copies of the game and switches to the backup one if expected input is not received input? This sounds completely bonkers If they are saying the truth that is the way to do it, yeah. Really efficient since each additional player in your (presumably) multi-player experience increases the need for extra computations exponentially.
|
|
#
?
Oct 10, 2019 02:48
|
|
|
Contingency posted:I wrote a bit about the Battlefield series and instead will say this won't help with the typical FPS game. It's instead all about the feedback loop. Latency is concealed by making MP games client-authoritative, but when you have a high-latency player, they end up doing crazy stuff like teleporting across the room to execute a one-hit knife kill, which you can't react to because on your end you're seeing his position 175 ms ago. If you project future movement, the feedback loop makes it fall apart. Let's say there's a doorway up ahead. I can dash in, I can hold at the entrance, or I can dash across it. The other guy is camping the doorway from the inside. If the game predicts I dash across and I don't, how do you handle that guy firing at my projected self? If I get shot anyway, I'm mad at the game. If his shots don't register, he's mad at the game and I get an easy kill since his accuracy is skewed. The best approach is to change gameplay--don't have instakill melee weapons or guns with a TTK lower than typical latency. quote:MP games client-authoritative, but when you have a high-latency player, they end up doing crazy stuff like teleporting across the room to execute a one-hit knife kill, It's the opposite, the tele-kill happens because the games are server-authoritative. (modern ones, stuff like doom and so on is peer-to-peer anything-goes bullshit) Anyways, we have reached such a level of self-parody as a species that I'm pretty sure that if someday we figure a way to time-travel the first thing the scientists will do is use it to shave a few extra milliseconds from the Super Mario Bros speedrun record.
|
|
#
?
Oct 10, 2019 02:56
|
|
|
|
| # ? Jun 5, 2024 15:26 |
|
|
Scruffpuff posted:There have been many peak Star Citizen moments, but for me one of the greatest will be the "let's buy the code and run private instances" attempts and petitions. Bonus points for the fact that once CIG goes tits-up, nobody will "acquire" the code base because it's worthless and broken, so it will just disappear into the air along with the dismantled SAN arrays when the material assets are auctioned off or destroyed. There will be nobody to petition. The humor score will triple every second when backers attempt to contact Crytek for the code base CIG first started on so they can replicate "StarEngine"'s development themselves. Yeah I said it a lot in the old thread, but I think a lot of the hardcore Citizens will never be able to give up the game fully. There will be mod projects/privately run servers for years, maybe even a kickstarter for some full fledged fan-led revival.
|
|
#
?
Oct 10, 2019 03:18
|
|
