|
UnknownTarget posted:I gotta ask - why is the physics grid thing so challenging/innovative? Physics Grids have been a thing since Unreal Tournament 1. Look at Super Mario Galaxy. Same as 2. Way too complicated for CI
|
#
?
Jan 2, 2020 17:47
|
|
|
|
| # ? May 15, 2024 00:46 |
|
|
AntherUslessPoster posted:Why do you need a 24/7 coke party stream from a yacht? Or half of all videogames ever published, come to think of it.
|
|
#
?
Jan 2, 2020 17:57
|
|
|
Agony Aunt posted:To be fiar, that's not that uncommon in software development. Yeah but see, you know what breaks poo poo, you may not know why or how to fix it, but you can point to the cause, CIG can't even do that. happyhippy posted:Was that the one where they were trying to debug carrying a box from the ground into the ship. I can't remember, the thread did laugh for 3 whole pages when the video was posted.
|
|
#
?
Jan 2, 2020 18:27
|
|
|
UnknownTarget posted:I gotta ask - why is the physics grid thing so challenging/innovative? Their engine isn't designed for it. Their engine is designed for massive outdoor areas, so 'everything' falls just DOWN, simplifying the code in the engine itself. SC took this engine, made the 'water' code part space, and is writing on top of that to make stuff like physics grids within physic grids within physic grids within physics grids. Then rewrote it when they found out 32 bits wasn't enough. So normal development may not be that bad compared to what they ended up doing. https://www.youtube.com/watch?v=ZsXnFzYu10k&t=3880s Take a look at this. Watch him run up the ramp, watch his gun. As it goes up into the next physics grid, it wobbles, and then snaps to is most likely a default position. Now that's just a SIMPLE item. The previous year they were whooping for joy getting up a loving box up the ramp. Now watch this: https://www.youtube.com/watch?v=ZsXnFzYu10k&t=3793s When the buggy fucks up minutes before. It has its own physics grid, and it fucks up trying to figure out which bit is in which grid.
|
|
#
?
Jan 2, 2020 18:29
|
|
|
Wow this sounds super familiar: https://www.youtube.com/watch?v=1KOkPEg5GUg
|
|
#
?
Jan 2, 2020 19:06
|
|
|
AntherUslessPoster posted:Why do you need a 24/7 coke party stream from a yacht? How can you have pipelines without a stream? Checkmate FUDster!
|
|
#
?
Jan 2, 2020 19:29
|
|
|
UnknownTarget posted:Wow this sounds super familiar:
|
|
#
?
Jan 2, 2020 19:30
|
|
|
This thread had some posts showing off hellion before it released right?
|
|
#
?
Jan 2, 2020 19:33
|
|
|
A  forums account is totally scammy, dude who owns thousands of dollars in spaceship jpgs that don't exist.You really got us!
|
|
#
?
Jan 2, 2020 19:37
|
|
|
UnknownTarget posted:I gotta ask - why is the physics grid thing so challenging/innovative? Preface: I am a professional programmer, and I have experience working on a space MMO It both is and isn't. Everything in videogames is challenging for reasons that don't make sense if you haven't worked on them. Each piece is generally pretty easy to implement (depending, physics grids are more complicated than some problems) but the problem is getting all your systems to work with each other. Building the physics grids such that your ships can use them, and also building your ships so that the physics grids can support them, and all that needs to come together in lockstep. That's software architecture, and why you pay a few people REALLY big bucks to design your system, and then have the code monkeys follow the architect's design so nothing stupid happens. (And its very very easy for stupid things to happen in code) The problem with just adding them together is that completely defeats the point of having physics grids, is that you can make localized calculations without having to know the full state of everything. If the physics grids are constantly overlapping, you basically just have one giant physics universe, there aren't any grids at that point. The whole point of grids is to be able to chunk things and solve them piecemeal to make faster calculations. "Adding them together" is ultimately what you have to do, but you have to go about it in a way that's efficient and performant. You can't afford to recalculate the inertia of every coffee pot or watermelon slice every frame based on the local gravity of the nearest moon or whatever.
|
|
#
?
Jan 2, 2020 19:42
|
|
|
Zaphod42 posted:You can't afford to recalculate the inertia of every coffee pot or watermelon slice every frame based on the local gravity of the nearest moon or whatever. brb going to ask chris to add n-body to star citizen
|
|
#
?
Jan 2, 2020 19:55
|
|
|
Zaphod42 posted:Preface: I am a professional programmer, and I have experience working on a space MMO Right, what I mean though is that you have a general "world grid" (i.e. everything is 0 acceleration). Then you have a bounding box around an object, and everything inside that box is calculated only to that grid. If two or more come together you're only adding those two grids' influence, you're not factoring in large objects that are far away. So the coffee pot would only be affected by ship 1 + 2, but not the moon they're around as long as they're outside that grid space. I suppose the issue becomes where you have two ships near each other inside a larger grid. Maybe at that point just implement a hierarchy that the larger volume always takes precedence over the smaller volume in a n-intersection. This is both realistic and would limit the physics overlapping to just two at any time. Btw, you guys looking any for any freelance spaceship modelers?
|
|
#
?
Jan 2, 2020 19:58
|
|
|
happyhippy posted:Their engine isn't designed for it. Not that Star Citizen's physics aren't like that normally. They were just extra weird.
|
|
#
?
Jan 2, 2020 20:09
|
|
|
Zaphod42 posted:Preface: I am a professional programmer, and I have experience working on a space MMO I hope you have submitted your credentials to the mods proving your background in space mmo development. As per a certain hardware driver tech at MS this is the only way to validate your viewpoint!
|
|
#
?
Jan 2, 2020 20:11
|
|
|
UnknownTarget posted:Right, what I mean though is that you have a general "world grid" (i.e. everything is 0 acceleration). Then you have a bounding box around an object, and everything inside that box is calculated only to that grid. If two or more come together you're only adding those two grids' influence, you're not factoring in large objects that are far away. So the coffee pot would only be affected by ship 1 + 2, but not the moon they're around as long as they're outside that grid space. Yeah that doesn't remotely solve all the scenarios you need for a space game. I kinda don't know where to begin. Ultimately your top level grid is "the universe" and all other grids are inside that. You have a moon grid, and inside is the ship grid, and inside is the player. Its not a series of grids passing each other, its a russian matryoshka doll of grids. That is the entire point and the challenge. If you have a space ship on a planet, the space ship is inside the planet grid, and inside the ship are boxes. Dude... you're not thinking this through at all. "maybe just implement a hierarchy" is how it already works, how it has to work. "the larger volume always takes precedence" so if a box is inside a ship which is on a planet, the planet always takes precedence? So... you don't have physics grids then. You just have one massive environment.
|
|
#
?
Jan 2, 2020 20:13
|
|
|
AntherUslessPoster posted:Why do you need a 24/7 coke party stream from a yacht? Imagine the endless refactoring. "This coke line needed extra work." "We're having to move the yacht today, because Chris didn't like where we were going" "We asked the backers and they want the yacht moved to Monaco"
|
|
#
?
Jan 2, 2020 20:20
|
|
|
Zaphod42 posted:Yeah that doesn't remotely solve all the scenarios you need for a space game. I kinda don't know where to begin. Ultimately your top level grid is "the universe" and all other grids are inside that. You have a moon grid, and inside is the ship grid, and inside is the player. I think you misunderstood me. I said the larger grid takes precedence when you have an n-intersection of grids. I.e. you have three grids intersecting, sizes 1, 1 and 3. In Spaceship 1, it will only calculate 1 + 3, but disregard the other 1. If you have a 1, 2 and 3 then you would toss 2 and only calculate 1 + 3 for that specific volume. For 2 you would toss 1 and only calculate 2 + 3. It's just to simplify things. So if a box is inside a ship which is on a planet, you only calculate the ship + the planet, and disregard any other intersecting grids. The Russian doll thing is literally what I said, as is the top-level grid vs moon grid vs .etc. You seem to be pretty upset about someone pontificating on implementations in a good-natured Internet discussion, I hope what I said isn't that bothersome to you and I'm just misinterpreting your tone.
|
|
#
?
Jan 2, 2020 20:24
|
|
|
UnknownTarget posted:You seem to be pretty upset about someone pontificating on implementations in a good-natured Internet discussion, I hope what I said isn't that bothersome to you and I'm just misinterpreting your tone. You are misinterpreting, skip the "you seem to be pretty upset" routine  Everybody ITT seems "pretty upset about a videogame" there's no point saying poo poo like that dude. Its text, you can read it different ways. I didn't call you names, move on. You offended I said "you're not thinking this through" ??? How do you talk to people?
|
|
#
?
Jan 2, 2020 20:30
|
|
|
Zaphod42 posted:You are misinterpreting, skip the "you seem to be pretty upset" routine Dude, I'm not trying to call you out or fight. I'm just talking dev poo poo on an internet forum thread about a stupid game. This is literally the most real content the thread has had in weeks. I'm not going to say "chill" because apparently you're interpreting it as me trying to capture you into a "ur mad lol" gambit. So, IDK, maybe just be constructive?
|
|
#
?
Jan 2, 2020 20:33
|
|
|
I made a post awhile back about seemingly failed kickstarters and was rebuked by plenty of great examples. Thinking about it some more I was primarily focused on video games that we're headed by a known name in the industry. If we use this criteria, the only one I can think of would be Koji Igarashi's Bloodstained? I personally played this one and had a shitload of fun and felt that it was worth the wait. It had regular updates, a side mini game that is an actual game and it's basically Castlevania 3 with modern mechanics. Who else is on the list of popular industry developers and producers? Keiji Inafune's Mighty No. 9 was pretty disappointing, with it's legendary mismanagement being more interesting than the game itself. Julian Gollop of X-Com Fame's Phoenix Point, same thing. Yu Suzuki's Shenmue 3 just makes me want to play Yakuza instead. Richard Garriot's Shroud of the Avatar has already been covered here and in the MMO thread, it's a horribly mismanaged flop. I'm sure there is a name or two I missed but they can't hold a candle to star citizen. The cream of the crop is Chris Roberts,who accidentally turned the game into a religion for it's fans and only had concepts and demos desguised as already functional mechanics in a video game and people fell for it. Hell, I almost fell for it back in early 2013.
|
|
#
?
Jan 2, 2020 20:34
|
|
|
UnknownTarget posted:Dude, I'm not trying to call you out or fight. I'm just talking dev poo poo on an internet forum thread about a stupid game. This is literally the most real content the thread has had in weeks. I'm not going to say "chill" because apparently you're interpreting it as me trying to capture you into a "ur mad lol" gambit. So, IDK, maybe just be constructive?  You get to talk poo poo but I have to be constructive.... you're weird dude. I'm done wasting time on you.
|
|
#
?
Jan 2, 2020 20:37
|
|
|
Zaphod42 posted:
That is the most, uh, selective reading of what I said. "Talking dev poo poo" as in "shooting the breeze". Have you never heard of this phrase? Glad you're done wasting time on me, this conversation was not enjoyable, which is a shame because with you as a professional programmer and me as a professional CG artist I wish it would have been more constructive. UnknownTarget fucked around with this message at 20:42 on Jan 2, 2020 |
|
#
?
Jan 2, 2020 20:40
|
|
|
The poo poo with the physics grids doesn't even have to happen. It has to happen because Chris is a moron. There are countless ways to simulate the experience that Star Citizen is trying to sell without all this superfluous bullshit. If ship cargo was just an inventory screen, or if the cargo was just stuck to the inside of the ship graphically, if for the people walking inside a ship it didn't matter if the ship was upside down in an atmosphere but to them it just treated the ship like a game level and let them walk upside down (call it artificial gravity or some poo poo). It could just be a representation of the ship, he doesn't literally need to be in the object itself. So many ifs. If if if if. There's a very particular species of brain worm that Chris has and it means that every physics calculation in his "game" has the be the same calculations that are done in real life. We already have a working model of infinite physics interactions in real-time - it's called "the actual real universe" and it's the one place Chris Roberts seems least familiar with. It's a loving game. Just make a loving game. Scruffpuff fucked around with this message at 20:47 on Jan 2, 2020 |
|
#
?
Jan 2, 2020 20:45
|
|
|
I mean look at the incredible amount of focus CIG has placed on showing demos where a commando picks up a box, tries to walk up a ramp, puts the box down, and the ramp goes up. In ANY other game, you drag the box between inventory screens. Why? Because the physical act of loading the loving boxes is not important. That's what the art of game development actually is - the art of keeping the player immersed by keeping the brain focused on the feeling and the content of the game that is memorable. Carrying a box up a ramp is not immersive; it detracts from immersion. You're supposed to capture the memorable moments, not the moment to moment drudgery of life that every sane person forgets 5 seconds after it happens. Nobody at FedEx talks about the time they wheeled a box up the ramp into the truck. But they might remember a delivery to a celebrity they got to meet. Chris does not understand the difference. Apparently the backers don't either.
|
|
#
?
Jan 2, 2020 20:54
|
|
|
Scruffpuff posted:If ship cargo was just an inventory screen, or if the cargo was just stuck to the inside of the ship graphically fyi thats how it works rn and it's a "bug" that will be fixed by making it so you have to hire people to load it up for you, it'll all be real time and take several (several) minutes and they're also making a system to let the player do it to help i think? this is star citizen you're talking about, if it doesn't take a longass time and isn't an incredibly boring, mundane activity sold as a revolution in High Fidelity Simulation(tm) it's not good enough
|
|
#
?
Jan 2, 2020 20:58
|
|
|
Haven't they been selling really expensive JPEG hauler ships that will carry giant stacks of shipping containers which will of course have to be loaded/unloaded excruciatingly slowly and in full FIDELITY with PHYSICS at a space station? Note how this is a question because nothing like this will actually exist in Star Citizen.
|
|
#
?
Jan 2, 2020 21:01
|
|
|
Popete posted:Haven't they been selling really expensive JPEG hauler ships that will carry giant stacks of shipping containers which will of course have to be loaded/unloaded excruciatingly slowly and in full FIDELITY with PHYSICS at a space station? naturally, immersion is the only thing my prostate responds to these days
|
|
#
?
Jan 2, 2020 21:07
|
|
|
Popete posted:Haven't they been selling really expensive JPEG hauler ships that will carry giant stacks of shipping containers which will of course have to be loaded/unloaded excruciatingly slowly and in full FIDELITY with PHYSICS at a space station? And the biggest most ridiculous thing about all this faux-immersion physics poo poo is that it's what's primarily holding their progress up. It's the central reason nothing in this game loving works. The ships could have had a decent simulated flight model if they hadn't insisted on having the ships themselves be these maps with their own physics poo poo and all that implies, and the greatest irony of all is that CIG has to constantly fake things in order to not fake other things. It's madness.
|
|
#
?
Jan 2, 2020 21:10
|
|
|
UnknownTarget posted:I gotta ask - why is the physics grid thing so challenging/innovative? In short, edge cases. You can really just apply vector transforms to do the necessary summing of forces, but at some point you'll hit arbitrary precision and limitations of the engines themselves because they're *imperfect* physical simulations. There's quite a lot of variability in the energy transfer between objects, for example, based on how much both objects absorb through deformation, and the elastic and inelastic component of the collisions. It's why crash test dummies usually follow the parametricized testing of the car structures. There's also a lot of handwavium that goes into physics engines to make them *feel* right, and the second you try to actually make things accurate, it turns out no, most Citizens can't run fifty meters.
|
|
#
?
Jan 2, 2020 21:17
|
|
|
Its weird that there's a lot that Chris is so adamant has to be ~fidelity~ and that's no doubt holding the game back in many ways. But at the same time, they have absolutely made some concessions on pure realism, you simply can't simulate all the atoms in a galaxy. So its fun seeing where they're willing to draw the line and where they aren't. And I imagine over time as things begin to look more dire and deadlines keep getting pushed back, the amount of handwavium allowed will increase.
|
|
#
?
Jan 2, 2020 21:20
|
|
|
Keep in mind that SC physics grid are programed by a guy who believes one can "swim" in space. It took actual space goer and fellow game kickstarter scammer, Richard Garriott to explain it to him.
|
|
#
?
Jan 2, 2020 21:20
|
|
|
Hav posted:In short, edge cases. You can really just apply vector transforms to do the necessary summing of forces, but at some point you'll hit arbitrary precision and limitations of the engines themselves because they're *imperfect* physical simulations. Yea, it seems to me like the better option would be to actually create a low-fidelity gravity model, create attractor plates and then kind of go from there. It seems like the volume-based approach is clunky and ends up causing a lot of other issues. I've played Children of a Dead Earth recently and it's definitely possible to do fully simulated gravity equations, though it occasionally slows down a lot. I'm glad to get so many answers to my question. It always seemed like they were basically recreating physics volumes from Unreal Tournament 1 but I wasn't sure if there was something I was missing. You definitely do need some sort of physics simulation per-ship for player entities, weapons, etc. Carrying cargo is unbelievably stupid and if they had had a proper alpha where they tested the idea I'm sure they'd have chopped that early. In areas that I am an expert in, one thing that SC did do well was how they integrated normal-mapped decals. They really have innovated on hard-surface modeling and you can see a lot of great stuff coming from their artists on ArtStation. Basically a normal-mapped decal is a small poly you can place over top of a hard metal surface to get the illusion of a metal indent or extrusion. IIRC Star Citizen was the first or one of the first to use it in a production pipeline. EDIT: Here's some more info: https://polycount.com/discussion/155894/decal-technique-from-star-citizen That being said, the biggest red flag to me has always been how they talked about repairing parts of the ship or whatever. By building out their assets in their entirety first, they definitely would need to roll back and fix stuff at the model end. Since every ship is clearly bespoke in their functionality and programming, this means that not only will the art asset need to be updated but everything will need to be re-tied to that new asset. I can't imagine working with those assets, they must be absolute nightmares to update and change. UnknownTarget fucked around with this message at 21:27 on Jan 2, 2020 |
|
#
?
Jan 2, 2020 21:24
|
|
|
Popete posted:Haven't they been selling really expensive JPEG hauler ships that will carry giant stacks of shipping containers which will of course have to be loaded/unloaded excruciatingly slowly and in full FIDELITY with PHYSICS at a space station? I believe that the selling point was the hauling logistics, which theory-crafted some hilarious things, such as requiring a couple of days to load the biggest hauler - This isn't entirely unusual for some of our terrestrial bulk container ships, but scale is one of the things that you have to consider rather than just making things 'hella big' if you want the fidelities. Scruffpuff posted:And the biggest most ridiculous thing about all this faux-immersion physics poo poo is that it's what's primarily holding their progress up. It's the central reason nothing in this game loving works. The ships could have had a decent simulated flight model if they hadn't insisted on having the ships themselves be these maps with their own physics poo poo and all that implies, and the greatest irony of all is that CIG has to constantly fake things in order to not fake other things. It's madness. They started with the detail. Nobody ever, in the history of the creative works of man, historic and prehistoric started off detailing someone's hand before generally planning out the broad strokes of something. One of the biggest goddamned sins that is on display here is the sheer chutzpah of trying to convince anyone that a company that is 30% ship designers that they've figured out the local market simulation to hand out the boxes, let alone support bulk carriers from day one.
|
|
#
?
Jan 2, 2020 21:29
|
|
|
Hav posted:They started with the detail. Nobody ever, in the history of the creative works of man, historic and prehistoric started off detailing someone's hand before generally planning out the broad strokes of something. One of the biggest goddamned sins that is on display here is the sheer chutzpah of trying to convince anyone that a company that is 30% ship designers that they've figured out the local market simulation to hand out the boxes, let alone support bulk carriers from day one. What bothers me is the "ship designers" they have are operating completely off of the "rule of cool". Literally nothing in their designs points towards even plausible approximations of how a spaceship would look in order to perform its tasks. One really good example? The loving gimballed thrusters on the Hornet. Instead of fixed RCS jets they go with the most complex to calculate/maintain/design/animate solution because they have no loving clue how spaceships work. Thanks Chris.
|
|
#
?
Jan 2, 2020 21:31
|
|
|
UnknownTarget posted:Yea, it seems to me like the better option would be to actually create a low-fidelity gravity model, create attractor plates and then kind of go from there. It seems like the volume-based approach is clunky and ends up causing a lot of other issues. I've played Children of a Dead Earth recently and it's definitely possible to do fully simulated gravity equations, though it occasionally slows down a lot. The real fundamental problem is the chain of events that led Chris to select Cryengine (They wrote the first spaceflight simulation for him), stick with Cryengine, then fall out with CryTek, requiring a sidestep to Lumberyard. They're not actually creating anything, they're utilizing and extending. We're not actually sure if they're up to date on the Lumberyard core because it did appear at one point they were modifying the core engine rather than extending it. UnknownTarget posted:I'm glad to get so many answers to my question. It always seemed like they were basically recreating physics volumes from Unreal Tournament 1 but I wasn't sure if there was something I was missing. No, that feels right. Another one of their cheats is their solar systems are around 1/6th scale, and that has direct knock-on effects for escape velocity for the individual planets, and the calculated gravitational gradient for anything flying around. UnknownTarget posted:In areas that I am an expert in, one thing that SC did do well was how they integrated normal-mapped decals. They really have innovated on hard-surface modeling and you can see a lot of great stuff coming from their artists on ArtStation. Basically a normal-mapped decal is a small poly you can place over top of a hard metal surface to get the illusion of a metal indent or extrusion. IIRC Star Citizen was the first or one of the first to use it in a production pipeline. How does it differ from a bump-mapped decal? I've been seeing those since DX10, although they were a little less subtle. UnknownTarget posted:That being said, the biggest red flag to me has always been how they talked about repairing parts of the ship or whatever. By building out their assets in their entirety first, they definitely would need to roll back and fix stuff at the model end. I'm a network programmer, which is a shorthand way to cover the past twenty-four years of the growth of the internet and my role in it. I have so many red flags about this project that range from project engineering to how they appear to be handling latency for a twitch game in space. I hasten to add that my longevity has nothing to do with competence, obviously, and big pinch of salt. They've already ripped out the docking collars that were on some ship designs, and the form/function of many of the previous sold ships has changed over time; it's fairly obvious that they're simply selling pictures and hoping that sunk cost keeps people coming back.
|
|
#
?
Jan 2, 2020 21:41
|
|
|
Hav posted:How does it differ from a bump-mapped decal? I've been seeing those since DX10, although they were a little less subtle. Normal maps just have more data, bump maps are a flat height gradient where normal maps you use 3 color channels for 3 axis, so you can store more data about angles and make better shapes. That's why everything that's bump mapped looks plasticy and smooth. Normal mapping is just an extension with greater accuracy. But you're right that normal mapped decals probably aren't anything new. I can't say for certain any games that used them but you'd have to have someone on the dev team say so to know for sure.
|
|
#
?
Jan 2, 2020 21:45
|
|
|
Hav posted:How does it differ from a bump-mapped decal? I've been seeing those since DX10, although they were a little less subtle. Zaphod's explanation of normal mapping is right. The innovation with SC is that they used the normal mapped cards to basically blend with the material below the card. This post explains it well: https://polycount.com/discussion/comment/2333156/#Comment_2333156 Here's a tutorial on how it's built in UE4: https://www.youtube.com/watch?v=66IGMnPgEW0
|
|
#
?
Jan 2, 2020 21:50
|
|
|
UnknownTarget posted:What bothers me is the "ship designers" they have are operating completely off of the "rule of cool". Literally nothing in their designs points towards even plausible approximations of how a spaceship would look in order to perform its tasks. One really good example? The loving gimballed thrusters on the Hornet. Instead of fixed RCS jets they go with the most complex to calculate/maintain/design/animate solution because they have no loving clue how spaceships work. Thanks Chris. To be fair, the idea that the thrusters would have some effect was doomed from the start, but it's one of his selling points. he should have come out and said that it wasn't going to be doable because ... well, none of the ships really lend themselves to anything but. That probably goes back to Wing Commander prophecy and the Mary-sue that was the Vampire.  The end pods did a thrust-vectoring thing that looked pretty cool, but was unconnected from the way it flew. Of course that all then goes back to the problem that Wing Commander is entirely based around the carrier supremacy of the Pacific theater - whilst simultaneously managing to miss the point of defensible choke points in every game - which was again partially a way around the whole idea of contiguous volumes, rather than 'hit 'A' for the next area'. They would have been re-imagining the background from some of the most basic 'Fire and Fusion'* questions, like 'FTL comms?', 'FTL travel?', how big is the smallest and largest FTL drive? Enough for a missile? Enough for a planet? What's your limiting factors on that? Power source? Power density? Instead, we have a situation where they now have to balance the needs and requirements of the people that have already bought ingame assets with an engagement rate of around 5%. That's a recipe for a lot of trouble when the vocal minority turns out to create a bad game. I'm reminded of Cliffs of Dover. That released in a barely functional state and was completely abandoned by the developers, getting a decent release when a group of modders first fixed it, then convinced Steam to merge the mod into the main branch. We used to joke that they spent weeks getting details like the right blue for hanger doors, and forgot to fix a bunch of the gameplay issues. Edit: https://bigbossbattle.com/il-2-sturmovik-cliffs-of-dover-blitz-review-a-fresh-start/ Hav fucked around with this message at 22:07 on Jan 2, 2020 |
|
#
?
Jan 2, 2020 22:01
|
|
|
Zaphod42 posted:Normal maps just have more data, bump maps are a flat height gradient where normal maps you use 3 color channels for 3 axis, so you can store more data about angles and make better shapes. Interesting; so an extension of the old alpha and bump mapping by using more layers of the gradient. Haven't done anything like this in anger since my entirely legitimate copy of Lightwave was being used to produce equipment mockups for the sales guys. 1994-ish.
|
|
#
?
Jan 2, 2020 22:05
|
|
|
|
| # ? May 15, 2024 00:46 |
|
|
Hav posted:I'm reminded of Cliffs of Dover. That released in a barely functional state and was completely abandoned by the developers, getting a decent release when a group of modders first fixed it, then convinced Steam to merge the mod into the main branch. We used to joke that they spent weeks getting details like the right blue for hanger doors, and forgot to fix a bunch of the gameplay issues. The pilot's pants texture was some ungodly resolution.
|
|
#
?
Jan 2, 2020 22:13
|
|
