• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Dreams Is The Most Important New Game In A Decade

Geki-D

Banned
Assuming the full release of Dreams will come later this year, is it possible Sony will port it to the PS5 as well (launch title)?
tumblr_o7r70he9qK1v3swszo2_500.gif

MM said in an interview that if there was a PS5 (this was before the reveal), they would aim to port it.
 
Last edited:

Vawn

Banned
How many Xbox fans are going to gloat when this game doesn't sell all that well? I'm waiting for the inevitable, "but all you Sony ponies swore this would be the most important game in a decade", when in reality that's just a few vocal Dreams fans, not PlayStation fans.
 

somerset

Member
Pre-hype = trash. Sorry guys, PR marketing methods are well understood. Viral marketing gets the hard-of-thinking worked up for substandard product, because said product cannot sell on its own merits.

It gets worse with no-brain 'creation' tools. In reality they are 'tools' aimed at taking money from the wallets of the truly foolish.

Long, long, long ago a famous computer magazine (not a 'gaming' magazine) had a cover story declaring 'The Last One' was the last coding tool anyone would ever need. This 'program' would allow no-hopers to 'create' every other business program they would need.

Wanna create a game? *Learn to code*. Which also means learn basic maths and logic. Coding languages and environments are *free*. Engines are *free* to work with. Good enough sound, 2d and 3d art tools are *free*. Starter art asset resources are *free*.

What is *not* free and *never* free? Brain power. Thinking. Mental work. Making a game is both 'cheap' (tools and assets) and *very* expensive (time and thought).

Best way to start? Buy/build a PC. Consoles- *forget* about them.
Then install game coding environments with example games and full source code. Do *not* try to code from scratch. Edit a known working source project.

Most of you will be godawful at the beginning. Some of you won't have the brain for coding/design (which only means your brain is built for other things). But some of you will start to get better. You'll start to realsie things you never before understood- and that feeds back into the coding and design.

No-one learns anything useful from the noddy garbage on consoles or phones- even when they *pretend* to have 'coding' by dragging 'code blocks' or the like. Learning to 'code' from doing such is like learning to write novels by doing crossword puzzles. Dumbed down pseudo coding environments are anti-coding environments- they actually make you worse.

A true computer language, even the 'easy' ones, are challenging to master. But master them you must. The cost of entry is a quite basic/cheap PC (thanks AMD). Something with a Ryzen APU and 8 gigs of RAM will do brilliantly. As I said, *all* the software tools (the equivalent in the 1970s would have cost you hundreds of thousands of dollars with far less functionality- since the equivalent back then would have been highly proprietary libraries and tools) are free. And every second you spend coding/designing on this PC will be a second more experience getting you into a real tech career.

Play *games* on your console. *work* (design/code) on your PC. Your brain will thank you. Your future self will thank you.
 

Three

Member
I will be genuinely impressed if I see someone faithfully and accurately recreate the core movement mechanics of Super Mario 64 in this system.

So far though, everything I've seen looks like the same kind of fairly generic third person character locomotion.
Someone is trying to do something like that already



The better discussion to be had is what part you consider that can't be recreated and finding ways around it.
 

Gamernyc78

Banned
How many Xbox fans are going to gloat when this game doesn't sell all that well? I'm waiting for the inevitable, "but all you Sony ponies swore this would be the most important game in a decade", when in reality that's just a few vocal Dreams fans, not PlayStation fans.

No one cares. Their gloating will be hollow. Lbp sold in the multi millions and Sony has Media Molecule as the type of devs to push innovation. This game will score high and be what Sony wants it to be and that is a different type of game that they can point to and say "we sell games, we have fps games, mp games and of course innovative games like Dreams". It looks great in their diverse portfolio. You really think Sony greenlit Tear away and now Dreams because they think it's going to sell a gazillion? You think they don't know based off tearaway?

Bottom line is when you point to those games like Sea of Thieves and Crackdown 3 they neither sold well or scored well that is sad. This game will score well, be innovative like most games under media molecule and sell decent so it'll be a win either way. Totally different scenario.
 
Last edited:

FranXico

Member
Project Spark 2.0: Sony big planet

This is one of the most artificially hyped up games I've ever seen in my life and it literally doesn't do anything mindblowing. The hype is betting on it being a Minecraft level mass market hit despite it having nothing accessible enough for the mass market.

This is like those thinking Console rpg maker would take off when you had less limitations and more control on PC. This is nothing more than what I like to call a "flashing game" where it's hype will give it higher than expected upfront sales but then people realize there's nothing there and then it dies.

I've looking into this game and messed with it myself, I don't see one thing here that will make people go "WOW" with this game, if they want to create worlds and such there are easier, more accessible, and no limit programs on PC.

If this didn't world for Microsoft what makes anyone hyping this game think it's going to work for Sony? Heck, PS had some PC support, what does Dreams due that makes it 500X better than PS? The Sony Logo? Didn't save Socom and LBP..

I'll admit it's getting more "early" word of mouth than PS, but at the same time it's almost all hyperbolic chamber shattering artificial hype that outside very few cases can't possibly be real.

Do you understand how much damage it will do to the game if its hyped up this much and it ends up just adequate like PS? Not even bad, just passable, that would basically kill its interest after the first few days of launch and that's exactly how I'm predicting this will go down, very early hype, with a few days of use then it crashed into 400 windows, 10 cars, and knocks over 5 buildings until it's suffocating under some rocks.

At this point there is zero way this game can deliver, it's tainted beyond repair. Yet you guys keep making it worse. What' more, some company will release some shitty half-assed product once the hype dies to capitalize on the artificial hype and will likely end up being the more successful product. Happens everytime.
The Blu-Rays got to you mate.
 

Shifty

Member
Someone is trying to do something like that already



The better discussion to be had is what part you consider that can't be recreated and finding ways around it.

As best I can tell this is the same basic Dreams puppet locomotion we've seen elsewhere. It looks a bit like the physics-based movement from Gang Beasts and Human: Fall Flat- characters will drunkenly lean during turns, and their limbs interact with the environment based on a global physics simulation.

Compare that to Mario 64, where Mario's direction changes separately from his velocity and his animations are all unaffected by external objects unless it's something explicitly programmed like pushing a block or carrying a box. These are pretty fundamental aspects of how the two implementations work, so you'd either have to create a bunch of logic on top of Dreams' built-in system to suppress undesirable behaviour, or build your own movement and animation from the ground up.

As to whether the latter is feasible, I'm not familiar enough with Dreams to be certain. To give you an idea of the scope involved, SM64 uses the same kind of 'kinematic' movement that you see in most games that aren't goofy physics indies- tightly controlled motion tied to various animations that cause a character to behave in certain ways:

Collision Detection
At the core is a collision detection and response system- the logic to figure out whether two objects intersect, and how to move them out of that intersection in a controllable and believable way.

Dreams obviously has this, but it's not clear how closely-tied it is to the global physics sim, which is a hard-to-control factor that you don't want anything to do with when implementing kinematic movement as all velocities need to be directly controlled by the movement logic.

Character State Machine
On top of that, you have mario's 'state machine'- a programming pattern that defines a bunch of states (standing, walking, running, jumping, punching, butt slamming, damage knockback, etc, etc, etc), how mario behaves in each of those states (not being able to accelerate while punching, not being affected by gravity while climbing a pole, jumping higher on the second and third hops, launching forward during a long jump or dive, etc.) and which states transition to other states under certain conditions (stand > walk > run, punch 1 > punch 2 > kick, jump > butt bounce, and so forth.)

Based on the crazy stuff people did with LBP I think you could probably pull this off with Dreams' equivalent of scripting, but it's the kind of pattern that tends to be much more complex to implement without traditional programming constructs so would probably end up being A. the bulk of the work and B. a nightmare to iterate on and maintain.

Animation
On top of that, you have the animation system. In a traditional implementation like this, animations are tightly coupled with states- if mario is walking, play the walk animation, if he's punching, play the punch animation, etc.

This is probably the simplest part, since you just rig up a "play animation on loop" event to each state change.

Bringing it together
Once you have all that in place, it'd be a matter of tweaking each state's behaviour until it's faithful to the original. Mario does a lot of stuff, so this would probably take a while by virtue of needing to do it by eye.

----------------------------

So all in all, it can probably be done, but almost definitely not with the built-in systems. The logic behind professionally-developed games is complex stuff, which is why game creation kits like Dreams (or The Games Factory, or Multimedia Fusion) attempt to simplify it down into off-the-shelf paradigms that are easy to grasp, but limited as a result. Like I said- I'll be very impressed if someone puts in the crazy amount of work necessary to pull it off :messenger_grinning:

Edit: Gribble's latest post in the OT is probably the most convincing example of a state-based character I've seen yet in Dreams, though it's a whole dimension (and tons of states) simpler than trying to accurately recreate a mario.
 
Last edited:

Three

Member
As best I can tell this is the same basic Dreams puppet locomotion we've seen elsewhere. It looks a bit like the physics-based movement from Gang Beasts and Human: Fall Flat- characters will drunkenly lean during turns, and their limbs interact with the environment based on a global physics simulation.

Compare that to Mario 64, where Mario's direction changes separately from his velocity and his animations are all unaffected by external objects unless it's something explicitly programmed like pushing a block or carrying a box. These are pretty fundamental aspects of how the two implementations work, so you'd either have to create a bunch of logic on top of Dreams' built-in system to suppress undesirable behaviour, or build your own movement and animation from the ground up.

As to whether the latter is feasible, I'm not familiar enough with Dreams to be certain. To give you an idea of the scope involved, SM64 uses the same kind of 'kinematic' movement that you see in most games that aren't goofy physics indies- tightly controlled motion tied to various animations that cause a character to behave in certain ways:

Collision Detection
At the core is a collision detection and response system- the logic to figure out whether two objects intersect, and how to move them out of that intersection in a controllable and believable way.

Dreams obviously has this, but it's not clear how closely-tied it is to the global physics sim, which is a hard-to-control factor that you don't want anything to do with when implementing kinematic movement as all velocities need to be directly controlled by the movement logic.

Character State Machine
On top of that, you have mario's 'state machine'- a programming pattern that defines a bunch of states (standing, walking, running, jumping, punching, butt slamming, damage knockback, etc, etc, etc), how mario behaves in each of those states (not being able to accelerate while punching, not being affected by gravity while climbing a pole, jumping higher on the second and third hops, launching forward during a long jump or dive, etc.) and which states transition to other states under certain conditions (stand > walk > run, punch 1 > punch 2 > kick, jump > butt bounce, and so forth.)

Based on the crazy stuff people did with LBP I think you could probably pull this off with Dreams' equivalent of scripting, but it's the kind of pattern that tends to be much more complex to implement without traditional programming constructs so would probably end up being A. the bulk of the work and B. a nightmare to iterate on and maintain.

Animation
On top of that, you have the animation system. In a traditional implementation like this, animations are tightly coupled with states- if mario is walking, play the walk animation, if he's punching, play the punch animation, etc.

This is probably the simplest part, since you just rig up a "play animation on loop" event to each state change.

Bringing it together
Once you have all that in place, it'd be a matter of tweaking each state's behaviour until it's faithful to the original. Mario does a lot of stuff, so this would probably take a while by virtue of needing to do it by eye.

----------------------------

So all in all, it can probably be done, but almost definitely not with the built-in systems. The logic behind professionally-developed games is complex stuff, which is why game creation kits like Dreams (or The Games Factory, or Multimedia Fusion) attempt to simplify it down into off-the-shelf paradigms that are easy to grasp, but limited as a result. Like I said- I'll be very impressed if someone puts in the crazy amount of work necessary to pull it off :messenger_grinning:

Edit: Gribble's latest post in the OT is probably the most convincing example of a state-based character I've seen yet in Dreams, though it's a whole dimension (and tons of states) simpler than trying to accurately recreate a mario.
Challenge accepted.

The second and third hop jumps were also done by someone.

 
Last edited:
Top Bottom