Faked "real" water.

[Yellonet]

To get somewhat realistic water, couldn't you make the water (in a tank for example) as one big jelly blob? It would of course be much more liquid than jelly so when you make a hole in the tank the jelly-water flows out? Or maybe it would just look stupid :dork:

 

[FoB_Ed]

then you would have to get "real" jelly

 

[M1RAG3]

There is already a thread regarding such discussion. Here.
:thumbs:

 

[[AUS]-TheOne]

wtf?
It's not to do with the substance mate, its do the with the physics... :/

 

[Feath]

Is this a suggestion for real life or Half-Life 2?

If it's for real life, why would you need fake water?
If it's for Half-Life 2, Jelly is still suject to the same fluid physics as water.

 

[Maxey]

Fluid dynamics are still far from being possible to reproduce realistically with the current average hardware.

 

[Yellonet]

wtf?
It's not to do with the substance mate, its do the with the physics... :/

Source should be able to handle jellyblob. The thing with "real" water would be that there would be too many particles.

 

[Yellonet]

Is this a suggestion for real life or Half-Life 2?

If it's for real life, why would you need fake water?
If it's for Half-Life 2, Jelly is still suject to the same fluid physics as water.

Why would you think I would talk about games when I'm on a HL² forum?

 

[Feath]

Source should be able to handle jellyblob. The thing with "real" water would be that there would be too many particles.

Jelly is just like water except it's more viscious. It's still a fluid. And it's still subject to the same physical law, and it still behaves the same, just slower. The problem is the amount of calculation needed for fluids, not the amount of calculation needed for water.

What you are suggesting is along the lines of:

*several years ago*
"Can the Quake II engine support cars with realistic physics"
"No, it can't, the computers of the moments can't handle the physics calculations needed"
"Then, how about you do a motorbike instead?"

 

[Danimal]

then you would have to get "real" jelly

lol, now that is logical thinking

 

[switch]

Water leaks can still be scripted, just dont expect when you shoot a hole in something water to flow out as its not going to happen.

 

[Concept]

Surely you could 'cheat' along those lines though? Shooting a barrel and replace the physics with a non-physics based animation of water pouring out. I think for the immediate future, illusions are stil going to have to be heavily maintained with regards to more complext physics.

 

[Brian Damage]

There's a game out at the moment called "Gish", where you play a blob of tar. It may only be 2D, but the player character acts as a rather well-simulated blob of fluid, and he can even change his consistency to make himself more or less viscous, and raise his friction against surfaces to the point of sticking to them. It's really fun.

http://www.chroniclogic.com/gish.htm

 

[hasan]

There's a game out at the moment called "Gish", where you play a blob of tar. It may only be 2D, but the player character acts as a rather well-simulated blob of fluid, and he can even change his consistency to make himself more or less viscous, and raise his friction against surfaces to the point of sticking to them. It's really fun.

http://www.chroniclogic.com/gish.htm

lol, first time I think of it.
You mean Gish is supposed to simulate a fluid? hmmm, now it makes sense.

 

[affen]

This is a piece from a post in the "Info you've recieved from Valve"-thread:

If I were to create a map in HL2 and build a giant wooden container, containing water and i would shoot out one of the sides, will the water flow out of the container?

Gabe's answer: no

 

[Brian Damage]

We're not debating whether Half-Life 2 has it or not affen, we're debating whether or not it's possible.

 

[Hyperion2010]

I personally think that it might be possable to render something like fluid dynamics by setting a very dense particle system to render water one the outside surface of all the particles, and you could set the detail based on verts or triangles generated around the PS. It would still be VERY computer intensive, but it might* be possible on source.

 

[affen]

IMO it doesn't make sense to fake water with water-like jelly since first you'll have to make a physics model of a fluid. By defining some variables in the model you determine if the fluid is water or something more viscous like jelly (they are both fluids), so you could straight go for water.
I posted that quote because Yellonet was refering to a container filled with fluid.

 

[Yellonet]

Jelly is just like water except it's more viscious. It's still a fluid. And it's still subject to the same physical law, and it still behaves the same, just slower. The problem is the amount of calculation needed for fluids, not the amount of calculation needed for water.

What you are suggesting is along the lines of:

*several years ago*
"Can the Quake II engine support cars with realistic physics"
"No, it can't, the computers of the moments can't handle the physics calculations needed"
"Then, how about you do a motorbike instead?"

No. The difference between water and what I'm proposing is that water is made up of many many particles and my "jelly" is only one deformable object.

 

[Demonmerc]

There will be no real water, period. It is simply too hard to do.

 

[Demonmerc]

Oh and so you know all about the complexity of programming a jelly-like substance eh? My guess is its not much easier to do than water.

 

[Yellonet]

There's a game out at the moment called "Gish", where you play a blob of tar. It may only be 2D, but the player character acts as a rather well-simulated blob of fluid, and he can even change his consistency to make himself more or less viscous, and raise his friction against surfaces to the point of sticking to them. It's really fun.

http://www.chroniclogic.com/gish.htm

Yep. And "Sumo" is a lot of fun

 

[Demonmerc]

Let me rephrase that... While it MAY be easier to program, making it work with the rest of HL2 physics objects is a monumental task.

 

[Yellonet]

This is a piece from a post in the "Info you've recieved from Valve"-thread:

We already knew that. We're not talking about what we have in HL2. We're talking about what could be done.

 

[bodhi]

No. The difference between water and what I'm proposing is that water is made up of many many particles and my "jelly" is only one deformable object.

it AMAZES me that you are still going on about this. you clearly have no idea about what your talking about. it doesnt matter what fluid it is, ("goo" or water), it doesnt matter how viscous it is (
"gooey" or otherwise). the hardware simply cant handle simulating fluids no matter what its made of. so stop pretending that your some genius that has solved the problems of realistically simulating fluids in real-time on anything other than a super-computer, and just accept the you have a jelly-fetish

 

[Incitatus]

then you would have to get "real" jelly

lol perfect answer

:afro:

 

[Yellonet]

it AMAZES me that you are still going on about this. you clearly have no idea about what your talking about. it doesnt matter what fluid it is, ("goo" or water), it doesnt matter how viscous it is (
"gooey" or otherwise). the hardware simply cant handle simulating fluids no matter what its made of.

I never said I wanted jelly. I just used a jelly-blob as an example how the water would be one deformable object. Clearly that was beyond you.

Realistic water is made up of lots of small particles - The hardware of today can't handle that.

Therefore I asked if it were possible to make a water blob that can be deformed - Engines of today can deform objects.

so stop pretending that your some genius

In your company I don't have to pretend.

 

[StardogChampion]

it AMAZES me that you are still going on about this. you clearly have no idea about what your talking about. it doesnt matter what fluid it is, ("goo" or water), it doesnt matter how viscous it is (
"gooey" or otherwise). the hardware simply cant handle simulating fluids no matter what its made of. so stop pretending that your some genius that has solved the problems of realistically simulating fluids in real-time on anything other than a super-computer, and just accept the you have a jelly-fetish

If you play that Gish game that got linked to, you might understand a bit more what he's talking about.

 

[affen]

Can you give an example of a current commercial game engine that deforms a jelly object?

 

[exoeight]

if anyone was ever "pwned" on a forum, that was it.

 

[eth8686]

The third time I post this but whatever:

http://www.nextlimit.com/

Check out realflow. There you go, realistic water.
Check out the water tank video (Yellow thing on the front page - top).

EDIT: Oh, and also check out Gallery --> Videos.
Incredible stuff...

 

[Neutrino]

No. The difference between water and what I'm proposing is that water is made up of many many particles and my "jelly" is only one deformable object.

Actually he was quite right. Your "jelly" is still a fluid and still requires fluid mechanics calculations. It would be no different than trying to create water really, which is of course not possible.

 

[eth8686]

Uhm, neutrino. Did you read my post? It's above yours.

 

[Demonmerc]

Its definitely possible... Just on a 8000ghz box, 5gb ram and 1gb gfx card.

 

[Yellonet]

The third time I post this but whatever:

http://www.nextlimit.com/

Check out realflow. There you go, realistic water.
Check out the water tank video (Yellow thing on the front page - top).

EDIT: Oh, and also check out Gallery --> Videos.
Incredible stuff...

It seems that you need some 3D-editing software to try the demo of realflow, correct?

 

[Demonmerc]

Its not that liquid is hard to simulate, IIRC wave race 64 used water dynamics, it is HOW you simulate it.

It's a simulation- it doesn't HAVE to act like water if programmed that way. The programmer will take shortcuts WHEREVER POSSIBLE. That means that if the game/movie/whatever doesn't have to drop things in the water, spray it it tiny bits, or spin it around it a funnel, it WILL NOT have that capability. It's not like once you model water its done and you can do whatever you want with it, the programmers must decide how it interacts with the world on every level.

BTW my previous post takes into account the rest of HL2 including water dynamics.

 

[Neutrino]

Uhm, neutrino. Did you read my post? It's above yours.

Yes I did. What about it?

If your referring to my comment about creating water not being possible, I was of course referring to that in the context of HL2, by which I meant using the current version of the source engine it's not possible to create realistic water dynamics on todays machines.

 

[Incitatus]

The third time I post this but whatever:

http://www.nextlimit.com/

Check out realflow. There you go, realistic water.
Check out the water tank video (Yellow thing on the front page - top).

EDIT: Oh, and also check out Gallery --> Videos.
Incredible stuff...

realflow and realwave aren't realtime

they still require a calculation pass

 

[Yellonet]

realflow and realwave aren't realtime

they still require a calculation pass

Are you saying that anything realtime doesn't need to be calculated?
But I guess you're right about Realflow and realwave not being realtime.

 

[cylyk]

worst idea ever