Developer Discussions: Alpha 17

[sillls]

That's exactly the problem.

There is no good way to look at one or a few blocks and change only those.

If you dig out a huge hole, like one of those open air bedrock bases and connect it to an infinite supply body of water, what happens?

 

Does water flow in there, which "Terraria water" would do? (which is one helluvalot cheaper in 2D, BTW, and Terraria has tiny bodies of water in comparison)

 

Does the water that flows inside become infinite supply water as well? At which point does it do that?

Because if it doesn't it remains "flowing" water and keeps costing CPU time.

 

Either way, water "flowing into" a huge hole is probably the worst case scenario here, akin to having a skyscraper or other large structure collapse.

There simply is no cheap way to do it and here you have to take into account that chunks can get unloaded anytime so "water filling a hole" would be suspended while no one is watching or only one half of a giant hole gets filled resulting in a huge vertical wall of water or such craziness.

 

Sure, it may "get better". Eventually.

But I'm certain that it will take a lot of brain sweat and work and not someone's bright idea for a quick fix. =P

 

Ok. So when it's flowing then it can be infinite water supply but only until the hole fills up. Then it stops flowing.

There would need to be a (water line) in which water does not flow over. Otherwise the water floods the world.

Any water above the water line would can be infinite but would not flow. So the player can (take) water from it but it wont go down.

 

Since all water seems to be at the same height in the game (not counting wells), that can be the water line. As for all the wells that dot the RG map, those could be preset to a specific were they are infinite but does not flow. So the player can take water from it and never drain but will never flood the world.

 

Osiris New Dawn has this feature when your digging up sand. You could dig all day and it won't change the shape of the ground.

 

So basically all water could be set to infinite but only flow if below the waterline.

 

OR you could just set water to infinite and not flow (or flood).

[uowaep]

That's exactly the problem.

There is no good way to look at one or a few blocks and change only those.

If you dig out a huge hole, like one of those open air bedrock bases and connect it to an infinite supply body of water, what happens?

 

Does water flow in there, which "Terraria water" would do? (which is one helluvalot cheaper in 2D, BTW, and Terraria has tiny bodies of water in comparison)

 

Does the water that flows inside become infinite supply water as well? At which point does it do that?

Because if it doesn't it remains "flowing" water and keeps costing CPU time.

 

Either way, water "flowing into" a huge hole is probably the worst case scenario here, akin to having a skyscraper or other large structure collapse.

There simply is no cheap way to do it and here you have to take into account that chunks can get unloaded anytime so "water filling a hole" would be suspended while no one is watching or only one half of a giant hole gets filled resulting in a huge vertical wall of water or such craziness.

 

Sure, it may "get better". Eventually.

But I'm certain that it will take a lot of brain sweat and work and not someone's bright idea for a quick fix. =P

 

One way to handle big operations like this is to limit the game to how many iterations it can make per frame toward this calculation. The result may be very slow moving water, but at least it would be able to calculate correctly without stealing overhead from the rest of the game. The issue with unloading chunks in large cases could be handled by not unloading chunks with water until flowing has stopped, and if flowing into an unloaded area, load that chunk temporarily, or at least a small portion of it. You don't really need to keep everything in the chunks loaded, only the parts relative to water flow.

 

- - - Updated - - -

 

IMO the only way this is happening is if Unity adds DIRECT water mechanics support for voxels to their engine. :-|

 

We all know that calculations made by the underlying engine of a game are waaay faster than those made by the game itself.

 

That's not entirely true. There is some truth, but it's not as huge of an impact as you think.

[Anabella]

On RWG in a17:

Random worlds in a17 onward will be fully "pregenerated". This means that the world generator will output the same data structure as navezgane but in a save folder. This means that any rwg world will load just as fast as navezgane going forward.

 

After fixing the vehicles and adding new ones it was determined that RWG's runtime code was just waaaay too slow. So now it instead gets all the data set up and exports it in a common format which allows for quicker chunk loading.

 

With the change to pregeneration and using the socket(tile) system over the entire world there will be some other changes as well. For the better. Over the next couple of days I will give a bit more detail on what else these changes to RWG mean for the average player, modders, and server owners. I think most will be very happy.

 

More to come...

 

That is awesome Kinyajuu. Looking forwarding to hearing all about the goodness coming from your work on this. And I am hoping that might mean we can get those lovely cave systems back that RWG use to have or something close. I am sure roads are going to get a ton of RWG love. Thanks so much for all the work and time.

 

- - - Updated - - -

 

Yep.

 

On the road again. Just can't wait to get back on the road again. lol

 

I was thinking it was more like "On a long and dusty highway".

[Anabella]

This is getting awkward...

 

hi Dagzambie <3.

That better?

 

- - - Updated - - -

 

Eep! *Runs to safe place!

 

LOL you are reminding me of the roadrunner mob in a game I am playing. It makes all this noise like it wants to be found then when you find it it goes running off all scared.

[Aldranon]

Is this Moderator Aldranon putting a forum user in his place for posting a Doom and Gloom message?

 

LOL

 

I've turned into Roland? Well there are worse things to turn into, I would imagine. Stale pizza comes to mind.

 

Oh yes, if I suddenly stop posting... Roland killed me, you will find my internet corpse where he buried the Raptor.

[Red Eagle LXIX]

Unless you get the water ski upgrade for it.

No, but there is ARRRRR which governs seamanship.

 

Water Ski Vehicle Mods confirmed for A17. Go-Go. Who needs boats when you can have a Jetski Jeep that goes true all terrain. j/k

 

And I'm caught up with the thread again.

 

Faster RWG, cool beans.

 

Water, I'm with others, even infinite filling water would be better than the water sinkholes until a solution could be done (I'd probably mind less if they were at least silent and did not have a drag effect beyond normal placid water instead of the effect they do have.).

[Jost Amman]

We do? That's... interesting news.

 

Not all of unity's stuff is lightning quick and great. Their terrain system hasn't been updated in far too long and there are several other alternatives which are far faster. One of which being the terrain system in 7dtd which is custom.

 

Wait until the 2018 update hits and everyone goes mental because their entire project has turned pink. The new shader system they've introduced is incompatible with the shader system people have been using for years now. The only solution is to rebuild all of your shaders in the new system using a pretty little graph layout which should annoy the most hardened shader coders.

I'm not talking about generic stuff, I'm talking about how water works now in 7DTD.

As far as I know 7DTD water is already a custom coded part of the game... and look at the results.

 

However if Unity does add voxel water physics to the engine, assuming they care to do that in an optimized and efficient way, it would be a win-win. Unity will have a nice new feature to offer to their clients while TFP will have a native voxel mechanics they will be able to use in game that will be updated and improved by Unity devs on each new release. :-]

 

I understand what you're saying is right on a more general level but I think that regarding the water physics we would be better off with the current custom-made code.

[Jost Amman]

You probably meant to say that calculations made by the GPU in parallel are waaay faster than those made by a single CPU.

 

Games programmers can use the GPU as well (if they don't mind the additional coding effort) and they could generate code that is very specific to their problem while an engine usually has a one-size-fits-them-all solution for any problem.

Well, common sense would suggest that a game engine company will try to code this kind of new mechanics in the most efficient and performing way they can, unless they want to damage their sales or worsen their reputation (even more). :-]

[The Gronk]

Well, common sense would suggest that a game engine company will try to code this kind of new mechanics in the most efficient and performing way they can, unless they want to damage their sales or worsen their reputation (even more). :-]

 

You're forgetting the prime motivator... money.

 

Unity's terrain system was bad when they first implemented it and it's been largely ignored for years now. This swiftly led to the development of a lot of free and paid alternatives, most of which were better than unity's ham-fisted attempt. Unity now makes a lot of money from these extensions without having to lift a finger, free dlc revenue for them.

 

What's more likely to happen is that unity will buy an existing terrain system and incorporate it into the main engine, in the same way that the "pro-builder" extension will be available in the new package manager for the 2018 build. It would save them a lot of work and they'll have an already popular system added to the engine.

[Roland]

Is this Moderator Aldranon putting a forum user in his place for posting a Doom and Gloom message?

 

lol...Doom and Gloom? Never. There is nothing more hopeful and exciting than having the option to mod something in or out of a game you don’t like. I give TFP full marks for giving us that ability.

 

I get snarky about the HP bar because I’ll never play a serious game using it. Either they will remove it once their balance testing is done because it is completely contrary to the survival aspect of the game or they’ll leave it in as an ode to classic RPG nonsense in which case I’ll wait till the HP bar removal mod is incorporated into Sphereii’s amazing and spectacular mod launcher.

 

Heck, for my serious play through of A17, I’ll be removing the entire HUD anyway.* It is the pinnacle of immersive and cinematic gameplay for me.

 

Whatever you decide to do TFP I appreciate you for allowing us to customize the experience!

 

 

 

 

*What I really mean is that I’ll be hoping the invisible HUD mod gets updated for A17 because that kind of modding is way beyond my meager skills.

[sillls]

I was thinking it was more like "On a long and dusty highway".

snip...

 

Or that really twisted roadway

Or that mismatch road with the connection twenty feet below.

Or the low road with the buildings right above you...

[Ninja_Freak]

This is my real account,i post only when its need or when i feel i need to do...1year its too much even for engine change or what they change...you dont need to agree with me..its my personal opinion.The fact i m still here shows i still interested by the game ,what i want to point is this methode with HP bar...more fair was to tell us a apx date(for ex : september-november)....

 

Totally agree... I think V16 is broken in many ways and like lordcrolwer, just because I don't sit here and post time after time doesn't mean I don't read the forums or like the game any longer.

 

It's taking a very long time IMHO that's why I mentioned it.

[Tin]

snip...

 

Or that really twisted roadway

Or that mismatch road with the connection twenty feet below.

Or the low road with the buildings right above you...

 

Massive Tectonic plate shifts and Aftershocks?

[Roland]

Totally agree... I think V16 is broken in many ways and like lordcrolwer, just because I don't sit here and post time after time doesn't mean I don't read the forums or like the game any longer.

 

It's taking a very long time IMHO that's why I mentioned it.

 

What’s the deal with all the impatient people saying 1 year dev cycle is too long? Whether it is or isn’t it’s only been about 7 months since the last update and since the team turned their complete focus to A17 development. I respect opinions that 1 year is too long but can we please either say that 7 months is too long or wait another 5 months to start opining about 1 year?

 

I know some of you don’t like to count the decimal point updates as “real” updates but....they are.

[Aurelius]

What’s the deal with all the impatient people saying 1 year dev cycle is too long? Whether it is or isn’t it’s only been about 7 months since the last update and since the team turned their complete focus to A17 development. I respect opinions that 1 year is too long but can we please either say that 7 months is too long or wait another 5 months to start opining about 1 year?

 

I know some of you don’t like to count the decimal point updates as “real” updates but....they are.

 

Doesn't 7 months round up to a year?

 

-A

[Jost Amman]

You're forgetting the prime motivator... money.

 

Unity's terrain system was bad when they first implemented it and it's been largely ignored for years now. This swiftly led to the development of a lot of free and paid alternatives, most of which were better than unity's ham-fisted attempt. Unity now makes a lot of money from these extensions without having to lift a finger, free dlc revenue for them.

 

What's more likely to happen is that unity will buy an existing terrain system and incorporate it into the main engine, in the same way that the "pro-builder" extension will be available in the new package manager for the 2018 build. It would save them a lot of work and they'll have an already popular system added to the engine.

If this turns out like you say and we get better voxel water because of it I'd be happy. An officially supported good extension would be on-par with the kind of quality and performance I was talking about before. If TFP want to buy/use something like that it would work like a charm IMO and benefit the game in many ways.

[Dagzambie]

hi Dagzambie <3.

That better?

 

- - - Updated - - -

 

 

 

LOL you are reminding me of the roadrunner mob in a game I am playing. It makes all this noise like it wants to be found then when you find it it goes running off all scared.

 

If you saw what I saw peeking into the window, you'd run too!

 

(Edit) We should have roadrunners in this game! They fit in the region and you could catch them, cut their heads off and drink the blood for a boost in running speed!

[DaVegaNL]

What’s the deal with all the impatient people saying 1 year dev cycle is too long? Whether it is or isn’t it’s only been about 7 months since the last update and since the team turned their complete focus to A17 development. I respect opinions that 1 year is too long but can we please either say that 7 months is too long or wait another 5 months to start opining about 1 year?

 

I know some of you don’t like to count the decimal point updates as “real” updates but....they are.

 

I heard my name?

 

The last decimal point patch was 7 months ago. The last full Alpha update was close to 11 months ago. That means that 4 patches came out in ~4 months. 1 patch/month. So they are 6 months overdue! If 7D2D was a woman, it would probably have to start buying diapers.

 

#indisputablemaths!

[Brett]

If anybody here was around at Alpha 1.1, they'll understand that extended patch updates is absolutely nothing to worry about.

[beerfly]

If anybody here was around at Alpha 1.1, they'll understand that extended patch updates is absolutely nothing to worry about.

 

Absolutely. Since the beginning of playing this game many of us knew the road they will travel on and we are all ready to wait as much as it is needed. Yeah, its how stuff works on this planet. It gets better, it will get much better.

[Crater Creator]

You could also stick a wound model to said model like they do in hollywood.

 

Okay, good point. I’ve never seen a game do it that way, but hey, new ideas are part of what forums are for.

[faatal]

IMO the only way this is happening is if Unity adds DIRECT water mechanics support for voxels to their engine. :-|

 

We all know that calculations made by the underlying engine of a game are waaay faster than those made by the game itself.

 

Unity has zero voxel support and I doubt they would ever add it, since it is very game specific, but they did add 2d tile support, so you never know.

 

Based on my experiments, native c++ code, like the Unity engine, is about 3 times faster than compiled c# code. The c# compiler (Microsoft) could be improved to generate better code and then there would not be much speed difference.

[Jost Amman]

Unity has zero voxel support and I doubt they would ever add it, since it is very game specific, but they did add 2d tile support, so you never know.

 

Based on my experiments, native c++ code, like the Unity engine, is about 3 times faster than compiled c# code. The c# compiler (Microsoft) could be improved to generate better code and then there would not be much speed difference.

It's a shame to hear Unity doesn't support voxels! :-|

 

So I guess it's all in the hands of Microsoft (*frawns*) or we need someone to code some very good (and fast) extension to support voxels and maybe specifically to support some kind of voxel water physics that could be "easily" integrated into the game.

[faatal]

A generic voxel implementation would probably run several times slower than a custom solution, so even if it was native c++, it would probably be around the same speed as our c# custom implementation.

 

It is in our hands to fix it and when higher priority issues are done, I'm sure we will deal with it.

[The Gronk]

It's a shame to hear Unity doesn't support voxels! :-|

 

So I guess it's all in the hands of Microsoft (*frawns*) or we need someone to code some very good (and fast) extension to support voxels and maybe specifically to support some kind of voxel water physics that could be "easily" integrated into the game.

 

You could always write your own implementation, as the Pimps are currently doing :-)

 

- - - Updated - - -

 

Oooh! The health bar dropped by... <takes off socks> Twelvtey!