Realism Or Colorful Stars, Which way should i go?

Oh nice, I like it! And what's the name of the song on that video? Love the song, it's relaxing haha

 

Journey Through Oblivion Sound Track

Synapsis - Universal Mind

They are free for commercial use so im using a couple of songs from this track in the game.

 

And it's getting even better!

I ran into a freaking editor glitch today which is slowing my work down though.


I got my Batching LOD system done for stars today and went to build and the editor took a complete dump on me through serialized destroyed errors out of no where.

 

Appreciate it man!

Hey you might want to find out about getting rights for this song

I'm obsessed with it, nice space/space conquest like military song lol.

 

Game looks beautiful -- never mind it being developed by a non-artist.

Besides the fantastic skybox and stars, the ship models look fantastic. You purchased them somewhere, or had them commissioned?

 

I bought them yes.

I am also about to buy 4,000$ in models from a professional artist on the asset store.

After the procedural galaxy, GUI's, intro screen, scene loading system.. a ton of stuff.. yah after all that and right before i start weapons systems ill get a bunch of models and get setup to installing them on the various ship classes.

The ship models will look good and you will probably recognize them but they are what makes this game possible to build besides my sheer will and determination.. and skill as a programmer.

these are the ships that will be in the game:
https://assetstore.unity.com/packages/3d/vehicles/space/spaceships-fleet-i-mega-pack-i-74979

I will also be using MANY of non ship models from him such as jump gates, asteroid stations, stations and other objects.

The domain name GalacticArmada.com cost me 2000$

You can say that i am committed to this project.

I was a web developer (LAMP) before going into game development. I got sick for a few years and wasn't working and fell behind my SEM research, so i decided it was time for a career change.

Luckily im in a position to where i have all the time that i need to build this game, even if that is a few years, working full time, lots of hours daily.

Others may doubt me but I do not doubt myself and that is what matters.

 

So anyway I just tested my Batching LOD system for stars which has 3 LOD modes Low/Med/High

ZERO effect on CPU use works perfectly.

As you travel the galaxy stars will automatically LOD around you, now im working on smooth fadein of flares.

It's just sexy i can't wait to show you guys.

I can literally add another 3000 stars and the lod system will have no impact on CPU time.

I also made 3 different shaders, LOW LOD shaders are simple, medium only use a single low res texture and no flipbook animation.

Highres is the complex shader.

I can make a HUGE galaxy filled with stars. I suppose i will let players choose if they want to play a game that will take months to beat in a massive scale galaxy.

Also I'm not entirely sure massive scale galaxies will work well in multiplayer.

Multiplayer is likely going to be different than single player as far as scale.

I am releasing a single player game first but am designing for multiplayer later on.

 

I was thinking of adding an instant travel system using jump gates (which can be destroyed permanently) in the game and will also have a cooldown, but will be an option to help the player travel faster as well as provide travel option if the players hyperdrive is damaged or if they are in a system that has star that is about to go super nova due to a space titan consuming it and they need to leave the area quickly to escape the super nova physics simulation which will destroy them and their fleet(s) / ships in the area if they do not exit quickly. Hyperdrive jumps may not get you out fast enough depending on your distance from the star when it explodes.

I am going to use mesh and shaders to make a massive scale super nova effect as well as a physics simulation on all ships, planets and objects in space caught within it.

I'm telling you guys, the game I'm building will be like no other that has ever been made in the space sim / 4x RT genre, perhaps the only one made that combines the two.

I'm not doing this to build some run of the mill repeat of some genre. I'm doing this to bring a never before seen experience to gaming in general.

Every space game I have ever played i have wanted, wished and waited for a developer to make something like what i am making and no one ever has.

When i played stellaris i wondered what it would be like to experience it in 3d, when i played x-series from egosoft i wondered what it would be like to have full conquest and tactics system designed for it from the start of development and not some "mod".

I love the game risk and i think i can blend a little bit of stellaris, a little bit of X, a little bit Risk, a tad bit of "freelancer" and a whole lot of FUN.

Unlike *Cough Chris Roberts i have played every space game that exists. I played WAY TOO MUCH.

If Chris was an avid gamer, star citizen wouldn't suck so bad.

One thing Chris taught me though, MMO's suck, sure they might be money making vampire sucking wallet parasites but they are boring.. play "fetch" games.

When not constrained to the box of an MMO you can build a truly fun and interesting game. It probably won't be as profitable but that is OK.

 

Just remember your making a game and not a tech demo. Lots of people fail to actually make a finished game and get caught up in the tech/visuals early on.

If it were me, I'd stop working on sun flares, LOD systems and colors of stars and start prototyping game play. Find the fun asap. If it's not fun with stars and planets as grey spheres, it still won't with pretty colors.

Not doubting your skills here, but this sounds like a massive undertaking. Even for a studio with 300+ employees and a budget close to a million dollars, 2.5 years for a final product would be quite ambitious. I mean, creating the kind of technology tree that exists in these 4x RT games, will all its complexity (multiple factions, balancing), bug testing, play testing, tweaking etc. Would be months of work for a studio, and you'll probably have to do that yourself!

Just saying, find your core game quickly then improve upon it. Otherwise you'll end up like so many others, nothing but a tech demo.

 

As usually so far great results.

Following to @CDF response.
I am not an artist either, but I was doing some 3D modeling in past, not great but enough.

In 3D art and CGI, often approach is to design the general look first. If doesn't feel right, will not sell.
Then doing texturing and shaders and whatever else.

So for example (not mine, googled)

Spoiler

Spoiler

And LOTR Smaug

Spoiler

The point is, to get genera feel for game play first, before focusing too much on visuals. I must agree with @CDF atm, you have good tech demo so far. But no game play yet. And already spending much time on visuals.

Don't get me wrong. I myself did fall in this trap too, since I just love to see what I can do with textures, shaders and whats not.
But I try constrain myself on visuals and focus on the mechanics etc.
However, this not doesn't make your question invalid. Good you ask early for a feedback.

From visual aspect, I would just try to resolve case with size of stars. You can simply use tricks, to push them further away, or for example exponentially change size of stars, base on distance to camera. You should be able to achieve something more realistic in this regards. If you want to of course. But that is my take on it.

Now I would like to see, how you go with basics mechanics and commands issuing. You got path finding etc. But lets say, you want have control of, where you armada or group of selected units should go. Or anything relevant, if that makes sense.

Other than that, keep up

 

Watch me make a miracle happen

I already know that I CAN do it. I have created countless prototypes on my learning journey.

I have already designed the game on paper i have over 50 charts of diagrams. I'm ready to build the game. And the game is built upon the galaxy.

I will get to gameplay soon enough. I know what I'm doing, I design my architecture in a way that is modular and will allow me to adapt ANY kind game that i can imagine.

I am not building a tech demo or a "prototype", what you see IS the game framework being developed not a prototype.

I have already made prototypes to see what was possible and what wasn't. You can see a bunch of them on my youtube channels.

I get what you are saying. And yes it IS massive and would take someone that is.. well probably insane..

To each his own.

Just watch the channel and watch me work and see how quickly I roll stuff out.

Now the GUI and galaxy map will take some time, some things will take more time than others, the GUI definitly will take more time than straight foward programming because im not a graphic artist but i am an expert in photoshop.

So, I can manipulate any template i purchase in anyway that i need too for the GUI. And yes ill be using a professionally made GUI template.

 

There will be two ways to issue orders to your fleet admirals, which will be through a Admiral LIST interface and through the Galaxy Map by clicking on the ICONS for fleets.

The galaxy map will be a 3d minified version of the galaxy that is made to be intractable and ill have a camera on it on a hidden layer.

You will be able to spin it around and manipulate it like a mini galaxy.

I have a plan.

Initially you won't be able to give orders to individual ships in fleets.

The first system will for giving empire orders and tactical orders to admirals. After that i will decide if i want to expand to allow the player to micro manage individual ships in fleets in combat.

You will also be able to drag and drop fleet icons and give orders in a similiar way to how it is done in stellaris.

Really the system is designed for the fleet admirals to command their subordinates. The empire leader will give bonuses to troop morale when present in an area with fleets.

The empire leader will also have a powerful prototype ship.

I will be able to add a tactical system for commanding individual ships later as im designing for it but not adding it initially, since admirals will handle giving orders using my advanced AI system.

 

Cool
Well then, let us know, when you got something working at some point.

 

Anyway,

So you guys understand im working this full time until its a game period.

If its a tech demo ill keep working on it until its a game. 2, 3 years. does not matter. It will be a game.

I have no choice but to make it a game. No choice at all.

So you will see it being made like my life depends upon it, because maybe it does, you don't know.

 

Don't knock cities. There's a story on This American Life about a class of school kids who started giggling at a scene in Schindler's List where a woman was shot in the head. The thing was, they were right. If you've seen someone get shot for real, which some of them had, the scene was so fake-looking that it was funny.

I think it's not so much games, as just TV and movies. The example where everyone has seen so much of that runny red fake blood on crime shows that real blood looks fake.

 

I don't get out much. I'm a dysfunctional genius. (or so I've been told). My IQ scores claim I'm in the top 3%, My personality test shows I'm in the bottom 2% of world population as an INTJ.


INTJ's don't tend to get along with other humans very well because we don't care about your feelings. We think in Logic and systems. Which is why we are called "The Architect" or "Systems Builders".

 

Anyone with a top 3% IQ should know that they Myers-Briggs test is a load of hoopla.

 

IQ tests are rubbish, they prove nothing.

 

Somebody tells me they're smart, I start looking for ways to prove them wrong!

 

LOL. Being smart is not being wise. I'm smart enough to know that I'm most likely not very wise.

I happen to think briggs is pretty accurate.

I'm just kidding. I'm not very intelligent. I have a hard time doing simple math and spelling my name.

I suppose i got bored waiting on the build and had to come put my lack of wisdom on display.

That's how it goes though most people aren't balanced.

Anyway ill be back when I'm ready to post a new video, lest I get my self into trouble with said lack of wisdom.

I've decided I'm not going to do the complex seamless transitions (yet), there good now there is a little popping and i need do a batched lerp on certain details.

It's ok for now. I can move on to doing sunshaft transitions and then ill post a new video. I have also updated the graphics of my stars (again) and now they are REALLY amazing!

I also achieved my goal of having 1,000 with a LOD system that processes them in batches but fast enough where it can keep up with extremely fast movement speed from the tracked object and still LOD stars on time.

And there is almost no performance use noticeable in FPS.

 

Have you met my friends?

That's Tom, Jerry, Will Ferrell, Einstein..............

 

Oh boy transitioning the sun shafts is going to be a bit tricky indeed, doing it efficiently without effecting performance at all

I'm going to pull it off with my magical jedi powers though very soon. Watch 500 ships with a galaxy lod system AND seamless sun shaft transitions as you traverse the galaxy, it will be magical at 200 frames per second.

 

Lol, I never went to sleep guys. I ended up working on it through the night so i basically worked for almost 24 hours straight, literally without doing anything else but eating a few meals.

I wanted to get the Galaxy LOD system done in one day but I ran into an editor bug that stumped me for a few hours.

And the LOD system is not a simple system, It uses a math base algorithm to find and predict position of the player object in order to determine the nearby stars and create a short list to find the nearest star to the player object in order to "not use any cpu time".

The reason is there are 1,000 stars and you simply cannot loop through in an update, so I combined a custom batching system with math based prediction algorithm.

With my system i can add a few thousand more stars if i want and see very little cpu time increase from the LOD and player object galaxy position and near-stars tracking system.

I have also done major graphics updates again and I believe i am settled on my art-style and the graphic quality i want to achieve for now as a minimum quality for the games alpha release eventually.

I also moved the testing area "star system" out into the galaxy away from the "galactic core", so you get a sense of how the stars are spread out in area the player will spend the most time in.

I may improve over time we shall see. I feel like it is pretty good now (leave a comment after you watch the next video i post).

I am uploading it now and im going to hit the hey and it should be done when i wake up. It's a rather large file i recorded it at 60 frames per second to reduce "chop" in the video because the game is running at 200+FPS.

I don't want you to think the game is choppy because it isn't. It's the frame rate difference the video recording can record at 120 frames per second but it would take me 3 days to upload the video to youtube.

All ye who doubt me will soon have all doubt removed, very soon.

My game will be as fun as it is mind blowingly beautiful "for a unity game", "made by a solo developer".

 

I don't think a person really needs to be that "smart" to make a game. Games are a means of communication with other humans (so throw away thoughts like "i'm not a people person". You must be a people person!) The factors that will make or break you as a developer are more general life skills, like organization, critical thinking, maintaining life balance to avoid burnout, staying focused on priorities, knowing when to cut your losses and when to power through tough spots, knowing when to make bold decisions and when to be extra cautious....

The programming -- figuring out the systems and designing them -- I suppose you can't be a total dummy to do that stuff, but that's only one small piece of the puzzle. A solo-developer needs to be eternally humble. Always repeat the mantra "I don't know." Making the game should be seen as an exercise in studying and learning about oneself -- as you are the tool that is doing 100% of the work.

 

Accurate. I always liked the story of the development of Myst. It started off with two brother that wanted to make a game, but either didn't want to, were not able to, or chose not to put time into learning how to code or learn how to create one. This led them to dismay of course, since they had ideas inside of them that they wanted to put out. Then some day one of the brothers found a program, I think it was maybe even for children actually in marketing and in practice, it was kind of like a power-point for story telling. If memory servers it was just a basic ass pseudo clip-art animator like they used to have on DOS, under that name I forgot. Well he saw it and thought, he just may be able to manipulate that into some sort of game. It would be doable since the extremely rudimentary framework was already there, they basically just had to figure out how to fit all their new found pieces together. From this point and click type children's program, they started writing and crafting what would be one of the best selling and most commercially successful titles of all time, across more than one different medium.

So like the Bible, Thomas The Train Engine, Harry Potter, and other commercially successful things it arose from the sheer will of the creator to tell a story instead of their self-perceived intellect. If you have a good idea, you're going to fulfill your destiny and make a good game. And regarding above, I've tested INTJ (sometime ENTJ) 50 + times throughout my whole life, I've never stated it on a forum or in person to statement a fact of intelligence because it's made up, just typing this sentence felt dirty. I believe IQ tests exist only to make people comfortable, anyone who can read and possess within them basic schools of thought. I believe this, mainly in part because anyone growing up with the internet is going to score around 170 anyway, and not all of us are smart. I sure make mistakes every day.

 

HyperCard was the name of the program and it most definitely was not anything like you described. It was a framework for rapid application development with heavy multimedia support. It was in many ways the precursor to Flash. It just lacked a massively popular platform like the Internet to support it.

https://en.wikipedia.org/wiki/HyperCard

 

Same basic story.

"HyperCard is based on the concept of a "stack" of virtual "cards".[7] Cards hold data, just as they would in a Rolodex card-filing device. Each card contains a set of interactive objects, including text fields, check boxes, buttons, and similar common graphical user interface (GUI) elements. Users "browse" the stack by navigating from card to card, using built-in navigation features, a powerful search mechanism, or through user-created scripts."

Edit: Cool read, I didn't know about hypertalk either.

 

Oh you may notice the stars are farther apart, there are a few around the player stars do cluster in some area but they aren't that close together. Some stars are "massive" in size and may appear close but is actually far away if it is a very large star, it will appear closer than it actually is.

I have moved the starting "test area" away from the galactic core out into the galaxy about 1/10th of the way out towards the current edge. Stars get even farther apart the farther you move out into the galaxy.

Go out another 100 million meters are stars will be pretty far apart.

 

For you guys that mentioned getting to "gameplay" as soon as possible well, I need the environment (procedural galaxy) to be done or at least functional before i can get do doing gameplay, combat, team systems (empires).

I don't necessarily need it to be 100% polished but, since im doing it i mise well do the best job i can within a reasonable time frame.

If i attempt to build "gameplay" systems out of the order that i have "designed", it will create a "massive" amount of addition unnecessary "work" and extend development time.

I already know that i can build what i am building. I have enough experience in unity to know this.

The question isn't "will it be fun" the question is more about once i have combat and game mechanics it will be about tweaking and balancing and making improvements to add to the experience in positive ways to create the the fun experience that i envision.

 

I don't think people mean to nay-say you -- just offering words of caution as it seems you have a lot of potential here. People want to see it be the best it can be.

Great work so far. Keep the updates rolling!

 

I can't wait to get planets done It will take a little bit of time, i have create at least a rough draft of the procedural generation and then the position system within star systems,

I'm going to be adding "Areas" too within star systems that will will be nav points later that will contain asteroid fields,large asteroid structures, nebula's, stations, Jump Gates as well other interesting environments including areas that will contain unique physics simulations within star systems.

I want to be able to later on create as much diversity of activities for the player outside of the scope of the conquest and combat systems. I'm looking at it in a modular approach where i can create and add new systems to the procedural generation, unique systems that are self contained.

This could be "missions", finding artifacts, mini-games ect.

 

I know and i appreciate the feedback, positive and negative even if i do get my panties in wad about it sometimes.

I worked really hard to get to the *level to where i can finally build the game.

You know once you really get intimate with Unity on that level you realize it's true power. It is truly a powerful engine.

My only dislike is the quality of directional lighting and lighting in general, i feel like it could be better.

Also the anti-aliasing is... not as good as it should be. But i will deal with these problems and make the most out of what I can do.

 

And i really need to do something with the fly cam so i have more precision navigating the galaxy for my testing purposes. Need to add a constant acceleration so i can traverse large distances quickly and change some speed profiles so i can look around solar systems and adjust my movement scales.

 

Look at Space Engine. Don't get confused with Space Engineers.
http://spaceengine.org/

See how they resolved traveling large distances.
btw. I like theirs software

 

I'll check out. that IS a custom game engine though not Unity.

Ill figure out traveling when i get to doing the hyperdrive travel system, I am also planning on a hyperjump near-instant travel system using jump-gates.

Once i get procedural planets going (it doesn't need to be finished) but it needs to operational. Then ill start on either pathfinding or hyperdrive and galaxy travel systems.

 

Yes, it is custom engine.
It doesn't matter however. Suggestion is, to get the concept, of how fast/slow space traveling works there.
Also, you may pick up other interesting things as well, if you give a chance to explore it a bit.

 

Watching some vids on it, yes i need a control system like that for my fly cam lol.

Also i want his milkyway skybox, its nicer and more high resolution than the free one i have.

 

The scale in that sim is incredible, i tip my hat to the developer.

 

Indeed.
And whats important, is based closely on accurate astronomical data.
There is some hypothetical aspect, like you can fly to planets for example. But you don't know for real, what kind of planets are in other solar systems. So a little fantasy is indeed. However, very well executed.

You want skybox, map it
I don't think he is using any skybox in fact.
Not sure how much trickery is there. But if you switch settings to see more stars, (too many) your PC start shaking

 

Well this is interesting, i just went to 250 BILLION meters and experienced no issues at all.

I did not think that was possible. Past 250 billion infinity errors or thrown.

It happened because im testing my new fly cam and i accelerated to insane speeds and ended up at 250 billion from origin.

This means with some further test i may be able to scale my galaxy size up as well as size of stars and planets.

 

It's less about the distance and more about precision. If you're working with centimeter precision you can go much farther than if you were measuring everything with millimeter accuracy. You have seven digits of accuracy. It doesn't matter how close those seven digits are to the decimal point.

 

Considering float is in range
-3.4 x 10^38 to + 3.4 x 10^38
And you keep ship at 0 origin
You have plenty distance to travel

 

And something about astronomical distances
http://earthsky.org/astronomy-essentials/how-far-is-a-light-year



And
https://www.calculateme.com/astronomy/light-years/to-kilometers/

 

I'm working with meters, i don't want deal with physics scaling.

I am using physics engine for my game

Yah I'm not seeing any precision issues moving my flycam under 250 billion from original origin position.

 

The one issue i did run into which ive worked sort of for now is the ships still running in the test area (when flying around the galaxy at super-speed),
when you float the origin and shift the scene it does a physics update and when you are traveling super fast you are doing alot of shifts really fast and my ships are extrapolating which causes an extrapolation phyics update on shift to sync the tranforms.

I have some ideas for solution. Actually later on what will happen is when the player leaves an area the ships in that area will be translated to the 2d simulation space (background simulation engine) which takes place on the galaxy map or is more specifically going to be represented on the galaxy map.

Ships will be translated from that system to 3d space as the player travels around so this will be a non issue at that point because the ships won't exist in space when the player is traveling between stars.

In my last video i disabled extrapolation and re-enable it when returning to the test area that stopped the physics load spike for the most part when moving at extreme speeds causing many origin updates.

In galactic armada it will be possible for 10,000's of ships to exist (i dont know the exact number yet but im shooting for at least 3000 minimum on an i7) in the galaxy being controlled by empires. However, these ships won't physically exist in space until the player is with visible range of them. Battles that player is not participating in will take place in the background simulation.

That background simulation engine is a topic for another time but lets just say i have a plan and i will design it with load dispersion and reaction techniques that will allow it to run in the background without causing load spikes visible to the player effecting gameplay. It will also heavily use jobs and parallel threading.

 

On a side note i may need some travel time between stars to prep the area the player is traveling too. I may need to slow walk instantiations to prevent any load spiking and do them in batches.

Well see how it goes ill figure out and make it work by god! Nothing stops this train of awesomeness!

 

This is what I have indicated in other thread, if you remember.
Just be careful now.
You are doing RTS supposedly. If you relay on physics, to calculate critical aspect of ships (fighting, hits, relation to other ships) and yet you want move to other solar systems, yet keeping physics going on far away system, or maybe even disabling what is happening on far systems, you may be loosing the concept of RTS. In the end RTS means Real Time Strategy.

From our previous discussion, I expect you want shift at some point in future to ECS. I think you can ditch most of physics in your game. Even now. And calculate stuff based on spatial mapping, or equivalent, plus own simplified physics mechanics, if needed.

 

No, im not keeping physics going in distance star systems.

It is RTS. Stellaris does not have a 3d component. It's a RTS game.

Think about this way. The game is just like any other 2d board game but has the ability to translate into 3d spacial environments.

You could play the game without even ever flying the space ship by controlling fleets and giving orders on the galaxy map.

But no, only ships within 3-600K range will be present in physical 3d space. Any ship outside of that range will be simulated in the background. This is no different than stellaris or a game like master of orion.

Battles that occur in the background simulation will be math based numbers games, just like Stellaris, when fleets are battling and the player isn't there, its a numbers game.

There will be incentives for the player to participate in the battles though.

-----------

What you are "thinking" about is not possible, not even with ECS. It would be to complex, too much CPU time, unless you were literally on a cloud or very expensive machine that could process it.

"a" galaxy full of thousands physical 3d ships with complex AI systems (in real time).. nah not going to happen.

Games like sins of a solar empire have hundreds of ships, not thousands and very, very simple AI systems.

 

Just like in stellaris there will be battle progress indicators on battles the players empire is engaged in as well as battles that are revealed by FOW or scouts.

 

ECS is good for moving massive amounts of objects ... but.. running complex code on 20,000 objects even with threading... I just don't see that working very well.

even with my system having 3K ships, from a programming standpoint it will be more about the "fleet" object than the individual ships in the background simulation engine.

I don't want to be iterating over the individual ships very often in the background sim. instead fleet strength and other data related to the fleet object that is constructed and rebuilt when ships are added and removed from fleets will be used in battle calculations for battles the player is not directly involved in.

The goal is to have a good simulation with maximum performance.

Also, I intend on having multiplayer be my frist "expansion". or one of them.
So I need stay aware of the data (size) that will potentially need to be transmitted over network and keep it as light as possible.

The way i am designing my game, takes this into account.