Concerns with Unity

News & General Discussion
1

OK, so I’ve been at this an entire 3 days now, but I’ve been a programmer for years.

I noticed a few things that have me a tad worried, the number of games that have turned to Vapor Ware designed in Unity that looked “almost done”. Some of these games were very graphically excellent, perhaps the reason for some of this is they were over-engineered aesthetically and going beyond the realistic “easy tuning” capabilities of the engine. I’m guessing the more higher res textures, the more you have to program around the engine with special optimizing code (so to speak).

It still concerns me that you need to watch the performance metrics carefully during design phases because over-engineering could be more costly than anything else (over-engineering for an INDIE leads to making Vapor Ware and that isn’t profitable).

That said, yah I’ve seen a lot of games completed with it, but seems the majority were on the simpler side, but that could just be because INDIE’s have fewer resources to make the more complex bigger games.

I don’t want to argue about which ENGINE is better, since it really probably depends what you are going for, but it looks like even with all these fancy tools, the biggest pain for me will be designing the terrains and all the experimenting with 3rd party tools until I figure out which ones I need.

Hmm…

2

Unity isn’t the all-in-one solution, you need some other software to make the art, and you have to work smart to maintain performance.

Yeah I know this sounds oversimplified, but game design is a complex process if you’re trying to get started. Have patience, watch and observe and absorb and you’ll be fine.

3

Do you have any actual examples of this or are you mostly just postulating this as a basis to ask questions about where people run into problems using Unity with higher end titles? I can’t recall any large games made with Unity that almost made it out but then collapsed under the weight of Unity’s shortcomings. I can however remember a lot of new users trying to make ridiculously over-ambitious RPGs with huge terrains without any considerations towards optimization. These projects obviously never came out.

Could you elaborate a bit on specific cases where good looking Unity games turned into vapor ware?

4

Sorry, I was basing it on googling and anecdotal evidence, not trying to state fact but was looking for input. I can’t remember all the names of the games, was crasher one?

It just seemed like the biggest games with the best graphics were either “in production”, “cancelled”, or a long ways off in general. I’m sure there were some bigger games made with great graphics, I just don’t have time to install them all to look (sometimes screenshots or videos do not do justice). When we are watching Youtube videos of games, it is hard to judge texture quality and graphics, because we are seeing the image smaller and if we FULL screen the image, then it is often not high enough quality to represent anything. I did see what appeared to be a FEW games with smaller maps or simple mechanics that had great AAA quality graphics, but it was RARE.

I greatly appreciate the community here, most of these questions I am asking are related to me knowing better then to get in over my head as one-person without FIRST seeking the advice of others that have been down this path already (humility from being a programmer for so long). I am building a sim game with some space / earth terrain mixed. I chose this game because I know space has fewer objects and that way I can compare “SPACE type performance” to “Land/Terrain performance” and this is a good way to learn a new game.

I am not a game programmer yet, but I’ve been programming for 25 years, so I’m not new at dealing with hundreds of different types of software, including imaging and rendering.

There are some brilliant minds on here and hopefully I can learn what I need BEFORE I make a huge mistake. I’m not one of those arrogant developers that thinks he knows everything, regardless of having been at this so long. I’m seriously just trying to get a feel or direction for what the heck I’m doing before wasting time on fruitless endeavors.

Anyways, thanks for the input so far, I’ll try to take it all in and be perfectly rational about what my expectations are.

5

That’s certainly the main reason. Today’s AAA games are done in a scale none “regular guy” can match. So it comes down to the decision between a polished “demo” with great graphics (your Vapor Ware) or a complete game with simpler assets.
So it has nothing to do with Unity as an engine, it’s a general thing.

6

^That sums it up IMO.

If your project is too overambitious, it’s likely that you won’t get it out of the door, regardless which engine you use.

No matter what you use, there is no magical button that cuts the work for you :slight_smile:

7

I’ve completed few games in unity that are on sale, lovin’ it. Nothing wrong with unity that can’t be worked around. But you need ability of course. And balls. You can’t finish a game without BALLS, son.

8

The magical button is the BUY NOW at the asset store :stuck_out_tongue:

Sorry for my long posts (usually I’m not this wordy)…

Right, I fully understand the scale of large games (some with costs exceeding 100 mil).

However, some of that productivity gained with huge teams is also lost in bureaucracy, red tape, corporate politics, hiring, HR. Then a lot of it is also lost in marketing costs, video production, demos, voice acting, sound, legal services, copyright, trademarks, publicity relations, SEO, web design, etc…

Depending on the company, some large companies actually lose 50% to 80% of their actual productivity due to all the red tape compared to smaller companies, and this applies to a lot of things, not just game design.

However, 20% of 100 mil is still 20 million :slight_smile:

I could go on and on. The key as Indie’s, cut out as much of that as you can, all of it almost. This doesn’t mean you can build a 100mil game, but I would think IF the tools are good enough, you can certainly give the mechanics and graphics a run for the money on a smaller scale.

BUT, are the tools good enough for the GAME I want to make (with semi-high Poly counts for PC only), the answer may still be no, I guess I’ll know in about 30-60 days after playing with terrains and models and seeing the performance hits in this engine.

1) I don’t want to get 80% into a project and find a deal-breaking performance issue that cannot be fixed due to lack of access to the source code, I guess that is why I brought this up.

2) The other question we have to ask ourselves, what is better Quantity or Quality?
Hence, should I make one mini-game alpha-funded that has super graphics, or an alpha-funded mini-game with more content.

I’m not trying to make SkyRim, but a Sim (4x game but with a tad more SIM comedy) with very good graphics. Programming is not an issue for me, but I’m more worried about the animation and game engine. I’ve been a lead architect before, consultant, but just not in this industry. I also do not plan on going this alone, but I have to organize things first and will be alone the first 6 months or so.

I know this engine is great for people building more classic style arcade games and even some basic FPS, but I guess I will have to find out the limitations for myself.

9

You’ll get much better performance using meshes made in a 3d modeling program instead of the built in terrain tools. I make heightmap based meshes in blender and it works great. I texture it in photoshop (really easy if you have a heightmap). I’m a programmer but learned enough to do that.

10

Thanks for the tip, and again sorry for the wordiness or long introduction about myself, I know there are people in here that have some nice accomplishments in the gaming industry (of which I have never been in).

Anyhow, if I can make a business out of this, it would be great. I’ve written probably a hundred thousand SQL stored procedures in my life, and I’m looking to get away from that :slight_smile:

11

The server side highscores on the wiki http://wiki.unity3d.com/index.php/Server_Side_Highscores would get you connected to mysql from unity.

I’m using mobile licenses so had to break down and learn basic modeling to reduce polygons on existing models and make maps.

You can cull meshes not being viewed by the camera so I make them in pieces so you can move around larger maps than it could draw at once, but since it’s only drawing the ones in view the performance stays fine. You can shorten the view distance of the camera, or cull objects by layer to render some only at shorter distance than others. Also spawn points to trigger spawning/despawning as you move around…

I’m on basic so don’t have the pro occlusion culling but its better and might work with the unity terrain but on the free version it draws the entire terrain if you see any piece of it.

12

This is something that is very significant. Not speaking about you specifically, but most people who have programming backgrounds, step into the game industry thinking, “They make games, so the work must be child’s play.” And those people couldn’t be more wrong.

The fact is, game programming can be, by a large margin, the most complex branch of programming there is. I’ve seen more than one enterprise programmer skip into game development and crumble under the pressure.

As to your observation, you should also consider that, because Unity is free and incredibly user-friendly, there are a TON more developers in this engine than most others. Therefore, there are a TON more kiddie programmers playing with it and making wild claims that will never actually produce a final product.

Like any engine, of course Unity has shortcomings, but, IMHO, it is the best, well-rounded engine there is. If this offers any reassurance, check out some of the completed games made with Unity here.

I look forward to seeing what you create. Maybe we can even work together some time. Good luck!

13

Right, I hear you. I’ve always known it was very complicated which is why I waited until the tools became more refined. I know there are limitations of what one person can do (I’m a middle-aged guy, us older folks know limitations, that can be good and bad). I’m looking more to learn enough to manage a project and get people on staff. I’ve got a nest egg I’d rather not FULLY invest into this, but really I’m hoping to prototype something and see what can be done. I’m also a long-time business person, so I’m aware of the pitfalls.

Though I would say that game programming is probably NOT the hardest, though it is hard. The hardest types of programming are generally in R&D type science apps, artificial intelligence trees, and experimental engineering designer simulations and interfaces like Intel uses to break new ground on CPU design. Also the guys that mod some of the hardware simulations and mechanics that go into building some of our hardware, like the video cards themselves. There are experimental fields in Biotech, Engineering, and the sciences that have some absolutely insane type of programs where you might need to consult with 20+ experts just to write anything.

That said, I do agree if you were writing game engines that gets pretty complicated and is probably a close second to some of those types of apps above, but I would put designing most games with pre-built engines at a distant 3rd or 4th to that kind of stuff. That said, it is still very complicated, no doubt at all, it’s in the top 5.

14

you know, if i read this reply a few weeks back, I would agree with you out loud, but disagree in my head. But after doing some studying on algorithm design, NP-complete problems, TSP problems and the like, I can really appreciate what other industries really bring to programming. In my experience, game programming is cool because it borrows from many different disciplines, but in the end, I see that game programming borrows a small subset of all these disciplines…usually a simpler form of bigger well known problems and algorithms.

To get back on topic, if kids can learn things just to make a game, with little or no background in programming then the tool you pick wont hinder you. Yes, unity is not suited for every type of game, and because of the simplicity of use, you will see more people making simple things with it. However, seeing as this is your first game, I don’t think you need to worry about complexity of the game. The programming part wont slow you down its everything else that goes into making a game that will get you. In other words dont worry about other peoples vapor ware lack of source code access…there is a way around most limitations you will find with your first game.

15

Thanks for the feedback. I think we agree if I understood your point (that some fields are just as hard or harder). Yah, I expect the coding to take 30% of the time and animation and graphics 70%, though I could be off, but thereabouts.

To others, I would just point out that we aren’t doing weather modeling with partially unsolved Navier Stokes equations, that stuff is as bad as it gets. Now you’re talking dealing with theoretical estimations, lots of statistical multivariates, and having to work around unsolved equations. Working on that kind of stuff is more specialists in their field that use programming as a tool, so I guess in a way it’s not real hard-core programming, just insane logic and math (so I guess you could argue in that sense it is not even programming).

I know people want to think game programming is the hardest (not trying to start an argument), but although theoretically someone could build AI into a game that came from some of the most intense AI coding there is, it just doesn’t happen. Yes even the AI in gaming can be hard, but it’s nowhere close to research projects at MIT on AI or things like Navier Stokes problems. I really doubt I will hit unsolvable derivatives in game programming and have to pass those back to a theoretical estimation model to clean the imperfect model I am dealing with.

Now I’ve never worked on any of the above stuff, but I’m aware of its existence. That stuff is pretty much incomprehensible formulas, at least I can comprehend most coding (usually). I’m far from a math genius, but I can usually find a way to solve a problem.

We’ll see how it goes…

16

right, game programming isn’t that hard compared to other programming disciplines out there. I think you are being generous saying its top 5 in difficulty. Again, before having my eyes opened on WHERE we got our cool programming algorithms from and a whole host of other algos and data structures i never heard of, I would just think “we’ll game programmers need to know AI techniques, graphics programming, physics programming, sound programming and a host of other things, therefore we are da best!!11!!” But seeing as we don’t need to know these things in depth, and many time we fake things (!!) i no longer think that.

With that said, game programmer tend to be pretty good at programming tho, because of the fact code has to be clean and concise to run the game properly. We dont need to use much Big-oh notation, because we can SEE when things are slow! This means we have to be good optimizers. We have to take advantage of the hardware to its fullest (maybe not with with using unity) to also get some of the results we want. Also, game programmer’s code tends to be very well tested and many game programmers can work well under pressure…I guess i see why i used to think game programmers were the best, lol.

Making games themselves is VERY difficult however. It really is both a science and an art, and again, thats where you will find your stumbling blocks, especially on your first rodeo.

17

over-engineered aesthetically, That sounds like it would be me, but nothing that reduction and optimization wouldn’t fix.

Personally i would be less concerned with optimization, And more clearly concerned with the game’s A.I. and play mechanics since you mentioned “typical sim like”. I don’t wish to Rant On about the Asset Store but there are few jewels in there such as RAIN link Walker Boys link and a few more variations on the theme. How you approach this is up-to previous skill set and figuring out the details on integration into you’re basic SIMs like thing?.

Welcome and Good Luck

18

Thanks for the input. I don’t mean like the Sims series, it’s going to be more like a sandbox of people following a pattern with some random events occurring at time intervals. I am not worried about the game’s AI or problems of logic. These are simple problems that do not require deep analysis by me in most cases.

My fears are texturing design, animations, sounds, etc…

As noted previously, I’ve been a programmer for 25 years (makes me feel old to be honest), I grew up in the hard days of coding though, programming is not my issue, everything else is. The logic was much more complex to do simpler things in the old days, because we essentially had no memory to store any STATE data, and the hard drives back then were so slow that using them as persisted storage was virtually impossible in any near real-time environment. Game logic coding is tedious, but for anyone coding as long as I have been, the programming side is going to be a cake-walk once I get used to Unity. Tedious yes, difficult no.

I considered moving to the CryTek3 engine, because it doesn’t really intimidate me from the coding side, only that there was so much less documentation and the community was smaller and more “secretive” about how to do things (well that’s the way I saw it, I never tried the engine though). Plus upon my very limited research I did on the C3 engine, it appeared I would have had to modify it too much outside of its FPS-style comfort zone. Whereas Unity is more fitting OOTB for the type of game I’m doing.

Now animations, renders, and getting the hang of a new dev environment (called Unity) is what will be hard. I’m set in my ways of using languages and usually having full control, I can see Unity takes some of that away.

Thou both giveth and taketh from me.

19

I looked in the asset store, there is some really cool stuff in there, but not enough items that I need.
I will probably fulfill less than 10% of my needed graphics from that store unfortunately, though I will look at other sites to buy models from as well.

20

if something reveiw this text from English to Russian у меня один синий экран и я не могу менять ракурс и двигаться помогите пожалуйста что делать: Смущен:: Смущен:: Смущен: