Does Unity make game dev too difficult?

I have thought several times over the years about Unity and wondered if perhaps it makes game dev more difficult than it should be.

Part of this is based on my own experience working with it although admittedly that was almost certainly due to not "getting it" and attempting to throw all Unity specific stuff out and use as if it was just a more straightforward api to load, play and render content.

Part of it is due to seeing many people around here working in Unity for years (and they seem to possess plenty of skill too) and yet they have completed no games (not even a tiny super simple game) as far as I know.

Yet I have also seen newbies come in and make their first (usually Mobile) game in less than a month.

I do think the whole "getting it" thing plays a major factor here. If a person really "connects" with Unity they will likely complete a game fairly quickly. If they don't truly connect with it well it seems likely they may never complete a game with it.

What is your view? Let's say you can use up to $100 worth of stuff from the asset store. Would you be able to complete a FPS with the scope of something like the old Wolfenstein 3D? If yes how long do you think it would take you? Again the key here is could you COMPLETE the game... not START the game or create a scene, etc.

 

go make a game in c++ deal with the windows lib API and then come back and answer your own post.

 

Modern gamer expectations and the impact that has on the developer makes game development too hard. Players have come to expect certain things from certain genres and now the developer has to bring those to the table in addition to any of the features they wanted to introduce to the game.

That said I don't believe Unity is blameless. It's making life more difficult but only because some of the functionality it has is seriously subpar and requires the developer to build their own solutions. Two quick examples would be the Input API and the terrain system. We've been promised replacements but many people need them now.

 

Completing projects doesn't have as much to do with the IDE as it does the Developer's motivation.

 

Unity makes game dev TO easy...

 

Thanks for sharing! Let me ask you... you have been playing around with Unity for quiite a long time. Can you complete a clone of Wolfenstein 3D? And if so how long do you think it would take you (again you can use up to $100 worth of store assets)?

 

Game dev is hard by nature. That's nothing to do with Unity.

 

How much of a clone would it have to be? Would you expect it to have the same overall feel? Because I'm almost entirely just a programmer. I have very little artistic knowledge and the same applies to level designing. You'd be getting a prototype very similar to the one @Arowx made a while back.

 

I could complete something with the scope of the old Wolf 3D without spending a dime. I figure that a complete clone would take about a month considering I'd have to make sixty levels more than anything. I imagine I could hack together some procgen that could make the levels for me and save some time but that'd be going a bit beyond the project's scope.

 

That $100 was with that in mind. It doesn't have to look exact or anything like that. It just has to be the game itself. The play experience. The theme, etc. If you have to get aliens instead of Nazis so be it. And by complete I mean the entire game.

Reason I chose this game is because this is one of the oldest FPS games so it would seem like if people have any hope of making any semblance of a more modern FPS they should be able to make W3D. Maybe that logic is flawed but is how I am viewing it.

 

A month or two for the core. Probably six months to release.

The 2D-3D hybrid would probably be the most technically challanging, as it's an archaic technique not directly supported by Unity.

And I personally wouldn't bother with the asset store for this one. The time taken to find, evaluate, intergrating and tweak any asset would probably be longer then the time taken to build the specific system.

 

I like that approach. I find the non visual dev appoach and being one with the code very simple. Not building a 3D engine. That would be silly in this age. Of course, there are pros and cons to everything. But mainly it seems to be a commonly held belief that Unity makes game dev very easy. If so it seems like every person who has messed with Unity for many months should be able to make a game along the lines of one created 25 years ago.

 

You don't need to do the pseudo 3d stuff. Can be entirely 3D low poly etc but with "real" art and animated.

 

Nah, the levels are simple tile maps, and since they're easily found on the internet, recreating them would take probably 10 minutes or so for each level. (I'd use my own SpriteTile for this and modify the code to produce 3D levels instead of 2D.) I'm going to say a couple of weeks for the whole thing.

--Eric

 

I suppose it'll take me a month or two to create something like that for me, and I believe it could be a testimonial to the conveniency of the platform, or the usefulness of Asset Store, since I only started writing games early this year.

If I don't have to create too many levels then probably I can just buy some good FPS template with something like Rewire, and maybe some other items which might make doing the UI and decals easier, then just write quick dirty scripts to glue them all together.

I think a seasoned Unity developer would probably need only couple of weeks rather than months, but I'm still learning the platform, so I expect to do more trial and errors.

For me, it's been more a problem of getting used to the way it's supposed to do in Unity, as many of such practices seem to be at odds with my instincts as a programmer, but I have no doubts that they help immensely in making game development so much easier, especially for those who don't have much experience in programming.

Probably, the recent surge of indie developers who have been churning out 10+ new games every month on Steam might be a sufficient proof for that.

 

Oh, I know. I was just operating under the assumption that since this game is a clone and not a direct reproduction that I'd have to make original levels.

 

Seasoned Unity developers have frameworks they've written over the years too. Being able to start from a common framework can save a great deal of time over someone who has to write theirs from scratch.

 

Well, that's a new one.

Unity3d basically shaves off a few years from development time. Basically, without the engine, a user will typically struggle several years with C++ programming, and maybe write a demo that will be missing 99% of ease of use provided by unity editor.

Chances are they'll learn how to make engines, without releasing a single name in process.

I think I mentioned this before... but with unity it is possible to make a 3d action game prototype in less than a day. Without unity it ain't happening (at least I won't believe it till I see it).

Youtube link appears to be either dead or private.

I think the whole wolfenstein can be recreated by a lone developer in less than three months if someone funding development. That's without ever buying anything from asset store.

 

That approach is suicide for anything serious, IMO. Or a "not invented here" syndrome.

I spent tremendous amount of time dealing with C++ and various internal engine aspects, by now I wish I spent that time doing something else.

Low poly 3d instead of sprites will increase development time. Because skeletal animation.

---
Either way a lot of people (including me) would have no problem recreating the whole game from scratch in unity without buying anything on asset store ... as long as someone is willing to pay for development time.

 

If anything I think from a big-picture industry-perspective Unity has made it too easy, but in all the wrong areas because making good games is still hard as hell. It has never been easier to get terrible games out into the wild, that's not good for the market imho.

I don't think I could ever find the motivation to do a super tiny simple game, because I couldn't get excited about it and wouldn't have the motivation to work on it at all. I'd rather experiment with lighting setups or build a complex terrain shader from scratch in Shaderforge for the third time.

I might never finish my own game because I'm terrible at managing scope, but just a few weeks ago I took a weekend to read up on multi-threading basics and put a part of my pathfinding into a separate thread and that gave a huge performance boost. This was a really cool experience because it was the first time I ever coded something threaded and it worked like a charm. It'll likely solve a lot of problems I had with my pathfinding method, because I now need to worry a lot less about how long my calculations take. I wouldn't trade that for having finished a super simple game instead, because I wouldn't be able to be proud of a super simple game.

 

Regarding the opening statements, I would say it depend if you against or with the flow of unity. All my complaints is because what I wanted goes against the flow ...

 

That probably adds to development time, rather then subtracts.

And as soon as you go full 3D, you've definetly created something that's not wolfienstien.

 

I think the main issue is simply time.

Tell me one creative endeavour that takes 6 months or a year that a significant number of people will succeed in doing to the end without losing motivation?

Tell me one business endeavour that people will do for that time without making any profit in the meantime (and probably losing quite a bit)?

Tell me anything that a lot of people will do consistently for that amount of time, where the reward comes mainly at the end?

When you think about it, it seems like an insane thing to even try to do. The only reason I'm still going is because I reassure myself that no matter how many mistakes I make, it's still helping me arrive at a more efficient way to do what I want to do.

 

Getting pregnant? I know it usually takes couple more month than that to release the end product, but I guess it could be because there's no engine or platform of any sorts in that field

 

Unity actually takes you by the hand and helps you cross the street.

I remember back in college I took a course on gaming and simulation, in a nutshell, it was an OpenGL course which uses C++. I loved the content we were learning, but it was a drag and frustrating at times, I thought it was cool to be able to develop a game from the ground up, but good luck making anything worthy of releasing, not saying it's impossible, just good luck, that's all.

Perhaps there are engines out there that are more suitable for your needs. I personally feel that Unity is the best in terms of balance, you get a bit of everything.

 

First game I worked on was with a large team building their own custom engine. The ultimately failed because of the cost, time and difficulty...took 10 years, but it did fail.

First game our new team worked on was with Torque3d. It was very difficult to make the game we wanted. Assets were especially difficult to find or even make because of the formats of assets used in the engine.

Unity...feels so easy comparatively. And yes, the assets and tools available on the store do cost and you really need some of them to make a decent game, especially terrain tools, shaders, etc., but compared to a custom engine and T3d, it is not impossible.

In our experience, Unity made it easier to make our game.

 

Only if you implement it as a hybrid solution. A raycaster works by shooting a ray for every column of pixels on the screen until they hit a wall and then drawing a textured line sized according to the distance of the wall. For my solution everything would be implemented as textured quads with an orthographic camera.

 

I actually did a huge chunk of this only a couple of weeks ago. Just two maps, one weapon, and no enemy yet, but it also didn't take long at all, maybe a couple of hours.

I estimate that if I were to complete a full equivalent game it'd take two or three weeks, and I think most of that would be working on maps and perhaps enemy AI.

 

Thanks for letting me know. I had no idea it played fine for me. I just swapped another YT in for it.

 

Thanks for all of the participation especially those of you that answered the question about can you make such a game in Unity and included an estimate of time.

I should probably have clarified things more. I am not looking for switching to something else. And not even saying some other way is better or faster than Unity. I do like the code oriented approach for the most part but I already figured out how to do that in Unity.

Mainly I was trying to assess the general value of Unity for many people here who are forum regulars.

I do find it interesting the range of estimates from 2 weeks to 6 months. Makes sense though based on overall experience, familiarity with Unity itself and closely related to that the whole how much a person is connecting with Unity as a dev tool. Overall, I think 8 weeks for something like this would be the max time I'd hope it would take. And hopefully one day that will be true for even the person who has only been using Unity for 1 month.

On the other hand, I was surprised that nobody said "I'd just grab so-and-so asset" and have the game basically completed in 5 days". Surprised only because with all of the stuff on the Unity store I would think there would almost be like some kind of kit available to make such a game these days.

The very odd thing is I remember seeing games like this created literally in a matter of days. I believe some have been made for LD. And I know things like NaaLaa6 come with built-in old school ray casting style 2D FPS support. I think some of the LD examples were made in Unity but I might be mistaken.

Seems like surely there is a very easy & fast way to get the bulk of such a game up n running in Unity as well... ? Thinking maybe there is an asset or two that would be kind of plug n play. If not... why?

 

If you don't care about content I'll make it in two days, and the game is simple enough that I'd rather write my own code than familiarize with someone else's. But if you're aware of the fact that Wolfenstein has six episodes of ten levels each... that's where most of the time is going. At an average of 60 minutes per level - which is really throwing them together - that's 60 hours right there, which is three quarters of my two week estimate.

Edit: Correction, three of those episodes are actually an expansion. I'd still leave the broad estimate at two weeks though, and give myself a lofty two hours to throw a level together.

 

This has more to do with your personal work ethic, motivation, dedication, etc. Most people want to make games but when they realize that their "simple" indie game will take 2 years to make run out of steam. It's very common for hobbyists and indies.

But you're so wrong on this. So very, very wrong. Unity is by far the easiest way I know of to make games that doesn't also cripple you as soon as you want to expand. There are technically easier ways like Construct or Game Maker, but I found they kneecap you, you're more or less stuck with what they give you in terms of capabilities. Unity is both very, very easy for beginners to get started and flexible/powerful enough to be used for "real" games.

Starting for what I know now, and assuming I had all the assets I needed, I could bang out something with the complexity of Wolf3D in maybe weeks. Wolf3D is such a simple game that most gamers these days wouldn't even be interested in it. This is irrelevant though, it's not a good benchmark at all.

Back to your original point, where are you coming from? I'll tell you where I'm coming from. I started making games in BASIC on the Commodore 64 in the early 90's. The next major attempt was in C with Allegro on DOS. Again, with C++ and SDL on Linux. And again with another library, another language, moving on and on. I suggest you try making a game from scratch with SDL and C++ and then come back and say Unity makes game dev too difficult. You will realize just how wrong you are. I was never able to get any traction with just a language and an API, it's not that it's flat out too difficult to do, but it takes serious dedication to go that route. As someone who spent decades banging their head on a wall doing it the hard way, I can only say that you are very, very, very wrong.

 

Sounds like I am coming from very much the same place as you are. Started on a TI-99/4a in early 80s then C64, etc. also did the DJGPP C compiler and Allegro thing, etc. When I say does Unity make it too difficult I mean generally speaking compared to other modern dev alternatives (or older alternatives) that have a solid game oriented API built in. Not talking about literally writing low level drawing routines, music playing routines in interrupts and so forth. Been there done that. I am talking about using a solid existing api.

Although saying that, writing a ray casting engine these days is not a big deal due to tons of documentation available online. Wasn't that way for us obviously long ago as you know. I knocked this out in AGK Basic last winter in maybe an hour or so.



Obviously, it is not a game and not textured but was a good proof of concept.

Anyway, I think mainly I am just wondering what tricks there might be to complete such a game in Unity quickly that I am not aware of. Could be store assets or maybe there is just a "best way" to approach such a game specifically in Unity. But if a person isn't doing that, maybe doesn't know about it, then it actually becomes much harder in Unity than it might be in something else. That kind of thing.

And more than that why isn't making such a game just dead easy these days in Unity. To me it seems like it should be. Kind of seems to me like beyond graphics, sound fx and so forth the entire game should be doable by nearly anyone & everyone in a week using Unity. But that doesn't seem to be the case. So I wonder if it is tools such as level design tools (although I agree with Eric just use tile maps for this. I made my level in AGK right in code in an array) or what exactly is the bottleneck.

 

And Unity doesn't? I'm honestly wondering what you think could be so much easier. Like I said, I've tried other engines like Construct and Game Maker that let you just add things like "Player Movement" to an object and then the player can move it. That's much easier than Unity, but if you want to have the player move in a different way you're immediately crippled. Things like that are easier when you're first starting out, but for anyone else it just gets in the way. It ironically makes things harder.

 

Ha ha. Maybe I am interpreting you incorrectly but I am getting a somewhat defensive vibe here. When you say
"And Unity doesn't?" Well I don't know. Does it? Can nearly anyone knock out such a game in say 32 hours?

And when you say "I'm honestly wondering what you think could be so much easier." Anything and everything that is currently an obstacle in being able to complete such a game in about 32 hours.

Maybe. Just maybe there are best ways to approach such a game in Unity that for whatever reason most people just do not know about. And knowing those things makes all of the difference between 32 hours being plenty of time for some and 320 hours being too little time for others.

 

Well what is "anything and everything?" Give me an example. Because the example I gave (the canned player controllers in Construct and Game Maker) to me fits squarely in this category and, as I said, those end up making the task either harder because you're trying to shoehorn your modifications into a black block player controller implementation, or neither harder or easier because you'll end up making your own from scratch anyway.

 

Well one way for example would be an asset that comes with a level designer so the user can quickly define the floor and ceiling textures perhaps even in sections and also can drop in walls, doors and switches. Then the "running" state would provide an easy way to render the current view based on position and direction facing. I have no idea if such a thing exists but that is an example of something that should greatly decrease the dev time of such a game. Of course a script can be dropped on every enemy and collectible to have them always face the camera (player).

And again this is why I mentioned previously even using 3D is fine. Because I think much of the level building I just described could be done right in the Unity Editor for 3D.

It seems like beyond time spent on graphics and audio it should be easy to knock out the core of this (as in 1 complete level and a player moving around with collectibles to pick up, and some enemies standing around) in an hour. I doubt it would be true though. But perhaps for someone here it is. And if so, it would be damn nice to know that and how they would do it.

 

Who is included in your definition of "anyone"? Someone who is knowledgeable should be able to make a prototype on par with the average entry to Ludum Dare in 32 hours if we assume that the average contestant gets four hours of sleep per day.

 

I think I had the wrong lead in for this thread. You're basically at the heart of it. What would be the easiest & fastest way to approach such a project in Unity and using that approach (not counting time spent on graphics and audio because we all know those can be thrown together quickly or iterated on forever and never done)?

Perfect scenario... it's Ludum Dare and you want to make a version of Wolfenstein 3D. How do you do it?

 

The project is too simple to grab assets for it. On other hand it is uncommon enough so there probably isn't a kit.

That's a prototype. Not a finished game. Ludum Dare is also mostly prototypes.
You're asking for a finished product.
It is possible to prototype 3rd person combat game in unity from scratch in 15 hours or less. With low-poly graphics and skeletal animation. However, total production time is prototyping time multiplied by some very large number (prototyping_time * 50, for example).

It should be definitely possible to prototype this kind of game in 32 hours or less, BUT that'll be still a prototype, rough around the edges. Polishing it will take time, and you can't really magically speed the process up.

 

I would sit down and just make the whole thing. Seriously.

75%..90% of time would be spent making art resources (several wall textures, sprites for guards/enemies, sprite for guns, and maybe animated portrait at the bottom if I have the time).

The rest is making bunch of capsules walk in random direction while polling environment for presence of player. The game would be probably 4..6 monobehaviors total. Player, Enemy, Door, Collectable, Combatant... something like this. Coding part is extremely simple and it should be possible to complete coding part of the game very quickly.

 

Well that's true. Again I should have clarified this. What I am getting at is like this example... the guy created this Wolfenstein 3D clone <-- that is link to playable web version... for LD28 using Javascript and OpenGL.

Yes it is rough and obviously graphics are primitive as well. But I am thinking the idea of Unity is that it should be allowing a person to develop much more easily than creating a game from scratch using Javascript and OpenGL, right? So it seems like if people can make stuff like this example over a LD weekend then a person using Unity would be able to either do the same in less time OR get more done in the same amount of time.

With the idea of the 32 hours being you have a playable level such as this one at the end. Then a person can take the next day and build out all of the other levels. Take the next day and work on all of the HUD. The next day focusing entirely only on the graphics. And so forth. And at this rate of development I would expect the full reasonably polished game would be complete in 7 to 10 days.

Would you say that is accurate or just depends entirely on the game being made?

 

I'm not sure what you're trying to get at here.

At a quick estimate coding part of unity3d wolfenstein prototype could easily take as little as 4 hours. Or less. The rest is content making, no coding involved. Of course it is a very quick estimate, that doesn't really take into account accidentally stepping onto some sort of unexpected landmines, but in all honesty, in ONE day all coding for this kind of game will be done.

Compared to traditional programming speed, unity development speed during prototyping phase is insane. It is at the level that can make "I'll write my own engine" programmer weep with envy, rethink their life, quit programming and become a carpenter instead. You should be at least twice faster than that javascript guy by default.

So, if the idea is "unity should be faster", then we're already there. It is faster - you can create a result equivalent to several months of work in a day. If you want it to be even faster than this, then it is pretty much asking for a "make a MMO" button.

 

My 2 cents:

If you use a tool for game design it will always be a way of compromise.

• With a low level code based approach, you will be able to get the maximum out of the hardware at the cost of complexity and re-working a lot of things without the need
• With a very specialized tool (I never used one, but I guess I am talking about something like RPG maker or other examples mentioned above) you will trade in flexibility for ease of development and you are basically only producing the game art and story
• I think it would be interesting to divide up the time for game development in categories like project management, game design, programming, graphics, sound, special effects, level design, etc. and see how much time is spend in each category. This will inform a little bit of how much time is in general impacted by the respective tool of choice to produce the specific asset.

In the end, I see Unity mainly as a tool to help manage the different assets in one place and linking them up. Yes, it can do much more but you are quite flexible in terms of what you want to use and what not. I do not see Unity as a "very specialized tool" for a specific type of genre or platform, hence whatever example you pick, there will likely be a better tool for it's specific case.

I have the feeling that you spent a lot of time identifying roadblocks / bottlenecks. Great, this is important to understand but I think your challenge is to produce something you really want (instead of small prototypes)...

 

I think those are solid points. Well this was more just wanting to understand how much value people are really getting from developing in Unity compared to alternatives. Like is Unity really making game dev easier and faster or do people just *think* that.

As for me I think I have pretty much come to the conclusion I will not make the kind / scope of game I'd like to make simply due to the amount of time it takes. Again this is where if there are things about Unity dev that greatly speed up the game dev I'd love to know what they are. I've not really seen them yet. Again not compared to writing 3D or 2D graphics engines from scratch. Never understand why people look at it that way as if the only two options are Unity or write it all from scratch from the ground up. lol

I do think Unity is likely one of, perhaps THE, fastest way to create 3D games. But I don't know that for sure. I've always focused on maximizing dev speed because it is so important to tackling larger projects.

Maybe I'll test it sometime to see how long it takes me to make a W3D style game in Unity. Maybe not. I just think it is odd that in all of these years making games today is not much, if any, faster than it was in AMOS long ago or Blitz on PC. It's just very odd.

 

I think the best would possibly be to define exactly what game you want to make, break down the work required (estimates) and then see how to best tackle them. Decide the tools to use, decide what to buy and what to create, etc.

Maybe a specialized engine exists for your game and you can focus on the story only + buy the assets? Or any other combination. Maybe even a close enough game exists and it's modding that gets you there...


Edit:
If the game you want to develop is not similar to w3d, what is the point in benchmarking the time it takes. I could still do many effects faster on a c64 with a bare bone assembler than doing them in unity, but at the cost of locking it to one platform and mainly because I understand the rendering pipeline much better (it's way simpler). Then again, it's time is over and there are many things I would never be able to do because things have moved on.

 

Hmm. Well, consider the following:

Would you get into developing at all without Unity? Would majority of people using Unity for their projects?

Looks like it made game development easy enough for you to pick it up
I personaly like it because it makes building for multiple platforms so much easier. Everyting else I need, I can script myself.

 

That's a really bad example. Pregnancy and child rearing takes a few minutes of activity, which is normally intensely rewarding. Once it's started you are committed.

I'm sure if my game started screaming at me every time I neglected it I would work on it more too.

 

I have to agree. And people will probably need to be more careful with their side projects then, unless they want to get involved in copyright disputes.

 

How many hours are you estimating for playtesting, QA and getting feedback from external playtesters? And how many for making the artwork, sound and music assets from scratch?