Seasoned Unity developer here. Be careful. Unity is a mess.

Upon trying the latest release of Unity to develop the next game, I'm appalled by the changes in this company and platform that have occurred over the past few years while I took a hiatus from game development in my spare time.

I started with Unity 3 in 2012, then moved up to Unity 4 Pro, and Unity 5 and have launched mobile apps. I took a two year break from game development and now I'm ready to jump in again, but everything in Unity is now a big fat mess.


DOCUMENTATION
The documentation is horrible and broken..

• On the Unity website from the main page, documentation links at the bottom of the page are broken.
• On the Unity 2019.3.13. install page, documentation links are broken
• From the manual, the drop down top-left to select older versions of manual or scripting reference doesn't work.
• Networking pages in both the manual and scripting reference is confusing or missing in action. Everything is marked 'obsolete' or that 'UNet is deprecated..."
• I played around with the new Input system and found it confusing to figure out what is going on. Documentation is missing, confusing, incomplete and there are no tutorials to get one started quickly. I got it working, but only after watching 3rd party YouTube videos. Most of the new input system is all over the place and remains confusing (or perhaps broken).

BACKWARDS COMPATIBILITY
Unity backwards compatibility is non-existent or broken. My old projects cannot be upgraded without major effort, because Unity keeps deprecating core elements and replacing them with entirely new ones. Here are some of the hitches I encountered in the past week:

• How many times has Unity replaced the particle system with zero backwards compatibility?
• Everything in Networking is either obsolete or deprecated. My games depended on P2P LAN. Now I want to include multiplayer. Neither seems possible looking at the documentation.
• UnityScript was very popular and my projects used both C# and UnityScript. UnityScript was dropped. (Personally, I think that UnityScript was one of Unity's strengths over UnrealEngine for beginners. Without it in Unity, beginners might as well learn C++ and go with UE4 instead.)
• Documentation in general seems to be part of an afterthought altogether.

For upgrading older projects, which could have hundreds of thousands of lines of code:

• The JS2C# converter doesn't work on the scripts by themselves (which would've been the preferred method). Instead, it expects the entire project to compile successfully in the later Unity version before it will work.
• Since the project can't compile due to all the deprecated elements in the old Unity version, there is no workflow (test, debug, install) to upgrade the project and its scripts into the latest version of Unity.
• Of course, older packages purchased on the asset store that a project depended on won't compile either, so they have to be deactivated or deleted before importing.

GETTING ASSISTANCE FROM UNITY
I've tried contacting Unity through the contact link at the bottom of the website a few times to get guidance and help, and nobody has responded. So, this company now also ignores developers trying to reach them.


GENERAL IMPRESSIONS
I consider myself fairly seasoned in Unity, but being confused myself by all the changes, lack of backwards compatibility, lack of documentation, and lack of support, I don't know where a beginner would even start. I wouldn't recommend anyone to start in Unity until they Unity got their act back together. I feel an urgent responsibility to warn people through op-eds and videos.

Developing a game or creative content takes great effort, money, and commitment. As an indie developer with a personal stake in the success your own projects, you also invest in the Unity as a company and, in essence, a partner. Everything you put your heart in, depend on them too. If Unity can't be trusted for support and for keeping developers needs in mind in terms of compatibility and upgradeability, to keep the pace with rapid changing technology, or they don't care about their developers, then there is no point in investing in this product.

Not responding to my requests for help goes one step further and got me angry. I've been one of the suckers who got fooled into investing in the Unity licensing cluster####s of the past, and having spent that money I at least expect to be able to get back some support when I ask for it. What does it take? Do you now have to go through your lawyer to get a response from this company, because I'm willing to go that route.

To me, these problems I encountered in just a week are signs of a company that may be mismanaged and possibly going under. I could be wrong, but that's just my impression being an old guy in his 50's who has seen that happen a few times before.

 

While not comforting in the short term, I do know these sorts of issues are making their way up the visibility chain at Unity. It is a big organization, with a lot going on, so it may take time for any course corrections.

 

Yes unity webmaster is a monkey. But it's your own fault if you use JS / UNet. You basicaly asking for trouble. Those were created for people that like challenge.

 

Its so funny having spent the last 6 months over in unreal land and seeing these same types of rants with the exact same points regarding unreal over there then come here and see the same thing.

 

That's because no software is perfect. Neat tech demos come out that woo people but they don't realize that every ecosystem has problems.

 

I mean, if the problem with Unity is that they're too big... Then they should be less big.

By the time things make their way up, there will be a new host of problems and issues that will be too late too fixed.

 

C# is absolutely not equivalent to C++. I say that as someone who uses C# all the time and struggles even understanding some of the C++ I've had to wrangle with.

 

A heck of a lot has changed since Unity 5. Do you think one week is really a reasonable amount of time to catch up with all that and upgrade your projects? I've been working in Unity more or less consistently since v2.4 (and on and off before that), and I wouldn't expect to upgrade a largish project in under a week, let alone if I had to learn about everything that's changed as I went.

When upgrading projects which have incompatible assets, the first thing I do is open the project in a version of the engine / Editor that the project was made for and swap out all of the incompatible stuff. You have to do that anyway, so you may as well do it in an environment that'll help you. That'll then mean that the updated version of Unity can help you get it the rest of the way.

Upgrading in steps can also help. I'd suggest stepping the project forward through the LTS releases rather than going from Unity 5 to Unity 2019/2020 in one step. That way you don't have to deal with all of the changes at once.

I've given them similar feedback. Isn't it still in preview, though?

I found it pretty quick to get set up and running and I quite like the design. But it does a lot more so there's more info to chew through before you can start implementing whatever is right for your project. For my use cases now it's far superior to the old system. However, I'm not using mobile / touch input so I've no idea how well it fares with that stuff.

 

Particle effects have turned up broken between different versions for us atleast. And then we do not talk about legacy particles.

 

No you are right, though thats only semantics its still very bad for us developers when it happen More so for us smaller indie teams that dont have particle talent on the team but rely on asset store for such things.

 

Use Photon or Mirror, they're quite good, mature and free. Though yes, I agree, the UNet deprecation was handled very badly by Unity.

There were quite a lot of good reasons Unity deprecated UnityScript. Only 3.6% of projects had more than 20% UnityScript included. Unity's going through a huge transition period with lots of scripting upgrades, the entire DOTS technology stack, assembly definition files etc. Having to support UnityScript would have slowed them down even further. If you look at the comments on the blog post, it was a popular move in the community too.

I used UnityScript when Unity 3 was out, too, but I don't really miss it. Personally I feel like C# is way more user-friendly than UE4/C++ too. At least the entire editor doesn't crash when you get a null pointer.

Finally, you don't have to upgrade if your old projects are already working fine. You can download basically any version of Unity you want.

 

UnityScript should not have been done in the first place!.. Just like HDRP, LWRP, URP , LT, ect. ect.ect it just goes on and on.. Messy Cluster F**K

 

UnityScript was all the way from Unity 1, though, not at all related to our current mess? There was Boo, UnityScript and C#. Unless I'm misunderstanding you?

 

I used csharpatron to update nimian Legends, hundreds of js files https://forum.unity.com/threads/csharpatron-the-definitive-unityscript-to-c-converter.259847/page-2

Unity dropped JS in 2018.2 so it works only up until unity 2018.1 see the forum link and you can DM him to see about getting a copy. It's daunting but doable.

Legacy particles are updateable using a unity provided script.https://forum.unity.com/threads/legacy-particle-system-updater.510879/ You have to use it in 2017

I would recommend downloading unity 2017.4 for these reasons and update these first.

Also you will need to remove any old UI and use the new unity 5 ui since legacy , the ones that use GUITextures and GUIText are removed in unity 2019.3 ( I wish this had not been done in an point release)

Unity has made upgrading to new versions with legacy projects extremely difficult. It is the reason I Postpone updating some of my games, so I feel your pain. But it is doable.

This is the thing I want unity to fix more than anything. Post processing 2, built in render, deferred shadows, these are some of the many things being removed and it makes development difficult even though the engine itself of great to work with when you are not fighting project breaking updates.

Breathe and good luck.

 

To be fair, 3.6% of devs using it heavily and 0.8% of devs using it exclusively in their projects doesn't really scream "very popular". It wasn't even dropped overnight, but phased out gradually. All of this was already known since August 2017.

 

I recommend you jump up a version at a time and to 2018.1, not .4 first, before the nested prefab system was added (unfortunately in middle of 2018 - .3?)
UnityScript/JS was an easy move from something like Flash's AS3 but it was deeply flawed in some areas with hacks to get it close (but never close enough) to C# which is much better tooled and more 'standard'

 

"By 2016, the company reported more than 5.5 million registered users."

That means 180,000+ using unityscript heavily. I was one of them and during my attempts to upgrade to c# which did successfully I encountered many dismissive comments about unity script users.

 

I feel your pain. In my case, I absolutely love C#. I never used UnityScript, but I realize that some people did. I have a lot of experience with C and C++ as well. C# is super easy compared to C++. Coming from UnityScript, it will be far easier for you to learn C# than C++. Also, C# can access the same Unity API that UnityScript could, so you will not need to re-learn the game engine's API. If you switch to UE4, then you would need to learn the UE4 API in addition to C++.

The old and new input systems are not ready for production use, yet. I strongly recommend the "Rewired" asset. It works really well, and is very flexible. I am hoping the new Unity Input system will eventually be as good as the Rewired asset.
https://assetstore.unity.com/packages/tools/utilities/rewired-21676

For networking, I implemented my own solution completely from scratch using UDP sockets in C#. I realize that may not be what everybody wants to do, though. One advantage of building my own networking solution was that I could build a perfectly optimized solution specifically for my game instead of trying to use a generic solution.

I love Unity, but I have seen a few major issues as well. For example, I ran into a bug where Unity does not allow for more than 4GB of assets in a scene. Here is a link if you want to read about that:
https://forum.unity.com/threads/bug-4gb-limit-to-textures-in-standalone-build.441116/

I originally posted about this bug in November 2016 when I ran into this problem using Unity 5.5. That bug remained in Unity until December 2019 when it was finally fixed in Unity 2020.1.0a19. The fix was never back ported into Unity 5.x, 2017.x, 2018.x, or 2019.x. The only way to get the fix is through the 2020.1 branch, which is still in beta. To this day, none of the production ready versions of Unity include the fix, so none of the production ready versions of Unity support scenes with more than 4GB worth of assets.

 

I was one of them too. The vast majority of BHB: BioHazard Bot was originally written in UnityScript. I’m perfectly comfortable with C# now but I did not like having that change forced on me. Adding new features is fine, taking away old ones is not.

 

There's always the option to stay with an older Unity version if necessary. In which version did you make the game? While the phase-out began with 2017.2, UnityScript stayed up to 2018.2. (

2017.2.0

got released on Oct 12, 2017, where as

2018.2

got released on Jul 10, 2018)

Depends. In case of UnityScript: Supporting other languages apart from C# means taking developer capacities away from things that may be more crucial (such as modernizing the Editor). It makes no sense to do that for a feature that has a relatively small user base.

 

Mobile developers don't have the option to stay with older versions of Unity. We are often forced to upgrade versions order to meet new requirements by Google and Apple, especially with Apple where Xcode, iOS and MacOS are all updated on a yearly cycle.

 

Let's be generous and say only 40,000 Unityscript users then.

 

No, there isn't always the option to stay with an older version of Unity. Platforms change, the steam plugin might only support the latest engine, etc. If you want to port your game to Switch or maybe the upcoming PS5, that might not be possible without the latest version of Unity.

This isn't about Unityscript specifically. Being forced to switch from legacy PS to Shuriken was a ton of work too. This is about the removal of old features and pushing everyone onto the latest and greatest even when it has a more convoluted API and sometimes doesn't even have full feature parity with the old versions. Or pushing it on people that might not even need the features of the new shiny thing and are perfectly happy with the old one.

Removing it (meaning not just Unityscript, but legacy PS or anything else that might get removed in the future) means making a huge chunk of learning resources and videos obsolete. It means cutting out a big chunk of the Asset Store that was built on it. It means breaking backwards compatibility for people who may have had years worth of reusable code and who planned on using that code for years or even decades to come.

 

OK - there were only 400 people using unity script when it was dropped.

 

I agree. One asset store dev recently announced they were forgoing HDRP support because of the complexity supporting multiple render pipelines, and that the asset update would drop support for Post Processing stack 2, which is relatively new, because of the incoming PPv3.

 

No, I don't expect to upgrade a project in one week. I said I encountered all these problems in just one week, which has nothing to do with my expectation for upgrading a project.

My experience with Unity is you can't swap out old stuff for new stuff in the old stuff

So your solution is to develop the game 4 times in order to get it upgraded to the latest version? WONDERFUL.

I tried the input system in Unity 2019.3 (latest release) in which it is not reported as 'preview'.
If it's still in preview then why doesn't it say so in the documentation and when you open the package manager?

I develop for mobile with touch and gamepad support, which brings me to another issue that I haven't mentioned before. Gamepads are also sparsely supported in the new system as far as I could tell from the testing I've done so far.

 

Which 'legacy particle system' are you talking about? There was a legacy particle system when I implemented a newer one, which now produces errors.

In your forum the first thing I read is this:

"Unfortunately, scripts cannot be updated automatically."
​

So it's not really a solution. Someone still has to update the 700 compiler errors in scripts and 2700 lines of code manually.

 

They just decided to cut off and release as 1.0 not too long ago. So technically it is not preview anymore.

And they never told that they support every single gamepad. Actually they're really open about it. They will work on that after bring up the documentation to par with the rest.
Until then, if it is not good enough for your use case, you can use the good old system without any changes.

 

I'll check them out, thanks.

Careful with statistics. Statistics are easily manipulated to fit a subjective narrative. (Except for pandemics, of course. Those are sacred.)

Let's see. I learned C (predecessor to C++) in 1988. Then C++ in the 90's.

I taught some classes in Unity and Unityscript was a great tool to teach programming, because it was readable and had very little overhead from the user's perspective.

Please explain that to Apple, Google, NVIDIA, SAMSUNG,... It's not just the software either. Hardware too. I'm sure Unity folks can chime in here, because they're also beholden to the Hardware and OS makers...

 

Just a heads up in case you revert to the old input manager:

In the old Input manager, Gamepads are currently broken for iOS and tvOS on all current Unity versions.

 

As I recall the only problem I encountered using UnityScript was implementing UDP and I ended up coding that in C. Other than that UnityScript was wonderful. And as I already said in an earlier reply - especially to teach coding.

 

If that works it'd be a lifesaver. I see it's no longer available in the Asset Store, but I will contact him.

I agree with you on the point release removals.

Thanks for the tips. The UI is another one, yes.

Yes, yes, and yes. I hope you are right.

 

Most of the old users know where he is coming from. Unity is horrible in many ways to say it nicely. C# is the only reason I still use Unity these days and it works if you are just making simple games. But you will run into walls quite quickly if you try to push it and no one will fix the obvious problems. The biggest problem is that, it feels like there is no ownership in this company, just bunch of marketing people plus some part-time developers.

 

I abandoned the idea of pulling in the assets intact early on, but wanted my scripts converted to C#. The reason is that Unity's habit of dropping essential core elements and replacing them with new ones without any backward compatibility, also meant that 3rd party packages purchased in the asset store that the project depends on also ended up deprecated.

Upgrading Unity projects is more of a SALVAGE OPERATION than anything else. You're left stumbling through the rubble trying to pick up things you may be able to use again.

 

You are upset because projects from 8 years ago don't compile and work out of the box in the latest version of Unity?

As described above there is a piecewise path forward, LTS version by LTS version to get where you need to go. It's not a small amount of work, but doable if you want to invest the time. OR, you could just work on the projects in their respective LTS version of the engine/editor.

Unity is a company of finite resources. I would rather see them focused on current quality, *recent* backward compatibility, and new feature development. Even if it were technically possible, had they done as you desire and spent hundred of person months of development ensuring 2.x to 2019.x auto-upgrade capability, what would be the ROI on that? The three people needing to do that would be happy, sure.......

 

And yet, when installing the latest release of Unity (2019.3), the new input system is not in preview.

Thanks for the rewired asset tip.

In Unity 4 I used UDP to initialize and then let Unity Networking do the rest. However, my UDP script (which is not in JS) produced several errors from the API Updater when trying to import the project. I haven't yet delved into it to see how to solve that.

I'm still confused about what's going on with Unity networking. Can anyone tell me what is not deprecated or obsolete? Going by the documentation, Unity networking is nonexistent.

A 4GB limit was already a big deal in 2016. You'd think a fix would take some priority.

 

This.

 

Nope, that option doesn't exist for mobile.

Taking resources away for modernizing the editor? Interesting. The editor is good and, aside from minor changes, has changed little as far as I can tell. Blender on the other hand, has made a quantum leap in changing the editor. For Unity, that would put that at the lowest priority. Unity seems to make changes in ways that suit the developers at Unity, instead of the developers who are their clients.

How do you know UnityScript had a small user base? Did you include the classes full of students learning with it? Did you consider thousands of people learning to code with it from online videos on YouTube?

I'm not particularly against a company's decision to move away from something like UnityScript. I am upset about how it was done by Unity. The JS2C# converter that works only in Unity 2018.x doesn't convert scripts by themselves, but expects a fully compiled project to function. But the project can't compile without a major overhaul. Therefore there is no acceptable workflow to upgrade a project.

 

So according to you there were only 180 UnityScript users.

OKAY.

 

I was one of them.

But whether a number as low as 180-400 or even 1000 is true or not, we're not complaining about a scripting language to USE. We're complaining about the difficulty UPGRADING from the old to the new.

I wonder. Was the number of published games using UnityScript factored into the decision????

 

Who cares about UnityScript, it's old news. THEY HAVE BIGGER PROBLEMS TRUST ME...

 

This is a huge issue for me as well.

 

Every word in this post.

 

See a few responses down from yours about how the old one is broken for mobile...