Official: How Can We Serve You Better?

I hear that Unity say plans... but Unity don't say "when".. Timeline is important in planning

plan is meaningless without timeline . tell us at least season(spring,summer,fall,winter)

 

Yeah. I love this community and how much passion and enthusiasm everybody has. It takes me literally hours each day to read through everything and always about halfway through I start reading that people think we left. Also, I see many posters sort of skim (which is to be expected) whereas many of us are actually reading them all in their entirety.

That said, there's some pretty clear patterns and requests and we're aware of them obviously, but it takes time to see the fruit of all that, plain and simple. Some things have already come, some coming each day, some will take days, some weeks and some even months. Same for your own development, same for our competitors too. We applaud Epic's recognition that you are a community worth having as customers, we obviously agree, and for all the things people don't like we have also helped a fair few people get into and prosper at game development.

 

I disagree. At least on the 'time' part (saying Unity 5.0 is okay). I think that giving timelines and not respecting them (while keeping the road map opaque) is exactly what made many in the community feel resentment towards Unity. (along other things)

Right now, they are sharing openly the 'what they want to do' (and a bit of the "what they _will_ do"). Adding a timeline alongside the current communication will only undermine the efforts to mend the past few years and feed trolls that will confront any/all progress with negativity when those timeline aren't met.

I believe this will eventually end with the "castration" of unity devs trying to be open with the community and we'll be back to square #1; "Don't talk to the community about internal dev plans".

It's not for me to do their communication strategy but it is for us, as a community, to not push them in the wrong direction either.

Edit: My advice, never plan ahead with critical parts being unreleased features. I've learnt this the hard way.

@Unity;
Thanks again for opening the lines of communication.

 

Timelines are fine if they explain why they've been missed. We never got that, especially with the GUI, which was always "coming soon, but it's been pushed back."

 

I half agree with this. Yes, saying GUI will be the part of 3.x cycle was a mistake. Obviously nothing was ready at that point, and we are only now getting the new GUI in 5.0. Yes, it'll be available in 4.6, but if it is going to be released around the same time (with 4.6 being a GUI exclusive release mind you), we might as well call that 5.x cycle.

We are all developers, we understand delays, the problem arises when mentioned features skip 2 upgrade releases. I don't think community would be up in arms if UNET arrived in 5.3 instead of say, 5.2, or even 5.1.

There are many mentioned improvements in this thread, most of them point to 5.x cycle. That seems OK, but you have to keep in mind, 5.x means anywhere between 8 months to 2.5 years. Saying UNET will be a part of 5.x means nothing, if the phase 1 comes with 5.6 (ala GUI).

I don't think anyone in this thread wants the specific day or week or even month of a feature release, but mentioning which 5.x release they plan a feature to be in, would go a long way.

Edit: Not to sound like a broken record, but Epic has given specific months to specific feature releases on their public roadmap, with a clear "this is not set in stone, things may happen to things" label on it. When you take that into consideration, saying "it'll be in 5.x" falls extremely short.

 

It's not easy to explain why they've been missed other than "It's taking longer than expected because of Bug X or missing feature Y". It ends up sounding like a broken record after the first ... couple times. The GUI is a special case where they had to scrap it and redo it (which they did say at multiple points over the last 2 years).

In relation to that subject, 20'ish pages ago I proposed an alternative solution to timelines and roadmaps; open up the Alpha/Beta groups to Pro users.
- Want to know the state of the next.. 2 version? Check out the Beta/Alpha group.
- Want to contribute? Check out the Beta/Alpha group.
- Want to try out new features? Check out the Beta/Alpha group.

... I think you understand the gist of it.

I agree. The "5.0" was meant as an example. My problem with timelines is with 'precise' _time_line as what thedreamer was requesting.

Fair point. I guess I have less expectations towards Unity and I'm more worried about the community not being able to handle "things may happen to things" ( the increased noise ratio that it involve). I think we can all agree that Epic never had to deal with such a massive community as Unity has now.

 

I read some of my previous posts, and they come across as quite terse.
In an effort to further help UT understand how this particular customer (paying for several Pro seats + pro addons, etc.) views the situation, I will attempt to cut to the chase:

It is frustrating to be heavily invested in and paying license fees for a solution that has so many fundamental holes to plug and so few options to influence the development or get support. I would rather have paid much more for the pro licenses if that meant that:

- we could avoid "having" to rely on the Asset Store to (efficiently) fill certain needs that should be in the core engine.

- bug reports would have been replied to and acted on quicker

- support was easier to come by

To be frank, we have not tested premium support, mostly because when we do run into problems we cannot solve it is due to suspected bugs, which premium support apparently does not concern itself with.

So, please consider providing better service for those that fall inbetween amateur bedroom developers and enterprise VIP companies.

 

This is great we finally get some response to these questions... except for the one glaring thing lacking, no specific time-frame on the mono upgrade. Please, please start giving direct, specific answers. In what time-frame a lot of these "features" will be implemented is very important to the development process of my team, and I am sure many others. If these things roll out with Unity5 awesome, I will upgrade my licenses and it will be used on our next project, if I have to wait another year for mono updates, then I am sorry It's just more of the same and I don't think I will be able to use Unity anymore in our development. I hate giving ultimatums, but our next project is rapidly approaching, and I cannot afford to hold off production because we don't know what state Unity5 will actually release in. Unity4 will not cut it any more.

 

Reply is waiting for you in that forum thread.

 

I think in one of the demo videos you could see a dropdown to change navigation direction between buttons (e.g. if you press right arrow or joystick, button y will be highlighted)

 

Thank you, my post here was to further elaborate on the frustration caused by ambiguous answers. I appreciate your response, and to a degree I understand, what I take issue with is that it has taken this long to even address many of these issues, and this is only happening because major competition has come to bare. If that were not the case we would still be in the dark hoping for nuggets of information.

I really do like Unity, I like a lot of things about it, prototyping is rapid and accessible, but there are so many barriers to making a complete, whole end product that its frustrating to no end.

 

Additional effectors would be cool, ala HumanIK. I'd love to see configurable IK solvers I could attach to generic skeletons, in addition to the dedicated humanoid/biped solver. Something like...

Code (csharp):

1.  
2. public override void OnStart()
3. {
4.     IkSolver arm = animator.AddIkSolver_2Bone(shoulderXform, wristXform, ...);
5.     IkSolver tentacle = animator.AddIkSolver_Chain(tentacleRootXform, tentacleTipXform, ...);
6.  
7.     arm.SetReachTarget(pos, quat);
8.     ...
9. }
10.  

Or even better, if something like my proposed low-level API happened, expose these as posing nodes so that I can control when and where during posing IK affects the pose, split IK into multiple passes (e.g. solving hips/legs, processing look-at, then solving arms) or even solve IK multiple times if necessary.

 

I think FINAL IK from assets store is good general IK solution - and if UT will ever implement better IK then i think UT should hire Partel Lang as similar to aren (GUIru) or as consultant

 

Good to see that mecanim will get a fair upgrade.

However, the point was not that Mecanim is unusable/broken, but that it was when launched.
That only adds to the impression that UT develops shiny features to get people upgrade their license.

It's fine if a product is not 100% perfect when it launch, or has some awkward interface. But when people upgrade specifically to se a new feature, it kind of sucks if they have to wait for the next major upgrade to really make use of it.

 

My two cents...

I am a 2d/3d artist at a smaller subsidiary of an already established game developer. This small team does not have the power to integrate the whole lot of tools that we need for our day to day production. So we heavily rely on the Asset Store. That's what we like about Unity. You'll most certainly find the right tools made by someone from the community. There are downsides to this approach, but more on this later.

The issue we quite often see with Unity is that it has grown into a mixture of good and not so good implemented feature sets over the years where some features severely lack in documentation and API functions - and UT's bug fixing takes ages.

Some examples:
Surface shaders - I've filed a bug report regarding missing fog rendering when using vertex colors, worldPos and worldRefl vectors and dynamic shadows at the same time. No idea if this is intentional or a bug. No workaround. Never heard from you again. Whenever a bug commited by a user gets updated internally it would be nice to get an automated feedback mail. So we know that it is on your roadmap... somewhere.
Shuriken particles - this system looks like it has been inspired by the UDK cascade editor. But then again it is lacking in functionality. Dynamic parameters for example would be great. Adding custom modules to the Shuriken system would be awesome... but little is known about the internal working of the particle renderer and even fewer is documented about its editor GUI; the API documentation could be more elaborated. Writing custom shaders for normal-mapped and lit particles for example requires deeper knowledge. In UDK this is a breeze.
Image effects - half the code is in js, half the code is in C#. Not helpful. Please offer both. Actually this applies to all assets and example scripts, please offer them in both JS and C#.
Terrain system - very old-school. Fortunately several assets made by community members exist to pimp this system. It would be nice to be able to clone stamp and rotate areas from a given heightmap area. Also a road tool would be awesome. Again, this can be bought from the Asset Store.
More sophisticated standard terrain materials would be nice. Global color map, global normal map, splat map with detail textures (diffuse, normal, spec)...
Water - one of the cooler assets. But still one that requires workarounds and has issues when working in tandem with the already listed standard asset... I am speaking of the tree billboards of the terrain system that can create halos when seen from far away against a water background.

In general dropping a standard asset into the scene view should just work. Regardless of whichever other standard asset has already been placed. There might be assets that exclude each other from being useful ... but that shouldn't be the case with water and terrain - most outdoor environments use both.

Speaking of the scene view. Please make this view live. All post effects of the main camera should just work on the scene view.

More filters for searching the scene hierarchy would be nice. Searching for objects with specific flags (static), specific components (even custom ones) would help alot.

Nested prefabs. Period.

Earlier I mentioned the downside of relying on the Asset Store too much. You as the supplier of Unity3D will most likely stay in business much longer than the creator of a 3rd party plugin. Also they (3rd party devs) might just lose interest and abandon their asset. So I would like to see your standard assets to be more polished (both in the editor and the documentation) from the get go.

I could go on, but it's bed time now
Thx for this opportunity.

 

I only support perpetual licensing, a subscription option is appreciated but would hate to see it become the only option.

 

"Give an inch, they'll take a mile." I'm all for increased communication, but I'm not going to ask for direct and specific answers where Unity isn't comfortable giving them. If they're not comfortable setting a date for something then there's a reason behind that, and if we force them to start slapping dates on them there's a whole bunch of negative side effects.

 

That may be fine for you, but I cannot make business decisions that not only affect me, but the livelihoods of my entire team without hard, factual data. I didn't ask for a date, I asked for a time-fram (days weeks, months) or if the mono update would be included in Unity5, which it wont be, so now I am left deciding if I need to switch engines for my next project to something that will be updated in a timely manner. Unity holding everything back so that they can be on the mobile bandwagon is not conducive to my strategy, mobile is a saturated market where you have to sell for pennies and hope to get lucky, or stock heavy in ads and again hope you get lucky.

I don't know if I am prepared to sit around hoping that they get to a mono updates in a timely manner so that we can use current .Net capabilities. I remember when we were promised a new GUI system back in 3.5, just look at how well that turned out... For solo developers, and hobbyists this kind of thing might be perfectly acceptable, for anyone seriously trying to run a business, its a joke, and I am always left feeling I am at the butt of it for paying for upgrades that do not net me the features and promises made.

 

You've got to understand, it's not Unity holding back Mono upgrades. It's the licensing terms of Mono holding back Mono.
So basically, what Unity is doing here is trying to resolve this by circumventing it altogether - they're translating IL code to C++, and then directly compiling that for the target platform.
Now, WebGL is first up to receive this treatment, because it's actually required in order for WebGL to work (they need the C++ in order to feed Emscripten which generates the asm.js code) and in fact is what spawned this work in the first place.
iOS is next, because iOS is ahead-of-time compiled and is the reason they haven't updated Mono (from what I understand, newer version of Mono require that you cannot statically link the library). So after they replace it with pure native C++, it won't even use Mono in the end - and therefore that frees them to update Mono on other platforms.
Then I suppose any platform where Mono is not supported out-of-the-box (anywhere Unity has to write their own Mono port) will get the treatment next, so that Unity doesn't have to spend all of their time maintaining custom JIT compilers.
And then I guess the remaining platforms as a build option, since it does provide some benefits over JIT code.

They've pretty much perfectly prioritized it. WebGL first (it's what spawned this work in the first place), iOS (it's restricting a Mono upgrade), I assume then a Mono upgrade after iOS is out of the way, and then the remaining platforms get IL2CPP support.

 

By the way, I actually know one way you can serve us better.

Take care of this finally: http://issuetracker.unity3d.com/issues/on-mac-os-frame-rate-constantly-spikes
Forum thread about the same issue: http://forum.unity3d.com/threads/216644-60fps-Android-game-without-spike-It-can-t-be-with-Unity

I'm facing same issue in any of my applications made with 3.5.7 as well as 4.3.4. FPS just twitches each second, no matter if application is running on Android device or right in Editor. But in editor, while running on i5 CPU, that spike is not as noticeable as on Android device. But it is still there. It is just annoying, it doesn't allow to feel game running smooth no matter of actual FPS, be it 120 frames per second or just 30.

I've seen questions about similar issue here and there, and looks like no one knows the solution. No explanations, now workarounds, that issue above is seems like it is ignored completely. 9 month already, anyone cares? All that talks about pricing, technologies, new features... Why need them if it just can't run smooth with a single cube in scene?

 

We have support packages aimed at indies, small teams, professional teams and enterprise customers. These are tailored to meet the needs of different teams, who have different budgets to spend on support. It certainly is the case that my team have assisted many customers identify whether problems they have are bugs in their game code or in Unity. It's a pretty equal split. However, of course I know that submitting bugs, even well described bugs that include everything the QA team need in order to replicate the bug, can lead to frustration when these bugs are fixed in a version of the product that may still be in beta and so not actually ship for 3 months. Or, worse, the bug sits around and simply isn't tackled. We've been pretty poor at feeding back to users the status of the bugs. As you may know, the QA and Support teams got a lot closer at the start of this year when I started reporting into ThomasP who handles QA. As a result of this, we have set up a sustained engineering team who will perform back-porting of bug fixes into released versions. We did this for the first time only last week, with 4.3.7p1. As has been said in other threads, this should be the start of a better bug fixing system, and mean customers, whether they have free, Pro, or support contracts start to see bug fixes happening more rapidly. It is not the case that *your* bug will be fixed more quickly, rather, we are looking at the bugs we have, and work out what can be fixed quickly, and impact the most customers. Because of the large backlog of bugs, we cannot fix them all, so our concentrating *for*the*moment* on the bugs that come up in our weekly calls with Enterprise Support customers.

Other support-related changes. Our free support system has started to become untenable as the customer base grows and grows. It can take up to 7 days to get a reply to a customer, and we think that is not acceptable. So, we've started asking customers who need help to post on the forums. My team will spend the time they used to spend answering free support questions on the forum, which means that the answers they give can be seen by everybody. Again, we think this is a win for all of our customers.

In cases where customers encounter bugs in Unity, my team will always try and recommend workarounds.

(We should start a new thread if we want to discuss these things in more detail - this thread has got big enough all by itself!)

 

1) Any chances of C++ in Unity 5. << this will bring all CE4 and UE4 users to Unity5for sure.

When UDK was UnrealScript people chose Cry engine because of performance and c++.

Game developing school teach c++ not c#, since c++ is industry main language.

Thank you, Unity Technologies.

I meant unity 5

 

If by "multiple times" you mean "three, unless you spend hours digging around the forums."

http://blogs.unity3d.com/2012/06/29/the-new-gui/ <- the first major news post about the new GUI
http://blogs.unity3d.com/2012/08/22/unite-12-day-1-summary/ <- a whopping three sentences about how the new GUI totally exists
http://blogs.unity3d.com/2013/08/29/keynote-insights-from-unite-2013/ <- silence until August 2013

The GUI is a point of contention because it's a massive representation at how awful Unity's communication skills have gotten. This is a feature that was initially expected to hit during the 3.x cycle and the only real communication we got about it was seeing the release notes and not seeing it.

 

Nope. No changes in Unity 4, additional cpp-file-support for il2pp enabled platforms in Unity 5.x.
You can read everything in the il2pp thread.

 

I certainly hope we begin to see more activity from Unity support in this area as I do agree these answers will help other users as well. For my part, it seems as if I have been unlucky lately as most of my questions simply go unanswered on the forums.

 

I mean unity 5 sorry

 

just my 2 cents

i was a game developer for about 15 years, starting with the first playstation, later dreamcast, xbox and so on. the last three years i worked on a unity-title, an online game. they shut down last july and i began working as a freelancer and because of the experience i had at this stage i carried on working with unity. i already had a license before, so this was a no brainer. i don`t do any games now, but visualisations. architectural, museums and so on. lots of stuff with oculus rift. a few month ago i tried some webgl-tools like goo, shiva or three.js and i think this is a great field to expand my portfolio (i lost some customers who wanted 3d on their website, but didn't want the plugin). i was really thrilled when unity announced the webgl-export in unity5 and updated my license the moment it was possible. few hours later i read, this would cost me extra and was really angry about this. next day epic announced their price and best of all, including everything i want. as a freelancer the royalties don't apply to me, so this comes on top.
i really like to work with unity, though i had some serious problems with it. lightmapping seems very unclean at times, building an executable gives crashes and i had to set up a new project for just one scene that wouldn't export..
but i always had the feeling, i could get nearly everything done in unity. i would love to stay with unity, but paying another 1500 for webgl when i get the same much cheaper with ue4 (and more..) ?
i am in the middle of a project i will finish with unity. but when i am through with it, i will take some time to look at ue4 and the next project will most likely be done in ue4...
but as i already own the unity5-license, i will use unity parallel to ue4. its not bad to have a choice and time will tell, which tool i will use more.

 

I'm in exactly the same situation here with my team. The thing is those downsides I mentioned. I'd rather deal with not knowing (and planning accordingly) than either a) thinking I know but finding out that it's wrong or unreliable later, or b) having people stick to timelines that result in an inferior product.

 

I think it's key to help out the users of Unity and work backwards, because more resources for us means better games, which means a better reputation for Unity. So it's great to see Unity seeking input from us in this topic.


I have two more suggestions I think people would find incredibly useful:

1) As a way to help users get the information they need, perhaps the Unity Answers site can be modified with some staff answers. Obviously most questions on Answers can be addressed by the loyal community there, but perhaps once a question has gone after x number of days without an answer, or perhaps it can log a certain number of community votes for being answered by staff, it can be put forward to a Unity team. Such a system would add valuable staff input while keeping their role to a minimum and only where it's needed.

2) More built-in assets, techniques, and FX bundled with the engine. Perhaps it will change in Unity 5, but many of the bundled assets haven't changed in a very long time and many of those have become outdated. The new Teleporter demo posted to the blog is a perfect example—it contains many techniques and FX, which could perhaps be refined and bundled as new standard assets (although I know Unity are now moving to 'Sample Assets' instead.)

 

New to Unity development but not new to software development. Mainly in the C/C++ and C# domain. Mostly for hobby currently and it is fun learning about all of the different things I can do.

(1) Platforms:
Main interest is the windows (PC) desktop but having the other platforms available is nice too.

(2) Current Pricing.
(2a) PRO pricing is reasonable and I like having the free version available. If someone asks me what tool I am using I can always point them to the free version. That seems to be unique that Unity allows commercial use with the free version.

(2b) No to any royalties.
(2c) No to subscriptions. Because you never own your own product once you quit subscribing.

(3) Bundling methods./ pricing thoughts

It would be nice to have more platforms included at the PRO price of $1500 or a super bundle for $2500 you get PRO and all mobile platforms including the new WebGL.

Maybe a subscription model that after you have paid the same amount of a subscription as a PRO model ($1500) you own the software as a PRO.

(4) Unity MVP
Microsoft has the MVP program for helpful volunteers (who are not unity employees). Does Unity have something similar? I noticed there are Unity Evangelists but that appears to be only Unity employees.

(5) Mono support
More support for later Microsoft .NET versions like those included in VS 2013. It would be nice to see .NET Native support when Microsoft has finished implementing it.

It would be nice to have Visual Studio integration officially from Unity. I know there are third party option for that but I think an official support would be fantastic.

(6) Asset Store:
This is great. I was thinking about all of the work I would have to do and found some great assets. Most of the decisions on which assets I picked were determined based on having active forum threads on unity.

(7) Asset Store (search) improvements.

I would like to be able to restrict the search to Unity Technologies only assets for example but I do not see anyway to do that. Also the Unity Asset license (says the asset may use the Unity license or their own license) which makes it difficult to figure out licensing. It should be clear what type of license you are dealing with.

It would be nice to have search options like these:

(7a) Checkbox to search only for Unity Technologies assets.
(7b) Checkbox to search for Assets that include full source code
(7c) Checkbox to search for Assets with Unity license
(7d) Checkbox to search for Assets with their own license.

Each asset also seems to be formatted differently when it comes to different things for support and demo. Some assets have demo and support links and unity forum threads at the top (before the description), some have them embedded in the middle, some at the bottom and some scattered all through out the description of the asset. It would be nice to have consistent look.

(8) Source Code Access.

This would be great. Since some of your code is based on mono is that available for viewing already or do you have that under a difference license model since Unity has it commercially?

But since you are making major changes for Unity5 I think you should wait till that is finished before working on source code access.

I think EPIC's source code model for UE4 is very interesting.

(9) 64-bit editor

That is something I really looking forward to in Unity5. I have had some crashes which is related to the 32-bit editor (3GB limit) .

(10) hardware platforms used by Unity for development.

Can you discuss hardware you use for development? It would be nice to see what type of hardware you are using how that compares to a default system.

(11) The Unity input system.

It is serviceable but I would like to see improvements. Like being able to load everything into the input section from a script rather than to have to type it by hand into the editor. For xbox-360 controller I found some code on the CMU website that helped define what I needed to do. I am surprised that there isn't some universal input standard (at least on the windows PC) to handle all of these track-balls (missile command, centipede), spinners (tempest), flight controllers, game controllers (xbox360/ps3), x-arcade, driving controllers, Atari Star Wars (arcade style controllers) etc. I am showing my Atari bias ... -)

(12) Developers in the forum.

Don't slow down development to spend time in forums.
If you had a Unity MVP style program like Microsoft that might help more. With people helping out in their area of expertise. I would prefer developers working on Unity5.

I do appreciate this thread it makes for good reading.

 

There are more example than this, especially if you were following the posts about Unity hiring the creator of NGUI + twitter accounts + forum posts.

No, they didn't communicate it clearly enough (blog post) that it was going to be a long-run development effort due to re-building it but, they did their mea culpa at Unite 2013 (you can watch the presentation) and since then we know it's coming but it's going to take time. They even gave us a hard release version timeline.

I also don't see anyone giving them credit for doing exactly what most of the critics here are requesting: "Do not release half-baked features." That's exactly what they did by going back to the drawing board and making sure the GUI (a critical feature mind you) is solid. The jury is still out until we actually can try it but at least we can assume that the 'previous' uGUI would had been more of a disappointment/painful then the waiting time we endured.

In the end, it all comes back to the same thing; don't plan development relying on unreleased features. If you have to do, make sure you follow closely the forums, gossip and twitter accounts of those in-the-know.

Right now, I see is Unity trying to amend the communication breakdown of the past and it's a good thing.

 

Oh, now I can spend hours digging around the forums AND twitter? Wow! That's some excellent communication skills there!

They didn't give us a hard release version timeline until nearly two years after they announced the new GUI.

Except the two aren't mutually exclusive. You can give release timelines and explain delays while also not half-assing things like mechanim.

Except I do follow these things and we never get even the most vague explanation of things.

 

Okay, this is going no-where.

They admitted they were wrong on the communication blackout. What else do you want from them?

If you feel like this impacted your projects as you decided to depend on an unreleased feature instead of using 3rd party assets that would had done the job right... well that's your fault.

This thread is about finding solutions and expressing what we think they should do to improve their relationship with us. It's not a trial about how badly they acted in the past (we can all agree that they failed).

- Be more open on the roadmap -> Check
- Talk more directly to the community about on-going features -> Check
- Post more blog posts -> Check
- Release bug-fix patches -> Check

So what the hell are we arguing about?
(my troll-o-dector is tingling)

 

What an excellent, 'Yes...And' answer. *Hats off to your communication skills sir*
Gigi

 

No wonder Unity lock up and don't listen. Lets try to keep this with a modicum of professional debate..

The Il2CPP thread was to announce the good work Unity were putting in behind the scenes, people jump on them for it. They try and open up, get feedback and there seems to be a lot of negativity. One of the main attractions to Unity is C# and Javascript and now if you MUST have C++ you can use UE4, the options we have in front of us now have never been better.

Unity gave indie developers a chance to follow a career most could never of dreamed of. Because Unity in some areas is a little in the dark era / rough around the edges right now it essentially dug our project into a hole, I'm not angry I want Unity to follow up on all the messages they are saying and come through in Unity 5.0. I still find Unity at its core the most easy to get along with, flexible / rapid engine I have ever come across.. It's a damn fine engine, spoilt by little blemishes and I would really like to continue on with them as for me and others it's the best chance of success.

Give them one last chance, it's not like we don't have another avenue any more if worse comes to worst.

 

You must be a hit at parties.

 

I'm really looking forward for Unity 5. even thought it was hard to make my team not leaving unity for Unreal after the silence of unity and the delays and those negative sides of unity, I hope, I hope Unity 5 would fill a lot of gasps and earn our team's trust again (as for my self i trust unity). I'm sure as for myself i would never use another Engine as long as Unity is around.

 

What? You were serious?

 

Oh, come on, do i look like I'm serious, hie bros, if any offense was taken, I'm sorry. ok. Let's keep up with the thread.

 

I'm looking forward to Unity 5.0 too, for me it's as simple as this. Epic have put there cards on the table, we like what we see.. Bar a couple of core functions (64-Bit editor, MThreading, Mono / Physx upgrade) the asset store has everything UE4 has covered. There is a cut scene editor, there is a material editor. Unity would be daft not to take over ShaderForge and finish it off, the work is pretty much done for them also it shouldn't be difficult to expand on whatever they have in the animation / sequencer department. It wouldn't harm them to add some upgrades to the terrain system either, then again the world machine / color / splat / height map works pretty well..

Then it's just a case of attacking bugs like there is no tomorrow. Am I oversimplifying this? Because none of that seems over the top difficult, everything else in UT5.0 (Core functions, speedtree, PBR, Post FX etc.) is all covered.

 

@ShadowK I agree that unity should bring ShaderForge to the core, ans the terrain needs serious improvements, but i think if the shader are changed to PBR then that should change the look of the terrain and SpeedTree should do something about the terrain quality too, right?

 

The problem with terrain isn't just how it looks, but other key factors like normal maps not working at all without a lot of hacking, a lack of terrain generation tools (I need to jump between worldmachine on my PC and my Mac each iteration) performance issues, and a complete lack of ability to handle things like caves and overhangs without a lot of work. It's an antiquated mess.

 

We have 24KM2 of terrain on async, the performance in game isn't bad if you're clever with it not saying it doesn't need improvements. You really need to learn how to ring every last drop of performance out of it, a lot of the issues we had were due to main thread latency from trees + colliders / shadows etc. Hopefully with Unity 5.0 this should be rectified (M-Threading / Physx 3.0) and terrain that size usually just makes the editor fall over anyway.

As for normal maps, yeah we had to make our own terrain shader. We extended what came with Marmoset, there is also one in LUX you can use which are far better (look nicer too). Grass required lots of tweaking too, I personally spent about a month just tweaking and improving.

Caves should be on the card, we added just some meshes coming out of the cave and a door leading into it. A lot of it just comes down to general improvements (like most things in Unity), you should be able to add noise / sculpt like you can in UE4. As you said it's a little old school..

SpeedTree is a must, it looks excellent but if you haven't used it before you have to be careful with it.

 

in my opinion the most of all of the quality of the terrain sucks, it needs to be Voxel-Based.... but I thought maybe the PBR would fix the visuals of the terrain. (though normal-maps work for me but it's not enough)

 

UE4 is still plane based and so is CryEngine, there were gorgeous terrains about long before PBR came out.

 

I don't think the terrain should be voxel-based (just stick with heightmaps for now).
What they need to do is fix terrain engine in general.
It's perfectly fine for small levels where you just need one terrain in the scene, but the whole thing falls apart the moment you try to add streaming of any sort.
In my opinion, here's what's wrong with the terrain engine:

* Poorly optimized. It's incredibly slow, even on my high-end rig. It's just impossible to work around ATM and in the end I gave up the game I was working on as a result.
* No multithreading support. What we really need is a way to asynchronously assign terrain data without blocking the main thread - otherwise, open world games just aren't feasible in Unity.
* Trees. God the billboards look awful. There's an asset on the asset store which makes the billboards perspective correct, UT needs to look at implementing it.

Anyway, my two cents of what they should fix.

 

And here's the other big problem with Unity. A lot of the time it feels like they get away with half-assing features because there's always something you can buy from somebody else to fix it.

 

Well I wish I could disagree with that, but I can't.