General feedback & other software.

Ok, tried out some stingray, these are my thoughts:

• I really like how easy it was to start up a template project. Unity should learn from this and offer several template projects in the start screen.
• Instead of the offering a preinitialized environment, you have to initialize it yourself (in code). That's kinda annoying.
• Adding a camera to an empty scene does not make it render the scene - you have to hook it up in an initialization script.
• You can only see local variables in the debugger. No watches.
• The intelli-sense is fake. Suggests all methods/properties for whatever symbol (existing or not).
• Very verbose:
  Game.camera_unit = Game.world.unit_by_name(Game.world, "camera")
  I haven't seen syntax like that since C days.
• A bit buggy regarding a lot of things, but that's to be expected of a first release so I won't dive into that.
  
• You cannot parent or unparent things in the editor.
• Does not seem to be component based regarding scripts / flow (visual scripting). Instead of creating behaviors that are added to any object, you define 1 Flow (visual scripting) for each model, or you can have a global script handling the behavior.
  
• Overall, the feeling I'm getting looking at their samples is just spaghetti:

Code (CSharp):

1. function Player.spawn_player(level, view_position, view_rotation)
2.     if not level then
3.         print "ERROR: No current level - cannot spawn"
4.         return
5.     end
6.  
7.     local player = Player()
8.  
9.     spawn_freecam_player(player, level, view_position, view_rotation)
10.     PlayerHud:init_hud()
11.     play_spawn_sound()
12.     enable_walk_mode(player)
13.     play_ambient_sound()
14.  
15.     Appkit.manage_level_object(level, Player, player)
16. end
17.  
18. local function spawn_player_character(self, pose)
19.     local world = SimpleProject.world
20.     local land_unit = World.spawn_unit(world, "content/models/character/skm_chr_genrig_3p", pose)
21.  
22.     local land_character = self.land_character
23.     land_character.unit = land_unit
24.  
25.     -- Add Character input movement.
26.     local controller = UnitController(land_character, land_unit, Appkit.input_mapper)
27.     controller:set_move_speed(character_move_speed:unbox())
28.     controller:set_yaw_speed(character_cam_yaw_speed)
29.     controller:set_pitch_speed(character_cam_pitch_speed)
30.     controller:set_mover("body")
31.     controller:set_gravity_enabled(true)
32. end

Might be a nice addition to the engine market, but it doesn't really interest me. It doesn't do anything better than any other engine afaik.

 

I tested UE on a bunch of hardware including PS4, which let's face it in terms of hardware on the market the PS4 is pretty poor.. I also got our game running on a GTX 280 / Radeon 4850 @ 30 FPS 1080P.. But you have to be very careful about every decision you make when you scale it down. There's INI locks for platforms you can extend to reduce texture sizes on the fly for specified platforms.

A 280 also isn't very powerful, but it will scale. If you're solely aiming low (like Intel HD up) then of course Unreal is simply a no go, it's putting yourself through additional hardship for no real reason.

Of course, end of the day it comes down to platform / type of game / target market and developers preference. If you have a pipeline set up in Unity, I don't see the sense in moving to Stingray. Unreal well that's another matter, but again it all depends on what you're trying to achieve.

@hippocoder

Yeah I agree, multi-scene seems pretty sweet.. Although it would be nice if more dynamic workflows (like navmesh) could take advantage. I still think they should implement Apex, they used it in the butterfly demo.. I wonder why they left it out?

@sunnyshine

There's object references per component script that do different things, so it kind of is.. I just checked out the Unit editor and you are applying VS to that specific object.

It's not like you just have one massive "game" .CPP file with your central loop in it, it's split up.

But I do agree, even in Unreal's BP editor. You can achieve more in less steps, even the Visual scripting tools seem a little bit "verbose".. Plus it does kind of just flow into each other quite a bit. I wouldn't say it's spaghetti, but it kind of feels like a lot like you're hand holding every part of it.

I think Unity just abstracts a lot of repetition from the end user, when I developed my own engine every new class had to have references back to the central loop to be initialised. Even though I had object picking available, I was constantly creating references to other classes (like tick (timeclass)) / (math e.g. Vector3 class) in the ref for every part of it.

Then again, I wasn't trying to make it user friendly ..

 

Yes. Much the same way that if you licensed all of Unity's middle-ware yourself the total price tag would come out much higher than the $1,500 cost of Unity Pro.

 

I fear that the majority of our players will be laptop users, and not hardcore gamer laptops so yeah, this is important. Not our only target but will be a big part.

 

You don't have to spend as much as you used to for a gaming-capable laptop. The MSI Leopard costs ~$800 and has a GeForce 950M in addition to a Core i5 and 8GB of memory. Laptops have really come down in cost.

http://www.amazon.com/MSI-Computer-15-6-Inch-GP60-LEOPARD-1053/dp/B00YQNS62Q/

 

Yeah, exactly. But it seems like it's just for that object. You wouldn't be able to have a generic flow and apply it to several objects, or have several flows on a single object. Just 1 per object afaik.

 

Yep you'd pretty much just stuff it all in one place it seems, which I agree could get pretty messy.. It's a little old school compared to Unity..

P.S I'd be VERY interested in editor extensibility.. Engines can't cater for every tool out there, it's impossible. Being honest I can't be bothered to find out, but it would be interesting to know.

 

The vast majority of indie games never break the $1k mark. And it is a safe bet that if you knocked it out in 3 months its not making a dime. Thousands of others are doing the same thing. $4500 in a year for a game made in 3 months isn't a minimum, it is actually a huge success, even though you could make more working at McDonalds.

 

That is true, my son has one. However, if we are going to make a niche game and target that particular audience, it seems unlikely most of them are going to replace their laptops for our game. I would guess our niche is not the usual gamers that care about having a powerful machine.

One thing I love about Unity is that we can make a game that will run on a computer that doesn't use DX 11.

 

@Daydreamer66

Why would the Stingray Wwise licence be any different than Cryengine? What exactly is their "core" licence?

 

As far as I can tell, the Core License is simply their standard license that lets you use WWise in your game. This means there will be no additional licensing fees when using their product in Stingray. Other licensing options available on their site include expansion packs, add-ons, support, etc.

As to why, I imagine Autodesk offered them a sweet deal. Or perhaps Audiokinetic simply wanted their tech in more hands. You'd have to ask the players involved.

 

Probaby because Autodesk is a multi-billion dollar company.! They already chuck expensive gameware everywhere, what's Wwise to them?

Anywho, I feel like this conversation has been very fruitful.. Got a lot out of it, very interesting points and @S.0.L.0. thanks for the tip again.!

Now my intrigue has been satisfied, it's BAU!.

 

Hello forums. I am seeking general feedback for my avatar updated based on @Ryiah 's suggestion.

 

Whether I like it or not depends on which software you used to create it. ;-)

 

~3 seconds in Gimp. Even upped the quality of the base image. But now there's a little pepe in there too

 

I like Gimp, but it isn't a AAA quality avatar. You should have used Photoshop. Haha. I kid. I kid. It looks like James Franco said something mean to the frog and is gloating about it.

 

It looks kinda like james franco has a Pepe sock puppet <3

 

tabs (type unreal) and better/review use interface
social content

better launcher

 

Okay, I don't get Stingray. I really, truly don't. What's their angle? What's their appeal?

Best I can tell is that it will be integrated with other Autodesk tools. And, well, that's about it. Everything else they show off in their promos are just standard practice at this point (such as global illumination).

However, at the same time I can't help but to notice that it has a staggering amount of things Unity has going for it. Getting the obvious ones out of the way: there is a current lack of support and a still new community. Other than that, the engine does not seem to allow editor extensions which is a MASSIVE feature that make Unity so great.

Then there's their asset store, which seems to be a gaggle of random art assets and a few example projects. And the store is not well implemented at all. There's very little navigation and only what they want you to see shoved front and center. If you want to look for other things, well good luck, because the store is encompassing of all of autodesk and you will no doubt have to shuffle through piles and piles of irrelevant content.

Lastly there's the system requirements. That's like 3x more than what Unity runs on.

So can anyone explain to me why anyone cares about this?

 

Just another engine into the mix. Another option. Use it or not. Totally up to you.
Gives you another option when looking at what to use and how you are using it.

 

Everyone knows unity's editor has low system requirements though. My 2012 laptop can run that and unreal as well. The fan turns on if I try to do unreal + anything else, though.

 

We are working on a Playable Graphs API (first experimental release in 5.2) which allows scripting of tree-based systems. From our Public Roadmap:
"This new C# API aims to provide a way for users to create tools, effects or other gameplay mechanism by organizing and evaluating things in a tree-like structure.

In its current form, the Playable API supports creating Animation blending graphs at runtime, providing the capacity to have full control over the AnimationClips played by the engine"

This is a step towards easily creating tree/node-based systems; for both ourselves and the community(And we have a few listed on our public roadmap, such as Visual Scripting: https://unity3d.com/unity/roadmap)

No current plans for this.

We are overhauling our terrain system; but no specific information yet.

Thats for Unity 36.4.5

No current plans for this.

 

The only thing killing me about unity right now is my unreasonable demand for a grid based multiplayer game without lag. When your movements are grid based, the slightest desync causes a huge difference on every machine. Interpolation is fine for physics based movement, you'd never notice the small slide on the other clients... but that's not the case for grids.

 

Oh man, again engine wars . Ok so far I have looked, stingray looks like an overpriced scaled down UE4 with exclusive live link feature to me.

And honestly I can't realize how you people get the time to evaluate monster like engines such as UE4, CE. Project after project after project.....going on like an infinite series. Things are going so fast that I can not grasp UT fully, let alone others. May be you all are god gifted or may be sheer amount of time is being wasted for comparison, rather than actually "developing" anything.

 

mmm maybe the wysiwyg editors make people too optimistic

 

They just do this for a living or are students. Working people just have enough time to read things let alone try things

 

I dont even have enough time to read UE4 docs. After minimum 4-5 weeks as far as I can remember, only tomorrow I will have a free weekend to do something out of routine. Too much in the list to do.

 

Will it be possible to animate a character with the experimental release using the "play graph trees"? What is the release cycle of Unity, ie. when is Unity 6 expected? It seems like visual scripting is going to be one of the main features of 6, also the networking v2 of 3, what else did they mention at Unite? New input system?

 

Everyone asked for a public roadmap, so we released one to answer most of your questions!
https://unity3d.com/unity/roadmap

Absolutely zero information on Unity 6.

 

I haven't played with the new playable graph trees yet but I assume you probably could based on this bit that I posted earlier:

 

Or we simply allow ShadowK to do the quick evaluation for us to determine if we should waste our time or not.

 

I'm not sure why some people are calling Stingray over priced. It costs $30/month, Unity pro with IOS and Android is $225/month. Not to mention you will be able to get both Stingray and Maya LT together for $30/month. Also, UE4 charges royalties. If you make a game that makes any real money Stingray looks like the cheapest option by far.

 

The only thing being synced is a Vector3 which determines which direction the player is moving in (in 1 of 4 directions). If the player bodies get too far apart on the client and server (right now checking for a distance of 0.25f) then positions get synced once.

It's a realtime pacman othello. I wish I could do something as simple as interpolate locally but players can collide so everything needs to be in the right place 100% of the time, or no more than 0.25 units apart for collisions to seem fair.

Thanks for responding, the issue is mostly dealt with at this time. I just hope I don't have to sync grid based movements in realtime in the future, it's a pain.

 

mm...basic hating maybe ?

 

It is even cheaper if you pay yearly. Only $240/yr or $20/mo.

 

I saw that. It is pretty amazing when you think that you will be able to get a Maya LT license that includes a Stingray license for that $240/year which comes out to $20/month. You could look at that as $10/month for Maya LT and $10/month for Stingray. As much as I am not an Autodesk fan, I think they have put together a pretty compelling deal.

 

For me, the ability to extend Unity's editor has become pretty indispensable, and definitely something that I consider to be one of Unity's strongest features. In "The Deep Paths" (current dungeon crawler project) I have panels that I can paste CSV dungeon maps into that then build my levels for me (sure I wrote the code and that took time, but making sequels to this game now will be a piece of cake). I'm literally making levels in Excel and just copying and pasting them across - it's AWESOME.

I've always been curious as to whether or not UE has something similar that's easily accessible?

 

@ShadowK
I didn't switched to UE4 and will probably won't in the foreseeable future. Not that don't like UE, it's really awesome and I have some prior experience with (i was part of the alpha team and was working with UE4 more than a year before release). We have two problems here, we've invested a lot of time in Unity, we've created a framework called FOG Engine. Build on top of Unity, it contains custom asset/pipeline tools, shaders, replaced renderers, post-procesors, systems mechanics, even a visual coding editor that runs at blazing speed and looks nice even on mobile. To give you an idea, take a look at the trailer attached to my signature, it was recorded on actual iOS devices, no edditing, no retouch, you can also look for Forgotten Memories walkthroughs on youtube, you'll see them all running on iOS with multiple light sources casting shadows. F.O.G scales (thanks to Unity) from mobile all way up to desktop and consoles (Forgotten Memories is now remastered for VITA/PS3/PS4).
Long story short, the second reason we can't (at least for now) is that it will take huge amount of time to train my employees to get up to speed with UE art/design pipeline tools. They already master Unity, they know every single little bit of it, they rarely call me for help, in fact when they do is nothing about Unity, instead something to do with our internal tools (F.O.G related stuff).
Unity is not perfect, we all know that and like a lot other users here we wish we can have more "out of the box features", the more we have those the faster iteration and production of game in Unity will be.
What I'm really wating for are features that will help me create large/big/rich streamable worlds, like the ones UE currently have. After that, I will become a hardcore die-hard fan of Unity forever!

 

damn yes, ..
i'm actually wondering, if i 'm going to get RTP or code my own terrain system and shaders... and that will take time.
just that very feature should come out of the box and allow to jump directly in dev.

 

Basically this is what i'm also waiting for unity, "More Built In tools and features"

 

@Andy Touch Will Mecanim be using the new graphs API?

More specifically, will we finally see zoom functionality in Mecanim's Animator panel? The inability to zoom out on large state graphs is easily Mecanim's most frustrating aspect.

 

Stingray is out now, i can just say : Next.
It has a long road before becoming solid and prooven on the game side. A serious limitations as no plugin is just a big no, even UE4 allowed you to create editor plugins from source or using blueprints.

I could not say better. I also hope to see better out of the box graphics because a lot must be baught on the store once you target great graphics ( better shadows , better PBR ; better full screen effects etc ..)

 

That's cool and I agree, it's always six of one or half a dozen of the other. Like Ark (survival), we used the UE quake style demo as a template.

But unlike Ark, we stripped all the .DLL's out removed all references on the SLN stripped out the re-cycle BP's and set it up the way we wanted it so we can have level designers work with BP's as well as us adding some of our own backend stuff.

At that point you really realise how simple Unity is, BP's are cool but if you're going to do something that warrants a full scale game then it's not very "indie friendly". The pawn system is confusing, it has really stupid quirks like setting a mesh as a variable doesn't work in a PC but adding a mesh (reference) to a PC does work (like a spawn). Which actually worked fine as a variable in previous versions??!

A bit of it is similar like adding a multi-function BP to a PC Actor or equivalent (so like multiple scripts referencing to self on a GO).

Whilst Unreal is flexible and powerful, has great rendering / particles and tools coming out the wazoo. It also alienates a lot of people, it's not easy to use in the slightest. Even for basic things there's 20 options for various things in different sub-menu's that are over-written somewhere else.. Which is VERY annoying.

So as you say, everything has it's pro's / cons.

 

Like any visual scripting it will add lot of overload, with multiple new variables and methods compared to your 3 lines of code , i would not recommend it also for large scale games only for small sized games.

It is sad they did not get rid of all FPS centric framework to propose a more universal one like gameObjects.

I find Unity simple , it does not deliver as great graphics ,tools, terrain, vegetation as UE4 or does not scale as well for open world ... but like many small indies i am not making a Skyrim like game scale and i am not targetting the best graphics for current projects, so UT5 is the best choice and i know it will run well on old or medium hardware, and i can make a multi version of a project like mobile and desktop both running good.

 

Yes they did it, the Unreal GameFramework now is a general purpose game framework. It's just a collection of basic derived classes (Object, Actor, Pawn, Controllers, etc) really small in fact:

You can see that as an evolution of the MonoBehavior class but with more organized data driven structure.
I've been asking this for a very long time in Unity. Unity really need it's own gameplay framework, it will help a lot of people organize their code. I won't say that all games will need this, but the majority of games will do. Anything that have controllable player(s) and/or AI(s) characters = sure they do!
Where's the UNITY GAME FRAMEWORK!?

 

We see a lot non character games made with Unity also, they just let us make our own character game needs and proposing basic working examples.

 

There's a bit more to it than that , I don't believe a typical GP framework makes much more sense if Unity had a proper extensible system. In which I mean like a code derivative AI BP State Tree system, on which you could modulate and create your own frameworks to do whatever the hell you wanted.. It's not that different conceptually from Mecanim, why not just expand it into a proper system?

A proper navigation solution you could #include into your script components and control, YES I know you control navigation for AI components via scripts. But it's not what I'm talking about, a proper extensible system which would support dynamic and extendable boundaries in which you code and select how it should all come together (radius calculated boundaries per async load on runtime mesh for e.g.). Which would also derive from the AI BP State machine, because that's what you want really! A capable all-in-one Nav / AI system..! Get the UI to trigger off AI ST actions etc. etc. Make it all derivative (to a certain extent), so it all kind of makes sense.

In summary I'm talking about an elegant solution to simply set up like what you're saying Tato a proper game framework that you control as opposed to one being dictated for you. Which if Unity went this direction, it would be much preferable to UE.

You do kind of get the feeling, UE designed there system to make games though because they make games.

In the end though, I get why Unity is popular.. I believe many indies will stick with it for good reason.