Search Unity

  1. Megacity Metro Demo now available. Download now.
    Dismiss Notice
  2. Unity support for visionOS is now available. Learn more in our blog post.
    Dismiss Notice

General feedback & other software.

Discussion in 'General Discussion' started by Aurore, Aug 14, 2015.

Thread Status:
Not open for further replies.
  1. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    It looks like Blueprint will be able to generate automatic c++ in around version 4.10.
    No more performance issues. Just as fast as c++, incredible.

    Go to 1:05:50
     
    Tomnnn, Ryiah, Deleted User and 2 others like this.
  2. antislash

    antislash

    Joined:
    Apr 23, 2015
    Posts:
    646
    interesting
     
  3. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    That is huge news. I hope we can see Unity integrating their on secret sauce for visual scripting and material editor. It looks like this is the future of game design and even applications. Clickteam Fusion can do this now with their point and click method. I hope Unity is taking notes here.
     
  4. darkhog

    darkhog

    Joined:
    Dec 4, 2012
    Posts:
    2,218
    I think what Unity needs is better navmesh solution. And by better I mean "with ability to generate navmesh at runtime".
     
    tatoforever likes this.
  5. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    The asset store has been an invaluable source for users and tools and contend creators and very profitable for Unity also.
    What Unity needs to do now is implement some of those great plugins or create their own. I'm talking about taking the best tools from the asset store, similar to how the new menu UI was contracted out.
     
    antislash, ZJP, tedthebug and 2 others like this.
  6. tatoforever

    tatoforever

    Joined:
    Apr 16, 2009
    Posts:
    4,364
    @ShadowK
    I forgot to mention that the solution I'm proposing for a Unity data driven Game Framework can't be forced to users. It should be a sort of core classes that you can quickly derivate from and will be integrated with other core stuff such as serialization, networking, etc but nothing will stop you from creating yours from scratch if you want (instead of being forced to use it in your games) for the simple reason that Unity is used in other areas outside games but games is a huge shunk in Unity and cannot be overlooked. :D
    This UGF (Unity Game Framework) will help tons of people create similar code structures that can be easily shareable, will help build up faster games by having a basic common bullet-prof data driven architecture.
    To give you an idea, take a look at similar game kits/examples in the AssetStore, you'll see how messy and different they are architected in their core libraries.
     
    Ryiah and Deleted User like this.
  7. ZJP

    ZJP

    Joined:
    Jan 22, 2010
    Posts:
    2,649
    This..
     
  8. Tomnnn

    Tomnnn

    Joined:
    May 23, 2013
    Posts:
    4,148
    I was already content with simply writing custom nodes, but code generation sounds pretty awesome.

    You mean like some other engines recently got the ability to do? :p
     
  9. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    You can bet the team at Epic won't stop until Blueprint is as powerful as c++, at least for their engine. I see blueprint one day able to achieve mmo games as efficient as c++. Is got a little while to go, but maybe 6 more months and it could be possible. I don't know c++ and don't care for it as long as I can use visual scripting to make games.

    Stingray is very buggy and they should of waited until, but hey. Unreal is light years ahead of Stingray.
     
    Tomnnn, Ryiah and Deleted User like this.
  10. Deleted User

    Deleted User

    Guest

    Funnily enough I'm an old style coder, as I've gotten older I don't care for it to be complex anymore. I just want to make games, if it's a possibility I'm seriously all for it..
     
    Ryiah and Nateply like this.
  11. GameDevGuy

    GameDevGuy

    Joined:
    Nov 4, 2012
    Posts:
    96
    Regarding crowd-sourcing docs, I would also suggest a few improvements to the Unify Community Wiki:

    1. Tie the login for the wiki to a user's Unity account, rather than them creating an entirely separate account just for the wiki. This will help lower the barrier to entry and potentially speed up contributions.

    2. A visible queue of articles that are out of date, needs cleaning, needs deletion, or just needs some corrections. This would make it easier for community members to pick low hanging fruits for doc needs.

    3. A bounty category/article with a solid template that covers estimates on difficulty, time, art requirements, and specifications.

    Edit:

    4. Some sort of vote system to port the best of the wiki into the official docs/tutorials section.
     
    Last edited: Aug 22, 2015
  12. cyberpunk

    cyberpunk

    Joined:
    Mar 20, 2013
    Posts:
    226
    So I've spent the last few days playing with Autodesk Stingray. I still want to spend more time with it to make a final decision, but I thought some people may like the feedback.

    The interop with the DCC programs is actually pretty cool. I'm using Maya LT and it seemed to work well. You can send assets to Maya, even if you don't have the original files (for example, if you purchased FBX files from a store). Editing is not exactly real-time, since you still have to press the update button, but it's pretty darn close. Keep in mind, it's still transferring FBX behind the scenes, so you don't get any features outside that. You just save the time and UI hassle of exporting, over-writing, re-importing, etc. every time. The camera linking seemed a little wonky, it would update only when you released the Alt key and felt a little choppy. I was sort of expecting a full speed real-time mirror, but it wasn't really that. And it seemed centered on the origin instead of the model I was focused on. Though the idea is there, and I'm sure this could be improved.

    The Stingray interface is clean and simple, but a little too simple. The options and settings for mostly everything felt bare-bones compared to Unity. Clean is nice, but there are a lot of little features missing. Just as an example, the default environment expects a skydome material. It didn't look like you could just set a solid color like in Unity. Materials, strangely, did not have a UV repeat parameter. At first, this seemed like a huge deal to me, but I was told you could edit the standard material and add that parameter. On one hand, it's nice that the engine is flexible enough for me to do this, but I still found it strange that such standard functionality was missing. One thing I did like is how you can preview assets in a little window in the editor, and you can take a picture and use that for the thumbnail.

    The visual scripting system, Flow, is nice but needs work. So there are three different sections a piece of Flow can live on. Either the level flow, which is one per level, the unit flow, which is per unit (like an actor or entity or whatever), and then the external flow (which can be referenced in either level or unit). It's actually pretty flexible, and I liked being able to define my own flow node. You can also create nodes with Lua code, but I found this process to be sort of complex. You have to define the interface in one file, write the code in another, add some glue into another file to load those files, etc. It's not user-friendly at all. I did eventually get it sort of working, but only after hours of trying (and it didn't help searching Google, as there were zero hits). At first I thought you couldn't search for nodes, but you can by pressing Tab and then typing. However it only works on the level flow, and the search is rudimentary. For example, searching for "set rotation" brings up no results, as they force you to type the entire sentence "set local rotation".

    Though the graphics don't look bad, I didn't find them any better than Unity or Unreal, at least with my simple testing art. The lighting, in particular, just doesn't look as good. They do have Beast, for light maps, and that's good. But the real-time lighting did not impress me like w/ Unity or Unreal. I also found some oddities. There were issues with alpha materials and depth sorting. Also, alpha materials did not seem to cast shadows correctly.

    In terms of documentation, there is a decent amount. They have the reference docs, articles, and video tutorials. However, some really basic fundamental topics are not addressed properly (though I guess this could be fixed pretty easily). For example, explaining how to create your own Lua script and linking it together with a Flow node. This is a critical thing and just seemed confusing to me. I mean, there was an article, but it could have been more detailed.

    Also, one of the big features I wanted was real-time editing. They claim this is supported with changing code but I didn't get to try that yet. However, I did have trouble editing the scene while the game was running. For example. I pressed play, moved the popup window over and then changed the scale of an object and it didn't update. This was a little disappointing.

    Right now I am struggling with the camera object. It's not easy like Unity where you just create a camera and it works. I created my camera, then you have to set it as active in code, which I did, but it gives an error. It's a really simple basic thing. Not sure if it's a bug, or some quirky file I didn't edit or whatever, but it's frustrating.

    There are things I didn't try, like making a UI or messing with the cutscene editor and probably a lot more stuff. So certainly don't think I learned everything in 3 days. I just got my feet wet with some of the basics.

    I'll still maybe spend a few more days in my evaluation, but so far I don't think it comes close to Unity or Unreal. There are things I like, especially the connection with Maya LT, but I don't believe that makes up for other things that are missing. It's not that the engine is bad, it surely has potential. Maybe Autodesk can fix the bugs, and add the features, and all that. However, my initial impression is that it needs some more baking in the oven.
     
    Last edited: Aug 22, 2015
  13. Master-Frog

    Master-Frog

    Joined:
    Jun 22, 2015
    Posts:
    2,302
    Anybody want to talk about web publishing with Unity?
     
  14. GameDevGuy

    GameDevGuy

    Joined:
    Nov 4, 2012
    Posts:
    96
    I'm just here to talk about docs. The topic is the least talked about, but just as important to me as any feature. It might be a nice change of pace for Unity staff to discuss. Unless they enjoy reading more posts about PBR, UE4 lighting, visual ___ editor, engine ____ release, etc, ad nauseam.
     
  15. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    Stingray is got a long journey if they want to catch up here. They should focus on developing the visual scripting and less on Lua, which is not really productive right now. I will not be wasting my time trying to learn another engine. Unreal and Unity is enough to keep me busy for a while.
     
    antislash likes this.
  16. Master-Frog

    Master-Frog

    Joined:
    Jun 22, 2015
    Posts:
    2,302
    Well, I'll say something at least. If nobody else cares, I won't bring it up further. From what I gather, WebGL is the answer. However, it takes 20 minutes to build and then crashes the browser. It requires extra work to get it to work... keep that in mind.

    Now, I am looking at simpler tool suites that are out there that export to formats that are actually supported. Granted, these are 2D toolsets... but that's what I've been doing so far, just because it suits my tastes. It will be more work to use something other than Unity, but as I mentioned... looks like I'm going to have to work either way to get past this Npapi issue.

    I would like to ask Unity... what do you recommend I do? Wait? If so, how long? Is the web player a lesser priority? Etc.
     
  17. darkhog

    darkhog

    Joined:
    Dec 4, 2012
    Posts:
    2,218
    If you're not going to make visual shader editor in the near future, at least make MonoDevelop autocomplete shader code.
     
    Tomnnn likes this.
  18. antislash

    antislash

    Joined:
    Apr 23, 2015
    Posts:
    646
    this,
    got my hands on it an disliked too much things, disapointed about its not simplicity ans easy editing... tried terrain system and got very disapointed, for now i prefer to stick with unity, i think AD missed the bullseye this time
     
  19. Tomnnn

    Tomnnn

    Joined:
    May 23, 2013
    Posts:
    4,148
    I wouldn't mind UnityEvents being able to take multiple parameters. If you try to break up "PlaySoundAfterDelay" which has the parameters audioclip and seconds... you're going to have a mess.
     
  20. Tomnnn

    Tomnnn

    Joined:
    May 23, 2013
    Posts:
    4,148
    Well, that's interesting. I usually use moveposition instead of addforce for more control when trying to make a controller with a rigidbody... and I thought moveposition did a sweep to check for passing through objects. With a speed of 10 (sprint), I go right through walls that are the width of a default cube.

    Do I really need to use addforce? I dislike using huge numbers to move a small distance.
     
  21. GameDevGuy

    GameDevGuy

    Joined:
    Nov 4, 2012
    Posts:
    96
    Wrong thread?
     
  22. Tomnnn

    Tomnnn

    Joined:
    May 23, 2013
    Posts:
    4,148
    General Discussion -> General Feedback. Is this the wrong place for anything?

    I'd also like to mention I tried the "don't go through things" method of casting between new and previous positions and I still flew through corners with no collision messages. Very sad.

    --edit

    Set the rigidbody's collision to continuous. Still easily able to run through a wall where corners are. Why :(

    --edit2

    Fixed. I don't want to think about what horror was causing the issue, because I fixed it by making the colliders extend very far into one another at corners. I feel bad for my character controller. The physics system thought his capsule could fit between a 0.1e-39 space where 1 collider ended and another began. That's gotta hurt.

    @GameDevGuy if you're ever confused by a post of mine, please consult either my avatar or my signature.
     
    Last edited: Aug 23, 2015
  23. GameDevGuy

    GameDevGuy

    Joined:
    Nov 4, 2012
    Posts:
    96
    It just seemed like a bug report/game specific issue. Didn't see a solution or request in the content. Seemed like it came out of nowhere.

    Edit - I've actually blocked your avatar from Chrome and don't put a lot of stock in flair.
     
  24. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    It feels to me like AD is branding this engine for architectural more than for game developers. If that is true, then they better button up cause UE4 is already all over it. They need to hire a bunch of programmers and get serious if they want users to take them serious. I wouldn't be surprise if there's only a couple of developers. They need muscle.
     
  25. Tomnnn

    Tomnnn

    Joined:
    May 23, 2013
    Posts:
    4,148
    Aw, that's a shame. You're missing out on quite the sight!

    That's why it's feedback :p

    Still probably true.

    No kidding, have you seen the unreal tour of paris? Never seen anything like it.
     
    GameDevGuy likes this.
  26. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    Yeah, is as real as it gets.
    If you take a peek at last GDC, there was talk about Wetta using Unreal for a vr demo from The Hobbit "Smaug".
    You will quickly realized where the industry is moving right away. Is not just VR, but vr meets Hollywood. Wetta and ILM will probably be one of the first pioneers here.
    I envision a future where game designers are making interactive movies, not in a fancy studio, but out of their garage. Utilizing mocap data, facerig pro, motion capture, vr hologram and whatever tools available in the arsenal. What ever make the work easier. UE4 would most likely be one of those tools, given their track record. UE3 has already been used in tv shows. Example take that Paris scene and turn that into a horror game. I think I would S*** my pants and could not finish the game.
     
    Last edited: Aug 23, 2015
  27. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
  28. Deleted User

    Deleted User

    Guest

    Not any time soon, just because the engine is capable of it. Doesn't mean it's practical in a game sense, it's coming though...

    Thing is, at least you have the option.. If you're aiming for a three year release cycle, at least you don't have to worry about UE looking out of date or taking advantage of future hardware.
     
  29. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    I'm 45 and been around gaming since colecovision, atari. My fisrt computer was a comodore64.
    I hope it doesn't take too long since this is really the next step of game development. I just want to participate and hopefully make money along the way. I'm not a coder, but when it comes to game concept I amaze myself. Still grinding away with Unreal and Unity. At this point I would like to lean torus unreal more, but was kind of turn off by the lack of performance on mobile. Is getting better and the file size seams to shrink every release. We all have different agendas and preferences. That's why their are so many game engines out there. There's no magic bullet, yet.
     
  30. Tomnnn

    Tomnnn

    Joined:
    May 23, 2013
    Posts:
    4,148
    Oh, you mean this?
     
  31. Deleted User

    Deleted User

    Guest

    Agreed, whilst I've said on a fair few occasions Unreal is harder to use (and it is).. I did manage to get through a character controller, weap selection, full animation state tree, muzzle flash, decal, projectile spawn, camera (including collision), decal setup in three days..

    I'm far from a master as well, plus the rendering side of things was pretty much simple to sort out once I knew how.. I spent weeks with CryEngine completely confused..
     
  32. Moonjump

    Moonjump

    Joined:
    Apr 15, 2010
    Posts:
    2,572
    Mobile build size has increased over the years for Unity, and then took a big jump for the 64-bit iOS requirement. IL2CPP has improved from where it was, and hopefully will continue to do so, but reducing build size further tops any new engine features added for me. Too much time is spent on optimisation to keep sizes reasonable. Hopefully support for Apple's App Thinning technologies will come sooner rather than later.
     
  33. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    Not sure if you guys have tried 2d game engines like clickteam, stencyl and construct 2 to name a few. I like messing around with these while practicing game logic. The file footprint is really low compare to UE4 which is hard to go under 50 megs. I which they had a simple switch to go from Gorilla to chimp when publishing mobile games.
     
  34. zombiegorilla

    zombiegorilla

    Moderator

    Joined:
    May 8, 2012
    Posts:
    9,042
    ILM has been working on this stuff for a while, as has Disney. It has been used a bit, and more is coming.
     
  35. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    Yeah, soon in the not so distance future, we will be the actors making movies and everything else. We are doing that right now when we make a game, we are telling a story. Some are simple and others require a dream team of artist and programmers. Think about making a game like GI Joe the movie and using tools that are so fluent and agile is like a composer with a wand.
     
  36. S-0-L-0

    S-0-L-0

    Joined:
    Nov 9, 2012
    Posts:
    163
    Source Control for Unreal Engine is going to be a lot simpler to work with soon:

     
    Ryiah likes this.
  37. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    This is great.
    Now the cons is that blueprints becomes quickly spaghetti nodes and becoming complex to read when some line of C# would be more fast and simple to read (i know UE4 is not C#).

    It's like comparing all bad generic DAZ characters all looking very uniform using same uniform textures with hand crafted characters and textures made by veterans 3d artists.
    You' ll get lot of bland generic GI Joe movies or games ; and you'll see on the other side amazing hand crafted games like The Witcher 3 :rolleyes:

    It seems Stingray is no more the main subject lol
     
    Last edited: Aug 23, 2015
  38. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    Perhaps at some point we will need to turn back to more basic 3d engines for mobile like Godot and some other free ones.
     
  39. Tomnnn

    Tomnnn

    Joined:
    May 23, 2013
    Posts:
    4,148
    Is it bad on big projects even if people create functions and custom nodes so there aren't a million connections all over the screen? I wonder how code generation will work with custom nodes... will it just give you your code back? lol
     
  40. Deleted User

    Deleted User

    Guest

    No you create different classes and functions that you reference, just like you would in C#.. So you can cut it up whichever way you like.
     
  41. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    Given the speed at which Epic is moving, it wont be long before blueprint gets tamed down and the spaghetti factor ironed out. If you look at the whole video I posted, they talk about many improvements besides the big one. Blueprint is the closest some of us will probably ever get to real programming. I'm talking for the rest of us that are new to coding and want to dive quickly making games, artist and non technical minded. Some folks can never fully understand programming, but are excellent at story creation, game play ideas or awesome artist.
    Anyways they are closing in the gap between c++ and blueprint. The blueprint system is very powerful and it empowers now programmers to be programmers. You're really programming, is just not in the classical way.
    By version 5 their will be no difference in performance and more intuitive way of working or prototyping games and don't forget fun.

    The engine is not free, they have to eat also, but the terms are very clear. You succeed we succeed, plain and simple.
    This is why the engine is free with a nice little kickback when you publish and make money. Apple takes what like 30% for what, a store display. Talk about highway robbery.

    Epic will polish the engine until the end time, why? They want us publishing with their engine, simple. They cannot afford to be anything, but EPIC.

    A win win for us the game designers, architects and possibly Hollywood.
    Unity and Epic has made life really easy for all of us and is only been a little more than a year ago since all this got started. The future looks very promising for the people that want to invest in them self and continue education. I wish I was 16 years old again. Is going to be a fun ride for those that love making games, more now then ever.

    edit--
    I'm currently working on an awesome game which is being prototype with construct 2. Yes is another 2d style Platformer, but the game play, concept is something I never seen in my 30 years of gaming. I hyperventilate while adding more mix to the game. I wish I could say more, but at least you know where you first heard it.
    btw. I recommend construct 2 for anyone thinking about prototyping 2d games. You could publish with construct, but it depend how big the game gets.
     
    Last edited: Aug 23, 2015
  42. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    They are making Gears of Wars 4 also or Fornite and future games, this is why they keep improving,

    Each tool has pros and cons , UE4 has also cons like heavy machine required or games not running fast on medium machines, or no scripting language for intermediate skilled programmers that don't want C++. or for some indies it can be a too complex powerfull tool and engine when they target non realistic games.
    Their marketplace prices are really too high , and they lack all Unity great assets editor extensions or other cheap models.

    It would be cool to find playable demos games made by people using stingray or having feedback on future plans for Stingray.
     
    Last edited: Aug 23, 2015
  43. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    I would love to see something running by them on the android store. If the engine can put out a nice looking game running on android, then it should be given a closer look. So far is only been a let down. I mean if you're in the game engine business and you don't have proven game like unreal tournament for example. This might leave a bad taste in some people. Stingray is only version 1, but given the power of AD they should of came out swinging hard. We have tools like Unity and UE4 and are very spoiled at this point in time. You better take that dog food back and figure out the proper way of presenting to a sophisticated audience.
     
  44. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    For Android you have already many choice such as Unity, Shiva, Godot etc ...
    The only interest is perhaps for Maya users.

    I am not convinced until i see solid PC games made with it running faster on medium hardware than UE4.
    About future plans from Autodesk , nothing is so sure for Stingray, i think they will not push it a lot.
    They don't have a big team like Epic games pushing the releases with such big features each time.
    But let's wit some months to see how things will evolve in what direction ?

    They are also releasing their engine for beta testing, early users reporting bugs and general problems.
    This is why even 30$ for beta testing on a non prooven 3d engine is really too much :rolleyes:
     
  45. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    I like to think about it like this. The tools are getting more and more automated, powerful and easier. Take substance painter. Maximo+substance painter/texture. I think it would be more about the story the idea and not the tool. The tool should never hinder the artist in any way. Is all about empowering even people that don't know computers that well. If you have a strong imagination and want to tell your version of the universe, go right ahead.
     
  46. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    They most likely don't and this engine is probably more for architectural folks using their other tools. It just have the buzz word "game engine" to wow the uninitiated. If they want to run with the big dogs then they need to put some muscle in the R&D department and quick.
     
  47. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    We already have thousand examples of examples of 2d engines for people to express themselves and it results on lot of garbage games. Or lot of 3d garbage indie games with generic characters.
    You must have some real modeling and texturing talent to give a personnal interesting style to your game otherwise it is generic and not interesting and it stays bland.

    A big example is RPG maker, i really do not alike all these generic games , the best 2d RPG games are crafted by talented people not automated.

    You need to take a look at hand crafted characters or take a look at some games to understand textures or level design and characters or effects are successfull because they were created by hand not automated.
    The Push button : "Make my game" is just a dream.
    But "Make my generic game" . yes this exists in some ways and develop and we have more generic bad games :D

    do you really believe they will bring a team as big as Epic lol
    and as talented :rolleyes:
     
    Tomnnn likes this.
  48. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    YES, I DO,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, just kidding.;)

    Is really too bad is really a nice shell they got there. Looks like what would happen if Unity and UE4 had a child. No guts no glory.
     
  49. Master-Frog

    Master-Frog

    Joined:
    Jun 22, 2015
    Posts:
    2,302
  50. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    very kind of you. Is not as fun as here.
     
    superpig, Ryiah and Master-Frog like this.
Thread Status:
Not open for further replies.