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

Unity Visual Scripting 2017?

Discussion in 'General Discussion' started by Tomasz_Pasterski, Mar 21, 2017.

Thread Status:
Not open for further replies.
  1. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    I'd still like to know what's wrong with the stuff already out there ...
     
  2. eskovas

    eskovas

    Joined:
    Dec 2, 2009
    Posts:
    1,373
    Don't forget about TerraFirma by @Murpenstien: https://forum.unity3d.com/threads/w...ne-replacement-and-extraction-toolset.448766/
    It's a complete terrain engine replacement that is looking really good and seems to be orders of magnitude better than Unity's terrain engine.
     
  3. gian-reto-alig

    gian-reto-alig

    Joined:
    Apr 30, 2013
    Posts:
    756
    Looks really good!

    Hey, maybe Unity waits for the opportunity to solve their terrain issue the same way they did with Textmesh: wait for a thirdparty dev to produce a high quality asset, employ the guy and buy his asset, make it a part of Unity without having to pay for the dev time fully.

    Could also give them a VS feature, seeing just how many people are working on VS Systems in the asset store.
     
    Lars-Steenhoff likes this.
  4. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521

    True the asset store has provided visual scripting because unity is lacking in this department, however the nice thing if its integrated in the engine is that all the scripts that are made for it are in a standard that has support for a long time in unity.

    So there is less risk for store packages to become deprecated or not updated for new unity versions.

    It will create a nice new community for sharing ready made scripts that interface in an easy way in the editor.

    Just unity needs to make sure they get it right, don't do it half baked and make sure it doesn't create unnecessary clicks.
     
  5. MV10

    MV10

    Joined:
    Nov 6, 2015
    Posts:
    1,889
    Quoting myself because I just read, apparently, the Blueprint / C++ codegen is an experimental feature. Regular Blueprint really does carry a 10X performance penalty.

    I had a thought last night ... creating a solid, reliable IL2CPP might be viewed internally as a necessary precursor to Unity visual programming (can I coin the acronym UVP?)... It all depends on what Unity intends to target under the hood. It may actually be easier to target IL which might make visual programming more interesting to real programmers, since IL can do things that C# cannot (admittedly, mostly edge case stuff), and writing in IL is nearly as painful as traditional assembly.

    It still doesn't make me think it's a worthwhile investment, I was just pondering whether there might be anything about it that I could find potentially interesting.
     
  6. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    I tend to think this is more of a "marketing coup". Unreal was going it at for a several years, and there are still glaring oversights in the technology. It looks like a massive waste of resources to me, and one baffling thing about it is that blueprint obsession started after epic games decided to get rid of unreal script. So... "We don't want to write everything twice (meanign bindings for scritps, etc), so we'll remove scripts. Then we'll spend several years making our own visual scripting language so we will have write everything twice again". This... doesn't make much sense. Still... it is their money and their project.

    I also think that everybody can code traditionally, to a degree.

    It does, but this place is highly parallel data processing and image manipulation and VERY high level things like AI decision trees. Not game logic - which is what people usually try to use it for. Usually people get distracted by visual tools and start trying to make a very low level stuff in them, down arithmetic operation level and loops. That is a waste of time, in my opinion. Past certain threshold text is much more expressive/compact.

    Yep. Basically, blueprint to C++ converter is something that Unreal team promised .... hmmm.... last year? And there were bunch of misguided people who thought that this will apparently allow them to write C++ programs.

    Blueprint node is essentially a class derived from an abstract base, and since you can't just grab function contents from C++ code, you can't easily turn this into a codegen. I suppose the codegen will rely on macros magic, otherwise they'll have to write blueprint node content twice.

    I think you're off by a mile on this one, and there's no direct connection between IL code and Visual "programming".
     
  7. elbows

    elbows

    Joined:
    Nov 28, 2009
    Posts:
    2,502
    They didnt mention many Unity 2017 features at the GDC keynote but they mentioned Visual Scripting - I really doubt they would have done that if they were not getting anywhere near to releasing it. As I've said before,the roadmap is out of date because it doesnt reflect any of the 2017 features at the moment.
     
    Ryiah likes this.
  8. tswalk

    tswalk

    Joined:
    Jul 27, 2013
    Posts:
    1,109
    are you certain about this?.. I personally haven't tried the engine yet, but have been watching plenty of the GDC demos from them and several are using blueprints.. the performance looked great.
     
  9. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    Here you go:
    https://forums.unrealengine.com/showthread.php?1105-Blueprint-Performance-Benchmark
    https://forums.unrealengine.com/showthread.php?117332-Just-how-slow-are-blueprints
    https://forums.unrealengine.com/sho...hursday-April-17&p=19464&viewfull=1#post19464

    Rule of the thumb: expect them to be 10x times slower, that's according to engine developer.

    Then there are people claiming 10x..250x slowdown. Well, this is definitely possible, if you start doing a lot of simple arithmetics in a loop where overhead of calling a blueprint function will be much higher than cost of operation itself.

    It is worth keeping in mind that gdc demos normally wouldn't be useing bazillions of active NPCs so their performance won't be affected by blueprints much.
     
    Ryiah likes this.
  10. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    According to the video in the first post there wont be a performance difference, so let's be optimistic and see how unity executes this

    it could just compile to optimsed code, even to assembly if needed :)
     
    Last edited: Mar 23, 2017
  11. elbows

    elbows

    Joined:
    Nov 28, 2009
    Posts:
    2,502
    Ten years ago I would have been part of the target market for visual scripting. I'm really not anymore, but that does not mean I agree with people in this thread who think visual scripting is a waste of effort on Unitys part. Or that performance implications of specific implementations in other engines make the exercise pointless, even if the same sort of performance radio's ended up applying to Unity (which is not a safe assumption).

    There are a bunch of scenarios, pros and cons and reasonable differences of opinion. Unity have a number of ways of judging demand, including how well visual scripting type assets have sold in the store over the years. I cant see vocal people who think its a waste of effort having much impact compared to those numbers, which will speak for themselves.

    I suppose that although I am not going to use visual scripting directly myself, I do have an interest in this stuff being included in Unity so I can potentially expose applicable areas of my systems to rewiring by those who dont like written code. There are some very useful scenarios here and I believe some posters in this thread already touched on them better than I can manage right now.
     
  12. thelebaron

    thelebaron

    Joined:
    Jun 2, 2013
    Posts:
    851
    @neginfinity
    Indeed, such a waste of time and marketing, that top tier companies have them in their engines(Frostbyte, Snowdrop, CryEngine, Decima, Stingray, Redengine), not to mention their occurrence in industry standard programs (3ds max, Houdini, Substance).
    I'm assuming you also hold programmers to the same standard(If everyone can program, everyone can also make art)?
     
  13. tswalk

    tswalk

    Joined:
    Jul 27, 2013
    Posts:
    1,109
    I see, I get it... I'm just thinking at what point, or at what scale would such a thing make a difference that requires going to a native code solution. If one was able to iterate quickly with a visual scripting solution, perhaps then a determination could be had to focus on optimizations.

    It seemed like (atleast in Epics case) that blueprint could utilize custom native modules (?) allowing one to maybe do this streamlining when optimizations are necessary?.. i'm just guessing, I honestly don't know.
     
  14. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,965
    You definitely are able to write custom Blueprint nodes in C++. You can do the same thing with PlayMaker too. Being able to easily create your own nodes is part of why many other assets integrate with PlayMaker and I'd expect Unity's to support it.
     
    tswalk and Lars-Steenhoff like this.
  15. OCASM

    OCASM

    Joined:
    Jan 12, 2011
    Posts:
    328
    You could always use visual scripting for quick prototyping and once you have you are feature complete optimize by convert to code.
     
  16. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    Let's not. All hail Murphy's Law.


    Yes, if they put enough hours into studying it.

    There's a huge difference between a proper programming and what you need to do within a game engine framework to make your game tick. While proper programming requires years to get good at it, in-game scripting is something you should be able to pick up in a week or so. No need to make mountains out of molehills and no point in trying to be sarcastic.

    If you can use visual programming tools, then you can write game scripts (and will be more efficient at it).
     
    zombiegorilla, MV10 and Ryiah like this.
  17. Stardog

    Stardog

    Joined:
    Jun 28, 2010
    Posts:
    1,910
    Hopefully we'll see some experimental version in 2017.
     
  18. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    Its not a question of skill, or lack of it
    its a question of different ways of getting to your end goal, pick what you prefer. and enjoy the journey
     
    elbows likes this.
  19. thelebaron

    thelebaron

    Joined:
    Jun 2, 2013
    Posts:
    851
    @neginfinity There's a lot more margin for error in writing a script from scratch than plugging in nodes in an editor. I wouldnt trust a programmer to learn to make competent game art within a week, and if I was tasked with leading a designer, given the choice of pushing a designer to spend a week learning to script vs handing them a visual editor I'd choose the editor every time.
    While learning to code might be a more efficient long term strategy, especially on your own time, if you are in an actual work environment and you had to move a product, you wouldn't force your designers to learn coding if an easier solution exists.
     
    elbows and Lars-Steenhoff like this.
  20. MV10

    MV10

    Joined:
    Nov 6, 2015
    Posts:
    1,889
    This must be the only industry in the world where it's apparently valid to get a job without actually having the skillset needed to do that job or any intention of developing that skillset.
     
    neginfinity likes this.
  21. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    Who thinks that games are only made by programmers ?

    the job is usually more broad, most games have sound right? how many of then use procedurally created sound effects written in code? certainly not all.


    why talk about skillsets? its not relevant.
     
    Last edited: Mar 23, 2017
    Tomasz_Pasterski likes this.
  22. thelebaron

    thelebaron

    Joined:
    Jun 2, 2013
    Posts:
    851
    No but apparently it is an industry where different disciplines don't fully comprehend other disciplines. A programmer is not necessarily a designer and vice versa.
     
    elbows and Lars-Steenhoff like this.
  23. tswalk

    tswalk

    Joined:
    Jul 27, 2013
    Posts:
    1,109
    well.. that boils down to your role though ya? If I was a level designer or such, I think having a node editor which encapsulates the games mechanics would certainly be faster than banging out code to make some changes. But again, you would need someone with the skills to create that framework to make iteration a benefit using the tool.

    I don't think its' an argument about which is better, and what skill set is required... I think it is more a question about so many other facets of workflow. If a node editor perhaps had the benefit of being able to pin-point an area that needs optimization, that would be a perfect case for a programmer to focus their attention.

    reading an earlier comment of yours.. I think we're both talking around the same thing :D
     
    Lars-Steenhoff and elbows like this.
  24. elbows

    elbows

    Joined:
    Nov 28, 2009
    Posts:
    2,502
    That doesnt really take account of different strengths and comfort zones of different human minds.

    There are many factors that make some humans not really any good at programming, or not want to do it, or lack confidence in doing it. Different strengths of memory for example - some people arent so good at learning and remembering syntax for example, and visual metaphors that arent quite as reliant on words, numbers and symbols might go down better.

    There are a bunch of barriers to understanding scripting, coding and programming that visual scripting cant solve. So I do not seek to pretend that it is a magic answer, and no doubt some people who try existing visual scripting stuff dont get too far because they arent willing or able to get their heads round the logic and flow, some aspects of problem solving or how quickly things can get complex and messy. But there are those who will benefit from it, whether it be a step in a journey to other forms of scripting and coding or whether it just allow some basic tinkering without being so intimidated.
     
  25. MV10

    MV10

    Joined:
    Nov 6, 2015
    Posts:
    1,889
    Nobody said that.

    thelebaron described a work environment in which a designer's tasks may require something which, apparently, designers aren't expected to know how to do without the use of a tool which generally doesn't exist.
     
  26. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    Sure, I think I misunderstood.

    Im just excited that something is coming to unity that is aimed at making new workflows possible.

    I can be a little bit passionate about that :)
     
  27. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    There are things called IDE. They provide multiple aids for this very purpose. There's no bigger margin for error. If you messed up the script frequently simply won't run and will tell you why it won't run.

    Programmers specialize in information gathering and figuring out how to do something. So, if you task a programmer with making a 8bit art for an arcanoid (which is, he or she should be able to cobble up something during this time. Won't be a masterpiece, of course.

    Look, by subscribing to use of visual tools you work inefficiently and waste time that could've been spent making art or designing stuff. Therefore it is your own interest to learn how to script. Nobody asks you to specialize with it. This is an equivalent of learning first letter of programmer's alphabet. And you basically say it is too hard. Well, if it is too hard, then perhaps someone else should be doing this.

    There ARE few niche applications of visual tools where they shine, but writing low-level game logic (which is what people try to use blueprints for all the time) in them seriously isn't one of those suitable applications.

    Yeah, exactly.

    I think that some people might be afraid of even basic programming. And in this case they definitely should overcome their fears.
     
  28. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    I can write game logic in playmaker no problem, I would not call it low level per se, and I would be happy with higher level visual scripting. Give me a few nicely prepared actions ( code ) and I can connect them in seconds and get the results I want. still the logic and the structure is basicly the same, its just how you acces and excecute it.

    For example a get collision event and after playing a sound, animation, a particle effect and a camera shake can be set up in playmaker in less then a minute.

    I can do the same in a c# script but i don't prefer to do that most of the time.

    Just my preference.

    and yes basic programming knowledge is good to spend time on. Im happy that I did. and im also happy that playmaker is available, just to link up peaces of code.

    Makes it easier to get an overview for me of whats running and walk trough the steps at runtime.

    Writing custom actions is also not that difficult.
    and are easy to reruse.

    I do know that there are pitfalls mostly todo with refactoring, it can take time to manually search and replace nodes that in code is just a second to do.

    same for version control, it can be way more easy to see in code what has changed than in a visual script hiding inside a scene.

    So pro's and cons. however I would choose to have the options to use or ignore them rather than not have then at all.
     
    Last edited: Mar 23, 2017
  29. thelebaron

    thelebaron

    Joined:
    Jun 2, 2013
    Posts:
    851
    @
    Yeah it will tell you why it didn't run, but if you don't have a programming background every little bump in the road is going to have you scouring the scripting api, unity answers, the forums maybe even checking out stackoverflow, when during that time you needed to be working on making a game fun.

    People specialize, not just programmers. I said If I tasked a designer, not if I were a one man shop making a game. I don't want my designer making art or rooting through technical documents when they should be designing a fun experience. If visual scripting enables this(which it obviously does, why else would companies like Ubisoft, EA or Guerrilla Games build it into their flagship engines) then why vehemently oppose this?

    Look, I personally and selfishly wish Unity would focus its efforts on terrain before a visual script editor because you know what? I learned to code through visual scripting. Zbrush and character art are my bread and butter, so I know what its like to go through the grind from pure ignorance to an understanding. Its why this thread oddly matters to me, and I hate seeing people say oh its so easy just go code instead, its really not. While I may not touch visual script editors anymore, I'm sure as hell not gonna fault someone who can make it work for themselves.

    Yeah people do try to do too much with them, but you know what, bad practices aren't limited to just one facet. There are lots of people making bad art and bad code, all created without the help of visual scripting. Visual scripting is just another tool in your toolbox and if it doesn't work for a given situation, then you don't have to use it. Coding doesn't automatically force good habits, in fact you'd probably have the latter by forcing non-programmers to program because its somehow in their best interest to.

    Unity is all about democratizing game development, and I fully believe that visual scripting would be another big part of that.
     
    elbows and Lars-Steenhoff like this.
  30. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    And? You'll go through EXACTLY the same process with visual scripting. How many node types do you think are there in Unreal 4? I think the number is definitely above a hundred. Probably below a thousand, though. You'll still need to do exactly the same thing...

    ... If you use visual tools. Visual Scripting is the same programming in disguise. Except slower.

    https://docs.unrealengine.com/latest/INT/BlueprintAPI/


    They're a big companies and can waste some cash on R&D. What they do is not necessarily a good thing for every other smaller developer. It is worth keeping in mind that those companies do not exactly have a flawless release record either. "If a big company does it, it must be the right thing to do " ... this kind of thinking is not a good idea.

    I value power over "accessible to everyone". Tools that are accessible for everyone tend to offer less power in order to simplify learning curve. Therefore I do not necessarily see attempts to "democratize" as a good thing.

    In the end the whole visual scripting can be used as a clever advertising trick. Promise people that they can build that amazing thing, without mentioning the complexity of the engine they'll still need to handle anyway.

    Because I see idea of visual scripting as wrong. It is literally a lot of people wasting time using inefficient approach - time they could've spent doing something else. Yes, there are practical applications (sound synthesis/priocessing, image processing and AI decision trees), but most people seem to sue visual scripting on simple things and only waste time, with visual programming tools giving only illusion of ease of use in return.

    Want a useful visual scripting system? Let people type code within the nodes and let them build those node networks quickly and efficiently with keyboard. Then we're talking.
     
    MV10 likes this.
  31. elbows

    elbows

    Joined:
    Nov 28, 2009
    Posts:
    2,502
    Well I dont need a visual forum to know we are probably stuck in an infinite loop with some of these points ;)
     
    Lars-Steenhoff likes this.
  32. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    If I had to say why, it would be because these AAA studios have many hundreds of employees working on different parts of a game. To be able to pass around a clunky, inefficient prototype that is approachable by everyone across level designers, sound designers, artists, technical artists, animators, UI designers, cinematic choreographers, programmers etc would be a huge advantage, compared to having a hundred people crowding the programmers desk at any given time wanting to tweak a variable.

    The problem is that visual scripting still isn't a great way to build a game and ship it, which no doubt is what most Unity users want it for.
     
    neginfinity likes this.
  33. N1warhead

    N1warhead

    Joined:
    Mar 12, 2014
    Posts:
    3,884
    Oh boy what did I start hahaha, I mentioned why I didn't like it and bam the next day, all over the place hahaha.

    I'll say it again, it has it's uses, for quick little things - which of course I'd still rather code it. But using it for every single thing in your game, espically if you're trying to make the next Witcher 3 (like a lot of people seem to want to do), then it's going to kill your performance quicker than tripping on a root.
     
  34. MV10

    MV10

    Joined:
    Nov 6, 2015
    Posts:
    1,889
    Marketing slogans are rarely a good basis for an argument. I appreciate that it's free to a relatively generous point, but if you think that, too, is anything but a marketing decision, you need to reevaluate your relationship with The Kool Aid.
     
    dadude123 and neginfinity like this.
  35. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    It is lot better for debugging visually what is going on.
     
  36. Tomasz_Pasterski

    Tomasz_Pasterski

    Joined:
    Aug 12, 2014
    Posts:
    99
    Wow this is madness, i dont even realize this thread will enable so much voices against, but most seems to be like trying to convince this is barely useless and unity should focus on other things.
    What i think about VS is that this is only (mostly) for prototyping tool for artists and i think Unity knows better about what they want to do with own tool and how to manage own resources better than most if not all users here.
    Thx all who take part in the discussion and lets see what Unity will deliver.
     
  37. stormwiz

    stormwiz

    Joined:
    Feb 27, 2014
    Posts:
    145
    From what i've seened so far vs is not slow anymore, it used to be about a year ago.
    Take Nottorus for example. It can compile to pure c# and can convert back to node.
    Blueprints (can compile to c++) is another one currently making improvements and refining the technology.
    Visual scripting these days can be use to make full games and not just simple games. Its no longer just a prototyping tool, but another way to code gameplay, etc.
    As these tools become more and more automated, text programming will slowly fade. For the new blood of game designers it will be quicker to master versus having to learn the old ways.
    The goal here is to get the game out as quickly as possible and coding it should not be a road block for anybody.
    Game designers can just focus on making the game beautiful, engaging, fun and efficient without the aid of a programmer.
    Wether we like it or not vs is here to stay and could one day possibly fade out traditional coding.

    I think Unity should of had this and many other new technologies already instead of piggyback ridding the asset store for so long. Unity move your ass and stop letting the grass grow.
     
    Last edited: Mar 25, 2017
  38. thelebaron

    thelebaron

    Joined:
    Jun 2, 2013
    Posts:
    851
    No but there's a thought process behind every marketing decision. Their goal is to make things easier so more people can use their product. Unity as a company is valued at over a billion dollars. They want more users, and more successful users mean they get to boast about how great their engine is and continue to grow. Are you that naive to think one of they just woke up and decided to implement a feature without seriously doing some research and debating the cost/benefits of it?
     
  39. MV10

    MV10

    Joined:
    Nov 6, 2015
    Posts:
    1,889
    I've been hearing that for 40 years.

    So we can at least agree it's primarily a marketing ploy rather than a solution to a serious need.
     
    hippocoder likes this.
  40. elbows

    elbows

    Joined:
    Nov 28, 2009
    Posts:
    2,502
    Not at all. Pointless trying to change other peoples minds though, so I will simply predict that the visual scripting, once mature and performant enough, will be quite a success. and useful. Time will tell eh!
     
  41. MV10

    MV10

    Joined:
    Nov 6, 2015
    Posts:
    1,889
    I'm just here for the conversation. I doubt Unity much cares what people like me (who are extremely unlikely to draw my software) think about it. Except perhaps the fact that if the intro tutorials were all about Programming With Legos, I probably would have uninstalled it and moved on.
     
  42. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    It looks you are someone stuck in it's own principles and ideas, there is people and artists making 2D and 3D games with visual programming, why should that bother you if Unity proposes a Visual Scripting tool ? You are a coder , not everyone is a coder or want to learn.
    Visual Scripting is used in AAA games, this helps a lot to get the game made.
    https://www.guerrilla-games.com/read/creating-a-tools-pipeline-for-horizon-zero-dawn
     
  43. cdarklock

    cdarklock

    Joined:
    Jan 3, 2016
    Posts:
    455
    I don't care much for those people. Anyone who consistently says "stop helping other people, you should be helping me" is a self-centred jerk.

    The visual scripting tools are not at all interesting to me, but I think it's a Good Thing that Unity is making them. They will probably help a lot of people. Sure, those people are not me, but I am still glad someone is helping them.
     
    Tomasz_Pasterski likes this.
  44. MV10

    MV10

    Joined:
    Nov 6, 2015
    Posts:
    1,889
    As I've stated many times, because Unity has bigger fish to fry.
     
  45. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    There is lot of features that Unity could dig, and Visual Scripting is one of them, this doesn't mean other features will not come.
     
    Last edited: Mar 25, 2017
  46. MV10

    MV10

    Joined:
    Nov 6, 2015
    Posts:
    1,889
    I keep unfollowing this thread... then somebody quotes me.
    :D
    1.png
     
    Ryiah likes this.
  47. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    All I saw there were specific workflow tools made for parts of the game, mainly an explicit state machine editor (which a node editor is perfectly suited for, e.g. mecanim) as well as a conversation editor (another pure state machine). There were a couple of random shots of what looked like general visual scripting but no context as to what it's used (or useful) for.

    I think the biggest problem with this thread (and all the other threads on this topic) is that there's a complete and utter lack of contribution from non-coders who have used any of the various tools out there on the asset store (or even talking in general terms about their experience with UE blueprints) which leads me to believe that they either cannot or will not use a visual scripting tool that is available to them, which begs why Unity should bother making one for them.

    I'm open to supporting a visual scripting tool if I can see that it's going to help a substantial amount of people, and for that I really need to see examples from people who have at least attempted a minimally sophisticated game made with visual scripting, and preferably a run down of what is good/bad/needed for a useable visual scripting editor. Until that happens I can only imagine that it's another one of these "if only I had X I could make a good game..."
     
  48. zenGarden

    zenGarden

    Joined:
    Mar 30, 2013
    Posts:
    4,538
    It's Unity business to make some search and decision to include a new feature, because it will be included in Unity i think they found that it matters to include it. I don't think they just said , it's a cool feature let's bring it.
    (And you are not forced to use it because it is included)

    You should look at what people are doing with Blueprint only in UE4, people using Blueprint only and people that mixed code and visual scripting to ship their game.

    Anyway , if Unity brings a shader editor next, we will have again people stuck on their principles and ideas saying it is useless because shader coding is better or no one will use it lol
     
  49. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    Sure it's unity's business, but I would be surprised if they don't pay attention to what people talk about on the forums. I notice that constructive technical discussions often quickly attract a Unity engineer.

    And that's why, if I was wanting visual scripting, I would be going all out to show what I could make with something like playmaker, listing out all the specifics of what I'd like it to have/not have, and basically show why it's useful or necessary to me. But as I said, I haven't seen in this or any of the discussions on this topic so much as an example node from someone using any of the fairly advanced tools that are already available.

    I'm not actually 'against' it as such, in my opinion it's just quite useless, but I'd like to get to know why people want it (besides the whiz bang factor) and I'm interested in anything that helps people to build games, so I would beat a drum in support it if someone could show me how it's possible to make something substantial with it. In fact I like the 'idea' of a visual script editor but in practice it doesn't seem very useful.

    PS if in fact it got relegated to the Research section of the roadmap, I would be wondering if they are actually planning to go ahead with it in the near future. So it's another reason why it would be good for people who want it to articulate a good argument in favor.
     
  50. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    For toy applications, maybe. Or that 1% exception where visual tools are actually useful (AI decision trees).
    For the most application the amount of information to "display visually" will be too high and text based approach with debug outputs and stacktraces are better.

    It is mostly useless, even for prototyping.

    The areas where visual scripting works are very high level logic like AI decision trees, and building highly parallel filter graphs for signal and image processing. That's it. However, this is not "blueprints" level - it is above it. Instead high-level work people usually try to implement low-level operations in blueprints which would've been faster as pure text approach.

    People usually view visual scripting as some sort of magical tools that would enable anyone to program without experiencing grueling horrors of programming. This idea is naive and misguided.

    You'll still have to deal with all the same stuff, with equivalently huge APIs ( see: https://docs.unrealengine.com/latest/INT/BlueprintAPI/ ), except that you'll be locked out of power tools available for normal language, and will be essentially locked in your "walled garden" kind of api.

    As I said before the one way to make blueprint-based approach useful is to allow people to write code directly within the nodes and make sure the approach works very efficient with keyboard. Or represent nodegraph internally as a normal language like lisp. Instead people worship nodes that will make their wishes come true.

    Wrong. The goal is to bring in more profit. Not to make it easier. It doesn't matter if the product is easy to use, what matters is that the customer arrives keeps paying and never leaves.

    Basically, from cynical/realistical viewpoint unity tech would be only interested in you paying subscription. Not having an easy time with the engine or ever finishing that game of yours. That's because your subscriptions bring in more profit while your game sales do not. Primary customers of interest would be enterprises that can handle "contact us" level of payments where you possibly get source code access. In the end quite a lot of indie projects are probably a small fish anyway. The subscription payment scheme can be abused in fairly diabolical ways, but thankfully unity does not seem to have interest in doing anything of this sort. Interestingly Unreal's royalty based approach gives developer a better financial incentive to help you succeed (because the more you earn the more they get).

    Either way, in the end Unity will do what Unity will do. "Visual programming so everyone can make thier game!" is a good marketing slogan (even though it does not tell the truth), even better if people start believing it.
     
Thread Status:
Not open for further replies.