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

Please stop making updates

Discussion in 'General Discussion' started by l33t_P4j33t, May 2, 2020.

  1. l33t_P4j33t

    l33t_P4j33t

    Joined:
    Jul 29, 2019
    Posts:
    232
    please, its too much in too short of timespan
    there are way too many enormous changes and features added in 2019.
    its too much too learn, my brain is gonna explode
     
  2. ShilohGames

    ShilohGames

    Joined:
    Mar 24, 2014
    Posts:
    3,015
    Let Unity make updates as fast as they want. All you need to do is stick with the most recent LTS version if you want no new features for a while. For example, the LTS right now would be Unity 2018.4.22.
     
  3. l33t_P4j33t

    l33t_P4j33t

    Joined:
    Jul 29, 2019
    Posts:
    232
    you can't seriously expect anyone to learn in less than a year all the following:

    Netcode
    Visual Effects Graph
    UI Elements
    Data Flow Graph
    Unity.Animations
    Unity.Transport
    Lightmapper Changes
    DSP Graph
    Project Tiny
    URP
    Unity.Physics
    Animation Rigging
    Input System
    a ton of 2d changes and packages
    + Allot of other small changes

    its too much, what if they release a similar amount next year. it'll create a vicious cycle.
     
  4. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,965
    No, but that's your mistake for thinking that you should.
     
  5. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,617
    No one person is expected to know every single part of Unity, or game development in general. It's not practical. For small projects just use what you need and focus only on relevant parts of the engine and tools. For large projects you'll need a team, and everyone should focus on just the parts relevant to their job.

    I've been working with Unity since before the Editor was on Windows, and there are plenty of parts I've touched barely or not at all along the way.

    Closely related, have a read of this thread.

    Edit: Hmm, that's already one of your threads. It's easy for me to say and hard for someone to do, but here goes anyway: stop worrying about knowing stuff, and just focus on finishing stuff. I've been doing this for nearly two decades now and I'm still constantly learning on the job. That won't ever stop, and to me it's a part of the fun.
     
  6. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    If you realized how many features were already in Unity, you'd really keel over. And that's not to mention all the features you could create after learning programming and how to write editor scripts, in which case you could pretty much build any tools and features into Unity that you want.

    You need to learn how to focus the scope of your learning to what is relevant to your goals, and explore shiny new things every once in a while. Stop trying to go to Unity University and start making whatever type of game interests you.
     
    Ryiah and angrypenguin like this.
  7. sxa

    sxa

    Joined:
    Aug 8, 2014
    Posts:
    741
    Its only you who seems to have that expectation. Noone else thinks you need to know all of each version of Unity.
     
  8. MDADigital

    MDADigital

    Joined:
    Apr 18, 2020
    Posts:
    2,198
    Unity needs to rework their QA strategies, they have a huge problem with regression. But that doesn't mean they should stop updating the software. That's not possible with any software more so in this industry.
     
  9. Olmi

    Olmi

    Joined:
    Nov 29, 2012
    Posts:
    1,553
    Not all of us can afford to think that way... Quite often people might ask you (based on experience), do you know this, that and something else about Unity, and then they valuate you based on that. No matter how fallacious such analysis about a person's abilities is, you most likely are left without work if you don't check enough boxes. Companies often advertise themselves by saying that they appreciate generic problem solving skills etc.etc. but in reality they are looking for a person that can use a tool X, right now, when they plant you on an office chair, and not in two weeks or later. So in that sense it's not that good from a single person's perspective there's a flood of new half baked tools and features...
     
  10. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    Why would somebody working on an art pipeline need to know Unity.Transport, the input system, netcode, or the data flow graph? Why would somebody working on networking need to learn the graphics features? In a hiring situation, these things don't have a lot of overlap, even for a generalized position. I know this because I've assisted in the hiring process for micro to medium-sized teams. In a large team, there's even less overlap.

    This is an imagined problem.
     
    Kiwasi and Ryiah like this.
  11. Olmi

    Olmi

    Joined:
    Nov 29, 2012
    Posts:
    1,553
    @Murgilod imagined problem how? I've bumped into that during last two years 10-15 times. Of course it depends on your social class, region of the world you live in and so on. (of course not *everything* is required like you implied, but way too big scope to be realistic...)
     
  12. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    Which is why you use things like API familiarity and basic problem solving skills in any interview process. Nobody is expected to know everything about an engine or even most of an engine, but they're expected to know how to learn the engine. If you've bumped into it in the past two years that many times you should be looking at an overlap of what the most common factors in and focusing on those specifically. Like, let's get into some specifics:
    • Project Tiny has literally no overlap with VFX Graph, URP, the new Physics system, or the new lightmapper. On top of that, it's still in fairly early preview. People aren't generally hiring for that at all.
    • The Data Flow Graph is a part of DOTS that is so deep in preview that we're not going to see this be in a remotely production ready state for probably over a year.
    • VFX graph is an artist specific tool that maps pretty evenly onto most VFX toolsets that already exist, so if you're expected to work with that you're expected to have familiar with tools outside of Unity already.
    • DSP Graph isn't even out of version 0.1.x yet. This is another thing that is so deep in preview that nobody is hiring for it and it's well over a year out.
    • The new input system is the kind of thing that you might get asked about, but you can learn it over the course of an afternoon because it's really not that complex. If you've used anything like Rewired or other input assets (which you probably will have if you've developed in Unity for any amount of time), you pretty much already get it.
    • "A ton of 2d changes and packages" doesn't really say anything. What changes? The scope of most of those changes are generally minor API factors, which every engine goes through.
    • "A lot of other small changes" are small changes and easy to adapt to because they're small.
    Not only are you not expected to know everything, but the actual scope of what you are expected to know is not the entire engine.
     
    Ryiah likes this.
  13. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    989
    skillset needed is overwhelming.
     
  14. ShilohGames

    ShilohGames

    Joined:
    Mar 24, 2014
    Posts:
    3,015
    You don't need to learn every new feature of the engine to make your game.

    Download an LTS release. Use that version to make your game. Don't update the engine version until there is a new feature that you absolutely need.
     
  15. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,965
    We say it a lot in threads where beginners ask how they can get started but it bears repeating it here given the topic and the number of beginners commenting on it. Spending an entire year learning the basics of a modern game engine is completely normal. Anyone expecting to learn everything they need for their first engine in a few months is kidding themselves.
     
  16. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    989
    Thank you. How many months is feasible to becoming expert?
     
  17. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,965
    How are we defining an expert here?
     
  18. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    989
    1-2 persons disciplined in skills listed in post 3.

    At present myself and friend feel we are lacking in level where we need to be compared to everyuone else.
     
  19. shion33

    shion33

    Joined:
    Oct 21, 2019
    Posts:
    14
    Many of those features are still quite early in development, and have been released more with the intent of getting community feedback than being used in production. And the devs have said already that they are going to be mostly focusing on getting these features ready rather than adding new ones. So there will be plenty of time to learn them before they're stable, and of course the legacy features are still here.
     
  20. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,965
    Joke is on you then. There are no developers out there familiar with every item in that list. In fact it's completely unrealistic to think that you can learn every item in that list. I've used Unity for more than eight years now and I don't know every item in that list nor would I expect to even with another eight years.
     
    Tzan likes this.
  21. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    989
    OK i will try to work harder and dedicate more time to learning.
     
  22. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,965
    For the record the vast majority of experts became experts not just because they invested a ton of time but because they chose a specialization and focused most of their time on that. If they had decided to learn everything most of them would have been an expert in nothing. There simply isn't sufficient time for the vast majority of us to be experts in everything.
     
    angrypenguin likes this.
  23. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    989
    Maybe not expert but close to expert. I feel some topics on list are redundant though. I feel it is best to learn more than one so you become better rounded developer. is this not true?
     
  24. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,965
    It's heavily if not completely dependent on the specialization you've chosen and what your employer expects of you. I'm primarily a programmer therefore any secondary skills are dependent on being a programmer. If the skill has little to no dependency on being a programmer it's a waste of my time to learn it unless we don't have anyone else that can handle it.

    Just looking at the list he created the following are dependent on being a programmer: (1) the input system, (2) UI Toolkit (previously known as UI Elements), (3) Unity.Transport, (4) Netcode, (5) Project Tiny. Only five of the thirteen immediately stand out to me as things that are relevant to my specialization.

    Everything else is dependent on my role at the company. Just as an example my current employer expects me to have an understanding of animations since I'm responsible for setting up just about everything related to user interfaces including their animations. He doesn't expect me to know the SRPs because we have an artist that knows them.
     
    Last edited: May 2, 2020
  25. MDADigital

    MDADigital

    Joined:
    Apr 18, 2020
    Posts:
    2,198
    You dont need to switch to every system at once. For example our game started its life on unity 5.2 and is now on 2019.3.11 but most systems don't utilize anything of the new stuff, we are looking at moving some systems like AI to ECS/DOTS and also looking at moving to URP. It's a effort over time.
     
  26. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    989
    Thank you for your opinion sir. So i understand you know programming and unity UI. That is very small subset of engine. But i sure you doing great job. we try to understand many more fields.

    But we are just hobbyist when pandemic is over and school starts again it is back to normal sadly.
     
  27. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    989

    yes it is not simple to move/upgrade so i am reading.
     
  28. Ryiah

    Ryiah

    Joined:
    Oct 11, 2012
    Posts:
    20,965
    Trying to understand many more fields can be worthwhile if you want to see which fields you enjoy and which ones you don't. Back in school I had access to both programming and content creation tools. I rapidly learned I liked programming more and that's why I'm doing it today.
     
  29. unit_dev123

    unit_dev123

    Joined:
    Feb 10, 2020
    Posts:
    989
    Thank you :)

    At the moment i enjoying cprogramming and making mechanic but my friend enjoy making graphics, which i do not care for. I hope to learn more and understand object orientating. But so far i use simple scripting.
     
  30. ShilohGames

    ShilohGames

    Joined:
    Mar 24, 2014
    Posts:
    3,015
    Typically speaking, it takes 10 years and/or 10,000 hours to become an expert at something.

    You can build a lot of fun things in Unity along the way, though. You don't need to train for 10 years before starting your project. Enjoy.
     
  31. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    Not really, no.

    https://www.6seconds.org/2018/02/09...hat-you-actually-need-to-know-about-practice/

     
  32. ShilohGames

    ShilohGames

    Joined:
    Mar 24, 2014
    Posts:
    3,015
    I am not trying to set a specific requirement of exactly 10,000 hours, and I am well aware that the exact number varies depending on the person and the knowledge domain. If somebody asks how many "months" are needed to be an expert, it is import to at least bring up the concept of 10 years and/or 10,000 hours. Nobody becomes an expert in a complex knowledge domain in just a few months.
     
  33. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    But "typically speaking," 10,000 hours isn't even remotely accurate is the thing.
     
  34. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    5,181
    "a watched pot never boils"
     
  35. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    A bird in the hand is worth two in the bush
     
  36. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,617
    If you're starting from nothing then it's not going to be months, it's going to be years.

    The thing is that you're not learning "Unity". You're learning "game development",which is a colossally huge field. Unity is just a tool you're using along the way.

    Forget about labeling yourself an "expert" anyway. Just make and finish stuff. Put it out there. Learn from it. Not only is that more important than what you call yourself, it's how you become a real expert anyway - by using the tools to do real work for real people.
    Have you thought that they might be asking you a super broad range of questions so they can find out what you do when you don't know the answer?

    When people ask me questions about things I don't know, I first explain that it's not something specifically covered by my experience, and then I tell them how I'd approach the problem to solve it anyway.

    Being an expert isn't about having an encyclopedic knowledge of all things, it means having the foundational knowledge and experience to reliably solve new problems for yourself. Trying to tick off that you "know" all features in an engine is missing the point, because it's not addressing the underlying foundations you need.
     
  37. ShilohGames

    ShilohGames

    Joined:
    Mar 24, 2014
    Posts:
    3,015
    For the purposes of this forum thread, it does not matter. It is not about nit picking the exact number of hours. I am well aware that the exact number varies depending on the person and the knowledge domain.

    The point of mentioning the 10 years and/or 10,000 hours concept in this forum thread is to illustrate that becoming an expert takes a lot longer than a few months. The question I replied to literally asked "How many months is feasible to becoming expert?" The general rule of thumb to become an actual expert is about a decade (120 months).
     
  38. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    It's a bad rule of thumb and a bad thing to say because, as stated in the article I linked:

    The 10,000 hours myth is a bad thing to promote because it is literally the definition of quantity over quality being treated as an axiom. 10,000 hours won't help people like unit_dev123 because it doesn't offer anything other than a shrug and "it's time consuming" while offering no actionable advice.

    "Just practice" means nothing. "Just practice" is why you can spend 10,000 hours on something and come out on the other end scarcely better than when you started.
     
    ChazBass likes this.
  39. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,617
    It's years since I read that book, but I could have sworn that he did qualify that the 20 hours per week had to be "deliberate practice", and talk a bit about what that meant.*

    In any case, I think it's pretty moot when you're in a competitive field for reasons which I hope are pretty obvious. Even if you meet some arbitrary benchmark there are plenty of other people who've done the same thing, and you still have to compete with them. There's no benchmark point of "if you're this good then you'll be successful". Whether you measure it in hours or by grading technical skill or whatever else, the fact is that you're competing against people who've put in far more, or scored far higher, than whatever that benchmark is.

    Being successful isn't about being technically better than other people. It's about figuring out how you can deliver value with what you can already do, and/or figuring out the path of least effort to be able to deliver value.

    *Edit: In the greater context of the book, I also don't remember the "10k hour rule" being used the way people are suggesting. The book is not saying "do one of these things and you'll be successful". The book is called Outliers and looks at people who were particularly successful at something, describing a number of things that many of those people have in common. I don't recall the message being "practice this much and you'll be successful". I remember it much more as "this much practice is one of multiple things that the group being studied had in common." One of the others was being born at the right time in relation to the thing you get good at. Try doing that deliberately...
     
    Last edited: May 3, 2020
    Kiwasi and Ryiah like this.
  40. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    You're missing the point. It's not an entire learning strategy, and there's no point trying to frame it that way. It's just a reminder of one aspect of learning, which is that (as I understand it) a large portion of learning is simply via exposure over time, since the brain is already equipped with all kinds of tools for background synthesis of raw information.

    I remember something from a book I read by David Eagleman about the brain, where during WW2 it was necessary to visually identify aircraft as friendly or enemy from a long distance to raise the alarm in advance. These aircraft would be a dot on the horizon, barely visible at all.

    Some people happened to be good at it, so a lot of effort and expense was spent on trying to figure out why, so others could be trained up quickly. The only thing that turned out to be useful was having the trainee spend a lot of time doing it with yes/no feedback from someone who was good at it. After some time they just 'knew' if the dot on the horizon was not an allied aircraft, without really knowing why.

    The point is that the brain is running learning and adaptation algorithms completely outside conscious thought and action, which determines part of successful learning. And these simply require lots of information to do their job.

    It's only part of the picture, big or small depending on your point of view. But still very important.
     
  41. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,617
    It gets you lots of clicks on your debunking articles. ;)

    On a slightly more serious note, to me half of the value in a Malcolm Gladwell piece is the discussion it spurs rather than taking it as gospel. I kind of feel like reading Outliers again now...
     
  42. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    I dumped Outliers because (like a lot of Gladwell's work, I quickly learned), his entire view on things is reductive and speculative based on the reductionist he uses to the point where it's effectively useless. No matter what he specifies, he's still wrong. From the same article:

    I can't help but stress that K. Anders Ericsson was the lead researcher on the thing that Gladwell was drawing from for his 10,000 hours figure and that Ericsson is also a repeated critic of Gladwell's interpretation of his work. Malcolm Gladwell is not a psychologist (Ericsson is), nor has he ever been trained as one. Multiple psychologists and scientists who he's cited have come out as criticising him for being extremely reductionist of their work and ignoring the actual results they saw in favour of anecdotal evidence. Gladwell is not a reliable source and the data he presents is generally flawed.

    Gladwell is a journalist, a writer, and a speaker, but he is nowhere near an expert on the things he speaks about, which is what makes treating the 10,000 hour myth as even a broad strokes concept a bad idea. If the person he got the numbers from disagrees with him because he so dramatically oversimplified the work, he's just not worth trusting.
     
  43. angrypenguin

    angrypenguin

    Joined:
    Dec 29, 2011
    Posts:
    15,617
    Exactly. When I want to read science papers I get science papers... from scientists. I'm not expecting white paper levels of detail, accuracy or precision in a little book that touches upon many different topics.
     
    Billy4184 likes this.
  44. BIGTIMEMASTER

    BIGTIMEMASTER

    Joined:
    Jun 1, 2017
    Posts:
    5,181
    @Murgilod , complaining without offering better solution is just whining.

    what's your advice? 20,000 hours? Vow of silence until the complete truth has been realized? Just give up?


    Plenty of good advice here. What good is semantic nitpicky if you don't have clear counter point to offer.
     
    angrypenguin likes this.
  45. ShilohGames

    ShilohGames

    Joined:
    Mar 24, 2014
    Posts:
    3,015
    You are getting completely sidetracked by the specifics here. The 10 year and/or 10,000 hour rule of thumb is definitely useful, because it is an easy way for people to visualize that it takes a long time to actually become a real expert on a topic. I am well aware that the exact number varies depending on the person and the knowledge domain. I don't go around quoting 700 to 16,000 hours, because that is a wide range and is harder to remember.

    If somebody hears 10,000 hours and then decides to shrug and say "it's time consuming", then that person does not get to become an expert in that knowledge domain. Being an expert on a topic take a lot of time and commitment. It is not easy. But as I already mentioned in this thread, they do not need to become an actual expert in order to use the product. Thousands of games have been successfully made with Unity by developers who were not already experts when they started their project.

    The important point for the OP to realize is being an actual expert take a lot of time, but it is not necessary to be an expert in order to make game. The OP literally asked for Unity to stop adding new features, since the OP assumed they would need to master every feature of the engine before making a game.
     
  46. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    A little bit of googling of the name of the person who actually did the study would help. I figured I could say "Ericsson, the person who did the study and is not a fan of what Gladwell reduced his work to, is an actual psychologist who did actual studies" would lead to people actually looking up his work.

    https://hbr.org/2007/07/the-making-of-an-expert

    This came the year before Outliers did and contains more actionable information than anything in the book.
     
  47. N1warhead

    N1warhead

    Joined:
    Mar 12, 2014
    Posts:
    3,884
    There's only one thing I dislike about updating, perhaps I'm just doing it wrong - but using Unity Collab makes it a total pain to do. (perhaps I'm doing it wrong and missing the point). But every single time I update I have to re-built the library, which is fine. But then when I get into Unity it has to re-download the entire project again from Collab and then there's always a merge conflict. Generally it's an easy fix, But not always. but perhaps I'm doing it wrong I have no idea. So I've decided im'a stay at 2019.3.6f1 until 2020 comes out.
     
  48. MDADigital

    MDADigital

    Joined:
    Apr 18, 2020
    Posts:
    2,198
    I havent used collab. But with git its important to exclude Library folder this is a local only cache folder and shouldnt be commited
     
    Kiwasi likes this.
  49. N1warhead

    N1warhead

    Joined:
    Mar 12, 2014
    Posts:
    3,884
    Not sure I can exclude it, i'll look into it. Thank you for the tip, I'll research collab a bit.
     
  50. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,084
    Collab uses the same formatting as .gitignore but it's called .collabignore

    https://docs.unity3d.com/Manual/UnityCollaborateIgnoreFiles.html

    You can just copy over the contents of any standard Unity .gitignore template into it, as I recall.
     
    Kiwasi likes this.