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

Official Workflow Case Study - Show us your pain!

Discussion in 'Editor & General Support' started by willgoldstone, Mar 1, 2019.

  1. JohnHudeski

    JohnHudeski

    Joined:
    Nov 18, 2017
    Posts:
    126

    This should be a standard editor feature. The inspector view is too limited and the preview is useless

    I should expand.
    Most games out there a focused on character/avatars (yeah there are puzzle games out there but even racing games can be considered character games). Since unity has recognized the need for game foundations I believe one general aspect they can make a difference is in a unified character editor system as the competition has.

    Right now we have unity ik rigs, the very very old ragdoll system, and mecanim.
    Setting up your character with all the bells and whistles (ik, vfx, sfx, event, hitbox general colliders, anim state machine) could be streamlined.
    The worst part about the current state of things is that we cannot preview character modifications and event alignment unless we make a test scene every time. The preview window is non interactive and a glorified video player.
    Multi editor screens might be a solution as unity realizes these with the mecanim rig editor.
     
    Last edited: Oct 3, 2019
    thelebaron, joshcamas and varfare like this.
  2. awesomedata

    awesomedata

    Joined:
    Oct 8, 2014
    Posts:
    1,419
    I hear ya.

    I _think_ Timeline was probably intended for this purpose (on some level).

    Might be useful if it was repurposed somewhat for a much better "animation clip preview" for modular animation IMO.
     
  3. willgoldstone

    willgoldstone

    Unity Technologies

    Joined:
    Oct 2, 2006
    Posts:
    794
    Hi there, we're resolving this challenge with Prefab mode in-context, a feature in Unity 2020.1. Unfortunately we weren't able to get this working for 2018, but it'll be ready to try in the 2020.1 beta by end of this year. I appreciate this is a big leap in version for projects.
     
  4. awesomedata

    awesomedata

    Joined:
    Oct 8, 2014
    Posts:
    1,419
    @willgoldstone:

    Is this perhaps a step in the direction of the editor being able to handle floating-point-independent level positioning for large / endless open worlds?


    Right now, this kind of streaming positioning/design is a seemingly insurmountable pain I'm facing every single day in designing and implementing an open world -- a world that I can design in the Editor. Will this kind of thing ever be officially supported in the Unity Editor?

    I want to design my world without assistance from something like Houdini, which is SO subpar on SO many fronts to Unity's new DOTS-based processing for realtime modeling and design. Ever since I saw MegaCity, I've wanted to use that Scene-based workflow for endless streaming worlds designed right in the editor. And with VirtualTextures -- I think you're on to something. That kind of procedural artwork/design for level/world/rule generation should be possible in the Editor using DOTS and NestedPrefabs and the new Collections system you guys are working on... Am I wrong for thinking this might be the direction you guys are looking toward for the future? -- MegaCity really did create some intense expectations on that front... Open world designs (especially of the HUGE variety) are generally HARD -- and Unity aims to solve hard technical problems, right? -- So maybe I'm thinking you guys might be looking to solve the hard (and constant) problem of not being able to easily author endless worlds inside the Editor?

    Any assistance on this front you guys might be willing to spare to make my life easier in designing art (and levels) for these kinds of worlds would so quickly bring tears of joy to my eyes. You have no idea.
     
  5. SirIntruder

    SirIntruder

    Joined:
    Aug 16, 2013
    Posts:
    49
    Will you please consider adding auto-save to scriptable objects, as we have with prefabs now?

    Now we have to spam "Save project" for changes to go down to the file system / git, and of course non-technical people/designers constantly forget that and wonder why their settings changes were not committed...
     
  6. willgoldstone

    willgoldstone

    Unity Technologies

    Joined:
    Oct 2, 2006
    Posts:
    794
    Hi! I've been discussing a few of your points of feedback with various parties - can you drop me a mail on 'will [at] unity3d.com' - I'd love to setup a call, I think it'll save time and easier than replying piecemeal to many of your posts.

    For those interested in my reply to this - huge worlds are exactly what we're looking at presently. Sub scene conversion workflow like you first saw in megacity is a big part of this, new asset pipeline is another, the list goes on - but we're building a big production internally for this exact reason, to really stress test where Unity falls down, and resolve it.
     
  7. transat

    transat

    Joined:
    May 5, 2018
    Posts:
    779
    Great news @willgoldstone . Here's a small workflow pain... Scrolling through the list of Scripting Define Symbols when you have more than a dozen define symbols. It's a pain! And sometimes I accidentally copy a few characters from one symbol and move them onto another. I'd love for the UI to be improved here.



    PS - another small pain: having to use a third-party image host to share the above image with you. :)
     
    Ruslank100, Timboc, Havokki and 7 others like this.
  8. stijnvdbmgsp

    stijnvdbmgsp

    Joined:
    Feb 6, 2017
    Posts:
    6
    Hi, here's one I'm sure you've heard before:

    Compiler warnings. Each new project we start we try to keep the warnings at zero for as long as possible, which is really not all that hard. Even though the "Field is never assigned to, and will always have its default value" one is often a false positive when said field gets assigned through the inspector, we can work around that by initializing the variable with a default value when declaring it.

    Generally there are zero warnings in our project during the prototyping phase. Then we import a 3rd party package, and there's ~874230.

    We can't fix these warnings ourselves, because that will all be undone once we update the package.
    In an ideal world, all package maintainers would be blocked from shipping packages that have compiler warnings in them.
    In the world we have, I'd be happy if we could get an inspector toggle in the Assembly Definition which allows us to ignore all warnings from the selected assembly:

    upload_2019-10-18_17-26-12.png
     
  9. nsmith1024

    nsmith1024

    Joined:
    Mar 18, 2014
    Posts:
    870
    Wouldnt it be nice if you could open two Unity Editors, then be able to copy a component or the rect transform from one unity editor to the other.
     
  10. JohnHudeski

    JohnHudeski

    Joined:
    Nov 18, 2017
    Posts:
    126
    upload_2019-10-24_2-20-40.png

    This double scroll bar causes actuall physical pain. The inner scrollbar is 1 item off-screen. just 1 but it takes precedence over the outer scroll bar. I mean we can reduce the image to give that extra room for the button

    This is the entire length of the scroll bar
    upload_2019-10-24_2-23-28.png
     
  11. Ne0mega

    Ne0mega

    Joined:
    Feb 18, 2018
    Posts:
    745
  12. joshcamas

    joshcamas

    Joined:
    Jun 16, 2017
    Posts:
    1,276
    This.
    This has gotten worse and worse over the past few years, it's getting unbearable.
     
    oxysofts and awesomedata like this.
  13. awesomedata

    awesomedata

    Joined:
    Oct 8, 2014
    Posts:
    1,419
    Also, this:


    There's the rest of my reply:
    https://forum.unity.com/threads/addressing-overall-problems-vs-specific-solutions.787007/


    I cringe when I see these kinds of responses from Unity's own internal developers.

    I originally pointed this out in the SceneVis thread, but at their request, they shifted me around to a new topic because this was a "more general" issue with Unity's tool solutions, so I figured I'd point it out in case anyone else wants to chime in.


    Honestly, the "programmer culture" (i.e. just make your own tools!) conflicts heavily with traditional artists and designers who deeply struggle to remain technical enough to keep up with Unity's ever-changing API landscape. Being such an innovative and multifaceted tool, Unity's internal developers are going to have to either learn to respond more flexibly to artists who use the software to design, or make many complex, innovative, and customizable tool "handles" that work for all sorts of 3d manipulation purposes to make in-scene tool creation much more accessible and robust in general, and put heavy training emphasis on scene-based workflows.


    Saying "we didn't design the tool to do this" just isn't valid to an artist --
    -- Did anyone tell Bob Ross he can't use a cake decorating scraper tool (the palette knife) to paint his "Happy Little Tree" branches?

    Artists typically grab the first thing that even remotely looks like it would solve their problems in a general and convenient way (see the sales on the Asset Store for art tools). When a tool doesn't work for us in certain scenarios, we either seek to improve the tool (trim it here, tweak it there), or find a separate one. In instances where we've got to go and manufacture our own tool from scratch (and unless we're planning a business empire already), we will typically draw the line there if at all possible. Only a few brave souls venture beyond this point.

    See "Peek" on the Asset Store.

    The developer did a great job, but is running into terrible issues (like some of the ones I have had with Snapcam) when using some of the native Unity (internal) functionality, which breaks scene-based tools like ours regularly.
    A scene-based workflow is obviously not supported, and yet, if you look at "Dreams" on the PS4, it has a highly robust and flexible, artist-friendly, UX / UI in-scene (I thought about "Peek" when I saw the workflow). I had almost made a tool like Peek or Dreams for myself, but after running into issues with Snapcam that I am still waiting to one day be resolved, I wasn't even going to open that Pandora's box of nighmares. :/

    The editor tooling for artists still needs improvement -- SceneVis is one area that needs to be improved, but as mentioned in my link above, for this to be possible (and to avoid shooting yourself in your foot in each and every step!), you guys need to reconsider your design mantra to account for artists.

    See my topic -- I go into detail on this exact issue. (But there are more.)
     
    Last edited: Dec 3, 2019
    Lars-Steenhoff likes this.
  14. Ne0mega

    Ne0mega

    Joined:
    Feb 18, 2018
    Posts:
    745
    I about lost my... ...cool... when I first ran into this. However, you can start to type out the name you are looking for, and it will take you right to it.
     
    artemio_morales and awesomedata like this.
  15. Flow-Fire-Games

    Flow-Fire-Games

    Joined:
    Jun 11, 2015
    Posts:
    305
    In 2019.3, unity still decides to place new objects at random locations on screen, instead at 0,0,0 of selected parent object/null object. Even in prefab scenes and even if you add a primitive from the top bar in a prefab scene.
    That feels very low quality but is in for many years now and still exists with the editor rework, please fix this weird and buggy looking behavior.

    It takes 3 clicks and 3 button presses every time each object placement to correct it.
    At least please add a zeroing button on the transform. There must be years of manhours wasted on zeroing objects globally.
     
    Last edited: Dec 5, 2019
    awesomedata and SisusCo like this.
  16. nsmith1024

    nsmith1024

    Joined:
    Mar 18, 2014
    Posts:
    870
    Hello,

    I have a large Unity project and i want to be able to enable or disable debug statements and also find out which line those debug statements were printed at.

    For example

    1.public class test1 {
    2. x=10
    3. MyDebug("this is a debug statement"); // <<<
    4. y=5;
    5. }

    1.public class test2 {
    2. x=10
    3. y=5;
    4. MyDebug("this is a debug statement"); // <<<<
    5. }

    1. class debug {
    2. public static void MyDebug(string msg)
    3. {
    4. if (true) {
    5. Debug.Log("msg");
    6. }
    7. }
    8. }

    When you run this code, it will print 2 debug lines in the unity console, one from class test1 and the other from class test 2.

    The problem is that in the unity console when you click on either of those debug output statements, the unity debugger will open Visual Studio and take you to the debug class line 5 because that's where the actual debug print statement is located.

    But i want it to actually take you to the original lines which are class test1 line3, and class test2 line 4.

    To do that i want seems i need to INLINE the debug MyDebug function so that when you click on the debug statement in the unity console, it will take you to the test 1 class line 3, or the test 2 class line 4.

    I have thousands of this type of thing in my project, and its a real pain because when i click on the debug message in the unity console, it takes it to the same line in the debug class even though that is not the correct place, i have to look at the stack and then figure out who called that line and then manually open that file and manually go to that line, it just takes a lot of time to do all that.

    Is it possible to get taken to the correct file and line when i click on the debug message in the console?
     
  17. Baste

    Baste

    Joined:
    Jan 24, 2013
    Posts:
    6,294
    If you're in 2019 or later, you should be able to just click the next level up in the callstack, as 2019 added links to all the Debug.Log callstack entries.
     
  18. Aeroxima

    Aeroxima

    Joined:
    Jul 19, 2016
    Posts:
    15
    I really just want to be able to set a key for going up/down in "fly mode" in the sceneview, in world space rather than relative to the current rotation. It seems small, but it's really annoying trying to finesse it where I wanted and getting OCD over it.

    It seems like the easiest fix in the world and it's the first time I wish I had access to the source, because I imagine it's just changing one variable out for Vector3.up or Space.World. This would make it match the standard FPS controls games always use, too. (I literally have not once seen the sceneview's idea of up used in any game I've played or watched in my life, save for the very rare case of weirdly implemented jumping.)

    I made a thread elsewhere that nobody cared about, and can't find a good place to get the input in an editor script where it's not just a shortcut (has numerous issues doing it that way of course), so I'm about ready to give up on it and just deal with the annoying controls. I just can't shake the feeling that it should be an easy fix, and somebody more knowledgeable could have had it sorted by now, but it's just barely out of reach.
     
  19. Ne0mega

    Ne0mega

    Joined:
    Feb 18, 2018
    Posts:
    745
    This one may require too much restructuring, since it requires cursor position feedback, but Blender makes ctrl+C ctrl+V a breeze by being able to execute those commands and copy values simply based on what editor fields the cursor is over. No double or triple clicking to highlight a field.

    Having this little goody in Blender makes working in Unity feel like pain. It saves a ton of time compared to "highlight, go into rename mode, ctrl+C, highlight other, go into rename mode ctrl+V" takes about 5 to 10 times longer.
     
    Vectrex likes this.
  20. Baste

    Baste

    Joined:
    Jan 24, 2013
    Posts:
    6,294
    Oh jeez, it might be because it's not something I use a lot, but every time I use blender, having things happen based on my mouse position instead of what panel is focused drives me up the walls. It's the worst. I move my mouse away from something I'm going to interact with so I can see it.

    But I guess it could be a mode?
     
  21. oxysofts

    oxysofts

    Joined:
    Dec 17, 2015
    Posts:
    124
    Wow finally it seems Unity is taking notice to the SHOCKINGLY unchanging UX of their editor. The workflow has been virtually unchanged for so long that the basics must be reviewed. That means the hierarchy, scene and asset browser. I will try to make a more detailed list on these subjects, but right away the biggest most insane feature that is missing out of the box in the asset browser is cut/paste. I was working with my brother recently on something and I witnessed how he left clicked and then moved his left hand to scroll the wheel down to the folder he needed, basically 2-handing the mouse so he could comfortably drag and scroll the assets and perform a long distance movement. Does that sound acceptable? This is overall very representative of the style of dysfunctional workflows in Unity, the tip of the iceberg really.

    Another horror I can quickly think of off the top of my head is having to move the player (an object) all the way to the other side of the map in edit mode to play a specific part of the map. Two ways I know:

    1- Zoom out the scene view a lot until you find the object, without losing sight of your destination place in the map. Drag the object while holding shift for collision snapping.

    2- Rotate the scene camera towards the object (if you know where it is) and then drag while holding shift.

    Why no context menu in scene view? It could be extendible and powerful, with searching too. Unity is creating such wonderful little plugins like Quick Search, we could do the same for other parts of the editor that are severely lacking.

    With a context menu, it could be as simple as right-clicking a location on the map and clicking "Move selected object here"
     
    Last edited: Dec 10, 2019
  22. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    Some of the tools I work with to take out some of the workflow pain, and are currently missing in unity.

    It hard to see a lot of them getting deprecated because they are time-consuming hard to maintain as unity changes so much.

     
    Last edited: Dec 12, 2019
    senkal_, tatoforever, transat and 2 others like this.
  23. joshcamas

    joshcamas

    Joined:
    Jun 16, 2017
    Posts:
    1,276
    That object groups asset is something I really, really need. I've been meaning to build one myself.
    My project folder consists of hundreds and hundreds of scriptable assets, models, and meshes. Trying to world build with all of that junk alongside it is such a pain.
     
    Lars-Steenhoff likes this.
  24. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    Exactly, It's what I need too, thats why I worked it out in concept, Now I need programmer help to make it happen, best case unity would make this native
     
    Last edited: Dec 10, 2019
  25. willgoldstone

    willgoldstone

    Unity Technologies

    Joined:
    Oct 2, 2006
    Posts:
    794
    Hey again Lars,

    Great to see you at Unite recently, we've been discussing your suggestions a lot internally. Thanks for summarising here it's super useful for us.

    Cheers

    Will
     
  26. joshcamas

    joshcamas

    Joined:
    Jun 16, 2017
    Posts:
    1,276
    One annoyance that really slows down using the project window is the fact that the filters (tags and type) simply adds the corresponding filter to the searchbar. The reason why this is incredibly annoying is that you can't simply set your filter and go on your way, you have to be very careful not to remove the filter in the searchbar as you search.

    I'm not saying "filter codes" isn't a good idea, those should stay. But the GUI should handle GUI, in the sense that the filter buttons should enable a filter on top of whatever filter codes there are, as opposed to simply setting them.

    This would allow for us to have multiple project windows, one with no filters, and another with filters, without the annoyance of being able to mess up the "filter codes".

    This alone would greatly boost the usability of the project window during world building.
     
    JBR-games likes this.
  27. awesomedata

    awesomedata

    Joined:
    Oct 8, 2014
    Posts:
    1,419
    I actually did something better than this in Snapcam.

    Basically, I use the middle-mouse drag (hand tool) to move the scene camera viewpoint up/down in World Space (rather than screen space). Unfortunately, this must inherently stop the camera motion by default. I would sometimes like to continue moving while dragging too though, but this would require another redesign/recode of the scene camera controller. This sucks because I had already needed to rewrite the entire scene camera controller to get the first behavior. I HATE that I cannot have access to override the scene camera controller's methods/shortcut-functionality/etc. It seems so simple, yet so vital to be able to do this... There are so many different scene camera controllers I can think of for so many different kinds of games. These should even be able to translate to in-game camera controllers if we wished.
     
    Last edited: Dec 13, 2019
  28. awesomedata

    awesomedata

    Joined:
    Oct 8, 2014
    Posts:
    1,419
    Also, @willgoldstone, I second pretty much all of Lars' suggestions -- Here are a few of mine:

    • more immersive "in-scene" workflows -- (see "Peek" asset from Ludiq -- then see "Dreams" on PS4)
    • better "in-scene" procedural tooling abilities -- i.e. specialized "handle" widgets for complex positioning during level design or animation with handles like 1.) a spline-repeater handle for authoring fences or wall-geometry -- 2.) face/vertex/edge/quad-level handles for operations on portions of these, including operations on the center of edge (or selection of edges) -- 3.) using animation curves as a handle to position things in the editor -- 4.) Custom procedural handles using something visual like Timeline / Events / Scripts / Animation Curves to author them! (This is what Houdini's "Digital Assets" are for, but Unity can do "Houdini" better than Houdini can, especially for realtime procedural authoring, assuming it really wanted to!!)
    • better support for LOD/Imposters -- (Generating a MegaCity-styled HLOD via Prefabs, Imposters, Automatic LOD generation) -- Remember that "performance by default" thing you guys promised? This is kinda part of that... D:
    • streaming scene visibility / editing capabilities -- (handling editing of HUGE streaming worlds in editor that go past floating-point limitations -- See my post above or in the SceneVis thread describing this)
    • automatic positioning of imported models to placeholder locations in editor, independent of whether scene is loaded -- (i.e. for openworld landform assets constantly authored or updated externally)
    • more general way to map and iterate/operate on hierarchical collections of things -- i.e. prefabs, instances, LODs, textures/materials, shaders, variables holding shaders that might need to be swapped with LODs, tiles, environments, layers, (or other custom concepts) so you can easily combine/sort/show/hide or perform operations on these, letting you act upon everything you want to act on together, quickly, simultaneously, and automatically, without having to dig into the API constantly -- (i.e. make a "Hierarchy Map" data type / concept to let us apply to various groups of assets / entities / textures / materials / etc. any operation so that you can, for example, keep your hierarchical database of assets/prefabs/objects/variables/classes/textures/materials/concepts easy to access through code -- This would aid in the creation of things like tools or scene visibility maps (SceneVis) or material maps that help ShaderGraph know what textures should be batched and atlased)
    • more and better control over shortcuts and hardware input in the scene via shortcut manager for editor tooling -- (i.e. pressing letter keys should have the option to be repeated at the repeat rate of the keyboard OR the repeat rate of the event, and we should have access to pen/stylus input that should be mappable to mouse movement, as well as mouse flicks or gesture support for tablet-based operations. Though, beyond this, things like VR or Wiimote/Sixaxis motion controls should be supported in the editor tooling and would be even more useful than a mouse/keyboard in certain design applications!)
    • workspace-centric custom shortcut manager and assets -- (i.e. I don't care about the scene camera controls or other windows when I'm working heavily in a behavior tree graph or need a specific view of a gameobject so I can rig it using some other asset. Probuilder, for example, is not something I need access to when I'm dropping instances around my level, and as such, it gets in the way when I'm working in the scene view. This is only compounded by other assets. However, Snapcam/NAVIGATION STUDIO, is still useful in that context, so it doesn't need its own workspace in general.)
    • procedural animation rigging needs to be bumped-up in priority for a better interface -- (i.e. hand-authored animation is going the way of the dodo for games, and motion-matching is only part of the solution -- procedural animation, that is modular rigging using a visual state and behaviour graph to handle logic, is the future. If you look at "Dreams" on the PS4, they have already beaten Unity -- a proper game engine! -- to this!)
    • override editor's scene camera controller with custom shortcuts and functionality -- (i.e. Snapcam/NAVIGATION STUDIO in my sig has an alternate version of the editor scene camera controller, but I've had requests for making it work in so many different ways that it currently does not, but I would have to rewrite AND work around the entire default Unity scene camera controller's current behavior to do this. This is not something I want to have to maintain when Unity inevitably breaks my asset again. On top of that, I'd like to be able to switch between different editor camera modes -- i.e. a topdown tiled area camera that fits the current gameobject within its screenspace bounds so that I can easily view a single-screen area in a topdown game, for example, and shift between the screens with a keypress)
    • hybrid voxel terrains for caves/overhangs -- (i.e. see Digger Pro asset.)


    Except for the very last one (Digger), there are no assets or standard workflows available for these kinds of things yet, and much of this stuff cannot be easily implemented by your typical developer.

    Speaking of assets though, something like "Peek" or "Snapcam" for example still require a lot of reworking because they tap into lots of internal editor stuff just to make their cool features possible. When the editor changes, this stuff breaks -- which can be very VERY frustrating for users and tool creators alike. It makes everything have an air of "unprofessional" -- and it stinks. :/
     
    Last edited: Dec 29, 2019
  29. Marcos-Elias

    Marcos-Elias

    Joined:
    Nov 1, 2014
    Posts:
    159
    I made a thread about my pain in another section, it would fit here.

    https://forum.unity.com/threads/more-simple-options-at-exporting-package-dialog.795624/

    Basically I would like to have more control over the asset exporting process, because I use it a lot to sync changes in my projects between desktop, Android and iOS platforms.

    I miss an “Export package (without dependencies)” menu that could save tons of hours, and a “no scripts” button, useful when I just want exporting a scene content or prefab with its artistic dependencies, not code.

    I use it everyday, these simple actions would improve the overall experience a lot. To export huge parts of the project it is a pain unchecking undesired things. And version control is not great in my use case since there are assets that I don’t want to replace on mobile versions, like resized textures, deleted folders of pc content, and so on...

    What do you think about it? Simple options to implement that would improve usability.
     
    Lars-Steenhoff likes this.
  30. illinar

    illinar

    Joined:
    Apr 6, 2011
    Posts:
    863
    Why is it necessary to make me dig trough this to find my script:
    upload_2019-12-28_16-1-43.png

    I don't care about external code, ever! I never need Engine files and their full directories. I only care about my scripts. I never in 9 years needed a full stack trace. Never. And I spent a lot of times digging through conslole like this to find my file.

    Can you at least highlight my files in a different color?

    There should be a good simplified mode for displaying this. Most of this information is never needed.

    A line of clickable file names would suffice for the simplified mode:

    DynamicBuffer:123/DynamicBuffer:41/ComponentSystemBase:233/MyClassHighlightedInADifferentColor:24/...
     
    Last edited: Dec 28, 2019
    Ruslank100, freso, stonstad and 5 others like this.
  31. nsmith1024

    nsmith1024

    Joined:
    Mar 18, 2014
    Posts:
    870
    Here is an easy feature to add to Unity. Being able to search in the console window.

    For example you run your project in the editor, then it generates lots of debug statements, you are trying to find a particular debug statement while debugging to see if your code printed it out. There is no way to do that unless you open the debug text in external editor and then search that. But its inconvenient especially if you are trying to debug something that is running where its constantly generating more and more debug output.

    What would be cool if you could enter a search string into the console window somewhere and then it could search for that string while the editor is running, similar to how the console log works in android studio. In android studio you can filter the debug statements and even search while the debugger is running, it would highlight or jump to whatever you search for instead of you trying to eyeball through hundreds of debug output statements.
     
    Last edited: Dec 28, 2019
    illinar likes this.
  32. Ne0mega

    Ne0mega

    Joined:
    Feb 18, 2018
    Posts:
    745
    There is a search bar in the console window.
     
  33. nsmith1024

    nsmith1024

    Joined:
    Mar 18, 2014
    Posts:
    870
    I have Unity 2018.3, i dont see any search bar on mine
     
  34. Neto_Kokku

    Neto_Kokku

    Joined:
    Feb 15, 2018
    Posts:
    1,751
    When you build a player, Unity processes scenes, assets and shaders before actually trying to compile the code for the target platform. Obviously, the odds of code compilation errors are usually much higher than some problem with asset packaging, while asset packaging often takes the bulk of the building time.

    Would it hurt that much to compile the player code before packaging the scenes? This specially irritating when building for a different platform for the first time, because you cannot use the "scripts only" option unless you have a successful build, and you cannot have a successful build until you fix all your script errors, which means the iteration time when porting a project to a different platform is much larger while you are in the step of fixing scripts for the first time.
     
    joshcamas likes this.
  35. joshcamas

    joshcamas

    Joined:
    Jun 16, 2017
    Posts:
    1,276
    I definitely agree. If there is some sort of reason for asset scanning before script compilation, it would still be nice to have the option to have an additional script compilation before asset building.
     
  36. tatoforever

    tatoforever

    Joined:
    Apr 16, 2009
    Posts:
    4,364
    My top painful list:
    - Reduce Editor critical bugs/instabilities (this should come first before any additional tool/feature)
    - More rapid level design tools such as instanced mesh painters or any tool that will improve the QoL of artists and designers
    - Terrain tools such as Road/road-side/River tools
    In general, people should not rely on the Asset Store for basic level design tools, this should be built-in by default. But again Editor bugs/instability should come first
     
  37. transat

    transat

    Joined:
    May 5, 2018
    Posts:
    779
    Totally agree @tatoforever . Editor stability should come first. Here's a screenshot of my editor from just a few minutes ago...



    This hasn't happened in a while but I know that when it does I shouldn't panic and instead keep on working until all the text goes suddenly back to normal after a couple of minutes. :)

    The whole Terrain component UX is quite dreadful. It's annoying to have to go looking for hidden-away menus to do basic things like painting textures/layers. Why not have a nice toolbar in the Scene view?! And there should ideally be many more options available for manually and automatically painting the terrain with trees, grass, details and objects (based on height, slope, cavity, map, underlying texture %, etc.) These could be hidden behind an "advanced" checkbox.

    Also, on the topic of terrain tooling pain - I'd love to be able to drag prefabs straight into the Trees and Details tabs of the Terrain, rather than having to click half a dozen times.

    There are third-party tools that fill this gap but one problem is that they can't keep up with the newer Unity versions and pipelines as the developers are wary of potential breaking changes. @tatoforever, case in point: I loved your soft shadows asset so much - it contributed a lot to the look of the game I've been working on. But I lost compatibility with your and other assets when I switched to URP. To focus on the level design UX stuff, I think Unity would do well to hire you and a few of the low-key but super-talented developers of other top weather, level, and terrain design assets and start a special dedicated "Open world environment team" focused on making all this interesting tech talk to each other seamlessly... and up-to-date with the latest official Unity versions / packages!
     
  38. tatoforever

    tatoforever

    Joined:
    Apr 16, 2009
    Posts:
    4,364
    @transat
    I will be happy to modify the current built-in renderer and integrate NGSS way deeper but don't think Unity will do that at this point. They have plans to deprecate built-in renderer and they are not interested on improve it anymore. As for URP, don't worry I'm getting there. Just trying to finish out few other stuff first (related to NGSS).
    But yeah, I do also believe Unity should hire a couple of talented folks from the Asset Store and deeply integrate some of their plugins/tools into Unity but only if they won't add to the current instability (check out if the integration can be done in isolation)?
    PS: The constant changes on the SRP is what kept me from releasing NGSS on it (too many rewrites, lot of wasted time, etc). Hopefully things will stabilize in 2020, I'm dying to let everyone taste my new half-omni shadowmaps on URP! ;)
     
    Lars-Steenhoff and transat like this.
  39. chrisk

    chrisk

    Joined:
    Jan 23, 2009
    Posts:
    704
    Here is one simple, old, dumb, annoying bug that has been laying around for years.
    Unity is often getting frozen I have to force-kill, however, there is no way to tell which is which. (image attached and can you tell them apart?)
    Having a project name on the process will help.


    Here are other annoying bugs that are really old.

    1. Scrollbar along "Reorder Script Execution List". It's almost impossible to use it when you have ~100s scripts.

    2. Mouse forward/backward to move through selection history. Fundamental to 3D editor especially for Unity since it lacks custom inspectors. Using a global inspector and loosing the last selected object often causing pains. There is undo already and why can't we move back to the last selection?

    3. Automatic Animation Play option when previewing animations. When you are previewing many animations, it's so annoying having to click on "Play" each time.

    Many are common sense and don't require thinking.
    If Unity worked in large production, it will cause enough pains and they already have fixed it, but it's unbelievable that it still works like 10+ years ago.

    Oh well, I expect the least but it will be a miracle if Unity does something about it.

    upload_2019-12-31_20-40-20.png
     
  40. willgoldstone

    willgoldstone

    Unity Technologies

    Joined:
    Oct 2, 2006
    Posts:
    794
    Hey @joshcamas we are working on something to solve this for the entire editor - a new query system. We have already built some of it the past few weeks and found a super performant way to do it which is great. We will build on top of this a system for making grouping in hierarchy and project such that you can simply filter things and let Unity handle the rest as your objects / assets grow and shrink. Thanks for raising it, it validates our ideas further.
     
  41. Neto_Kokku

    Neto_Kokku

    Joined:
    Feb 15, 2018
    Posts:
    1,751
    At work I'm involved in projects in many different engines. Some of the things I miss the most in Unity after a few days days in UE4 are the robust asset management tools: being able to see which assets reference a particular asset (without having to use Notepad++ to search for the guid in all files), seeing asset details like texture resolution and polygon count directly in asset listing as columns and edit their properties in a table (so I can quickly find and correct oversized assets).
     
    Peter77, awesomedata, Ferazel and 2 others like this.
  42. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    awesomedata likes this.
  43. awesomedata

    awesomedata

    Joined:
    Oct 8, 2014
    Posts:
    1,419
    A shout-out to this topic here since it is 2020, and 2020 is all about addressing issues with "large" worlds.

    Here's an excerpt:


    I mentioned this to @willgoldstone at one point. He responded rather clearly that Unity aims to support large worlds in 2020, though I don't know specifically what "large" means to Unity in terms of editor tool support since I've heard no concrete plans for "floating origin" support in the editor, and none of the SceneVis team (for example) seem to think in-editor visibility of far away areas of a huge world is a critical task to be discussed.

    I beg to differ.

    We have the mysterious new "Environment" system upcoming that has a "layers" concept, but as far as I understand it, this is still just heightmaps (which is still just INSANE in 2020, I might add). No ability to setup an array of tiles, sculpt them as a mesh in something like Blender/Zbrush (i.e. holes/caves overhangs), reimport it, then having the new mesh sit firmly planted at a 10000-15000 meter position in the world. Without jitters/artifacts from lighting/animation. Then, using editor tooling, sculpting slabs of geometry for very distant buildings or other landscape details via Probuilder (or other assets) at a 5000+ meter distance using an asset like Snapcam to zip around the world instantly to tweak it by hand, having the floating origin be taken care of for me by the Editor itself -- something akin to how it's being done with Virtual Texturing, except I don't think anyone is currently thinking through the whole workflow situation quite this deeply.

    The editor coordinates themselves being a pivot point for multi-departmental tools (and future toolsets) is kind of a new road -- However, it is also essential and critical to get this right from the outset. For example, it's already a bad start when I know SceneVis isn't doing anything to mitigate the inevitable visibility issues that will arise with huge coordinates for object/entity locations in open worlds, even with the task of gameobject/scene "visibility" as their core principle. The new Blender-like "Collections" system they're working on happens to be about visibility, but it overlooks the basics of what that "visibility" needs to look like in a 2020.x release focusing on large worlds. Because of this, I'm not convinced the design work has been laid out across the departments at Unity to take on a feature of this magnitude (and the implications it has) -- though I could be wrong. I would love to be wrong. @Unity, please tell me I'm wrong?
     
    Marcos-Elias and Lars-Steenhoff like this.
  44. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    Good points!

    For me the today biggest issue is texture compression, it took me the whole day to compress the textures for ios, and in the meantime I could not work on my project.

    Please prioritize background compression without blocking the editor.

    It completely takes me out of the workflow.
    Thanks!
     
    Neto_Kokku and awesomedata like this.
  45. Marcos-Elias

    Marcos-Elias

    Joined:
    Nov 1, 2014
    Posts:
    159
    Just an origin shifting like Unreal would solve 99,999...% of problems with both rendering and physics for open world. I just don't understand why Unity does not care about it at all. Physx has the shiftOrigin function, if Unity had the source code available I'm sure I could implement it.
     
  46. Neto_Kokku

    Neto_Kokku

    Joined:
    Feb 15, 2018
    Posts:
    1,751
    Another missing piece of the open world puzzle is dynamic occlusion culling. Right now the only native occlusion culling solution is the pre-baked Umbra occlusion, which is not compatible with any DIY world origin shifting and does not support streaming (the occlusion data for all sub-scenes must be baked at once and stored into a single file).

    Hardware occlusion queries is pretty much a necessity for good performance on open world games. And we are in 2020, open world is not AAA-only anymore, with small teams and indie devs making such games at a regular basis (with great pains if they happen to use Unity, but still).
     
    Marcos-Elias likes this.
  47. Marcos-Elias

    Marcos-Elias

    Joined:
    Nov 1, 2014
    Posts:
    159
    When I tried occlusion culling with static smaller maps my experience was bad, I got less fps and needed more time to bake things. I don't know if I did something wrong, by the way I avoid using it at all. On mobile I found that the processing needed to OC to work was worse then just showing all the polygons that it saves from rendering.

    There are other solutions based on raycast, or it may be possible creating custom solutions that disable entire game objects based on distance or trigger. I find they easier to code, and provide also good performance with zero baking time to disable interior of buildings. This is generally used in procedural worlds too.

    I don't miss it too much also because my project is a driving simulator, so I need rear mirrors, there are objects to draw in every direction here. Raycast based OC had worse performance in this scenario, but also most of my buildings don't have interiors.

    For projects like buildings with floors and internal objects, disabling/enabling them with triggers or distance works fine for me.
     
  48. awesomedata

    awesomedata

    Joined:
    Oct 8, 2014
    Posts:
    1,419
    This is a very good point.

    This is why I think it should be handled specifically in-Editor and have occlusion (and lightmap or lightprobe) data baked in the background (along with physics?) dynamically for individual structures. Maybe using a hook into the new Asset pipeline to handle when the baking needs to happen.



    @willgoldstone:

    I'm not sure if you've considered this, but that huge file kills the editor on a regular basis for me as I make updates to the landforms and physics in my open world game.

    My use-case is unusual (but not far-fetched) because not only do I have one set of landforms in my open world, but I actually need each of my landforms (i.e. meshes, not terrains) to have multiple shapes and will change to these to different shapes as the game progresses. I would need both Occlusion Culling _and_ Lightmaps (and maybe physics?) to be baked for multiple versions of my open world's individual landforms. I would basically have "sub-tiles" for every tile, which would each need their own openworld data (and imposters). I'd be willing to explain this in detail if you're interested, but as you can probably tell by this, I tend to push engines to their absolute breaking points with my game designs. I think it's a gift! You should probably find some way to hire me to create (realistic) game designs that will specifically break Unity's different components (so that you can learn how to unify them better!) -- I can do it! -- This is of course if you really want to improve Unity and make it future-proof for both games and interactive technology. I love both. I can make any design you have fall apart at the seams (if it is flawed), show you exactly how those seams arose, and then show how to redesign it in a fundamental way that makes it not only future-proof, but also much more holistic and structurally-sound (and look nice!) across the whole product -- all in amazingly creative and clever ways. I'm just saying. You guys need me. :)



    This, I can agree with. This is why the Unity Editor makes me cry most days. :'(
     
    Last edited: Jan 7, 2020
  49. Neto_Kokku

    Neto_Kokku

    Joined:
    Feb 15, 2018
    Posts:
    1,751
  50. Lars-Steenhoff

    Lars-Steenhoff

    Joined:
    Aug 7, 2007
    Posts:
    3,521
    Very happy to see this window in 2020.1 18

    Screenshot 2020-01-08 at 15.19.30.png