The interface is an extra step if you want to avoid coupling and it can be a bit easier to call a function, since you only need get_game_mode. There is also an argument about exposing only necessary functions this way. This was an example of a very specific way to create singleton, but most of the small/medium games do not require abstractions like this. In my videos I like to show random and interesting ways that you can do something, so I wouldn't use any particular example as a gospel or THE way to do something, more like just another way.

​@@petar4621 Could u check the following use-case please? Actor A is my Player Controller Actor B is a Sword [Interact -> Take] Actor C is a Pen [Interact -> Take] Actor D is a Door [Interact -> Open/Close] You would use a Interact_Pickable Component for B and C, but not D. While Implementing Intractable Interface to B C and D. In Case of B and C you would call the Component functions from the Interface-Implementation. While on D you would call the Actors Native Implementation from the Interface-Implementation. This is actually the one and only way I could now image an use of both abstractions. Though, using Abstract Classes would be more preferable from my side of view - especially for this case.

@@user-rg7ux5hn5t There are many ways to set up interactions, but in this case it seems like an interactable component is an overkill. Why not just have interface and interact function? Another way might be to use some sort of inheritance if that is something you prefer, like a interactable base class. You can also have just one class for all interactable items that stores data about the item and a component in player controller(eg. interact component) that gets the data when interacted with and figures out what type of interaction it is. Usually you would have limited types of interactions in a game(pickup, play animation like opening doors, interact with NPC,...) The main reason for the interface and component in the video was just convenience and decoupling if needed, but in this case it seems like it would just overcomplicate it.

@@petar4621 What would you tell me to the following configuration: It's very interesting, as I am searching for a way to shrink the amount of code I have to write. I do have about 10 Items, those all inherit from actors, that are not intractable. All 10 Items do not relate to each other. I also do have an Interaction Component, that has toggle-able availability of interactions (Pickup, Rotate, Scale, Colorize, Default Interaction etc.) So not all interactions are available to all items, but some might be. I do have all the common interactions implemented in the component, and everything that is not commonly used, is triggered by delegates on the items individually. I guess, if I use one Intractable Interface, I would be forced to implement every kind of interaction, that has to be available for each item individually, what would result in a copy-paste, or a library as a dependency. ' That library dependency would break the whole decoupling, in the same way components do it already now - wouldn't it?

I haven't tested the performance impact of duplicating variables, but usually you want pass by ref. Just keep in mind that changing that reference is changing the original variable.

I have and its good, I am just waiting for it to become more user friendly and to be fully exposed to blueprints. It is more versetile than CMC and has better replication, but it is also way harder to use, since right now you have to create every movement mode from scratch(you have to do all of the input offsets, physics, solve issues like wall sliding, step up,...).

@ ey I didn’t realize it was that raw still where you have to solve physics and sliding issues. I’m developing an arena shooter like Quake but more intense movement. Do you think I should go the route of customizing the CMC, try Mover, or maybe spend on the GMCv2 plugin? I made basic gameplay with GAS but am not sure if I should pivot to something else for advanced movement features 🤔

@@everything-bagel For quake type game, CMC is probably fine. CMC mainly starts breaking when you need more physics based character, or when you have to add a lot of custom solutions on top of it. Mover is great in concept, and they are providing some functions to help with different components of how a character works, but it is more like legos, where you have all of these small pieces and you need to slot them together. CMC is more monolithic, but solves a lot of those smaller things by default, but extension or modification of CMC can give you issues. Not sure for GMCv2 though, never tried it.

@@petar4621 Thanks for the response. The game isn't planned to be physics based, however I do want several movement abilities that are outside the default CMC: air dashing, wall running/jumping, flying (like in the game Spellbreak), drop-stomping out of the air (like in Ultrakill), sliding. It is the issues with modifying the CMC that you described that is giving me pause at the moment. I'm a software engineer but I can imagine it can be a large task to implement custom movement if I have to account for a lot extraneous networking issues. I was hoping Mover might make things easier but it sound like it is not there yet? Perhaps I just need to work deeper on CMC and see my own limitations in developing what I want.

@@everything-bagel Mover is in early stages, so its probably a good idea to wait for 5.6 and see if it is viable then 👍

Thanks, I am not a fan of generic videos, so when I make these I usually put more niche stuff or something that is actually interesting to me.

@polysiens UE desperately needs more niche and advanced tutorials. Every other UE youtuber repeats the basics 20 times for each advanced topic they cover. Hopefully you get a bigger audience this way since if you did beginner topics, you would have to compete with thousands other youtubers and now you have next to no competition

@@tommydell9191 Yes, there is plethora of beginner content and not enough intermediate/advanced. To be fair the pool of beginner devs will always be way bigger than other categories, so I see why the youtube is saturated with basic or beginner videos.

Should be pretty soon. I think they said at the end of october.

Hi, I don't speak russian, only serbian and english.

@@polysiens I get it. You just sound like a russian speaking English)

I also would like to see that :D

The issue is with creating my own game instance. I just reset it to the default game instance and it works, will have to figure out how to do the rest outside the game instance, or create the game framework component manager within the game instance maybe.

You always have to cast directly to the object that has the dispatch event, otherwise you can't bind events. Towards the end I even mention that casting is necessary for binding and that it is one of the downsides.

@@polysiens no, you can use actor components to call dispatchers without casting directly to the actor. Sorry friend, but I believe you need a tad more experience on this matter. Tutorials like this will only teach bad practice.

@@koldx111 If you have an actor component attached to the actor, that means that actor has direct reference to that component. If you want to reference actor component that is on another actor, you have to reference directly that actor to get access to its components(Or use get component of class, which does not reference actor directly, but it does reference the component directly). This is the reason people try to get around this direct reference issue by creating Event components, so everyone directly references the Event component, but not each other. It still needs direct reference though, but at least that event component is lighter than the actors that are referencing it. I am not a fan of it because it exponentially increases complexity and adds a layer of misdirection.

If you are telling someone that a code is wrong just because simply it is not favorable to you, then that's literally what programming is. There are multiple ways to get the same results.

@@easternunit2009 I agree, that is why I gave pros and cons for event dispatchers and where they are useful. This was a video about event dispatchers specifically, so even though you can create different configurations, the video was focused on basic way to use event dispatchers for people that are not familiar with them.

Can you expand on that? Not sure what exactly you mean.

No problem, thanks for watching 👍

Super Cool to see actual impacts

Thanks, I generally don't like to just talk conceptually and try to ground it with examples. It does make my a bit longer than I would like, but hopefully tradeoff is worth it :D

@polysiens a helpful suggestion to help grow your audience. Make youtube shorts of the key stuff and quick did you know, tips, i.e snippets of your main long videos. As well as other short just 6 to 8min quick tips..those get quite good viewership via algorithm. That way you can capture a vast majority of your unreal user audience as well as new unity engine users transitioning over. Hopefully other users are sending you suggestions too which could really help with direction viewership and elevate your channel. All the best 👍 yes totally worth it

@@warren01 Yea I was thinking about shorts(maybe less than 6-8, that sounds more like a short video), but for me issue is time to make and video type. Since most of my commentary is made for long form with a lot of context that you need around it, I would probably have to record separately for shorts to make it coherent, which takes time. Maybe at some point I will hire an editor or something. I have a general idea on how to grow more, which is make more generic, shorter videos that appeal more to beginners. When I look at my analytics, most of the best performing videos will be a tutorial of a basic X feature in Unreal. The issue with that is that I don't find it fun to make those type of videos, but appealing to more advanced users is basically me targeting 10-20% of the Unreal community and cutting out the rest, which could still work, but is way harder to appeal to and get.

Yes, you can check out my fiverr profile in the desciption 👍

@@polysiens I've never used Fiverr - is Discord / PayPal an option?

​@@elganzandere I don't do direct payment, sorry.

Since unreal doesn't know what struct you stored or want out, they have to give you a wildcard, so makes sense you have to specify.

@@polysiens wildcards can be refs but if its not we cant use SetMembersInStruct this is what i mean

I've never heard of blobs. It might be interesting to test, especially the performance difference between them and Instanced structs. Thanks for the suggestion!

I second this ! Thanks brother

Thanks, I try to stay away from generic topics and do more interesting stuff. Not sure how good it is for the view count, but it's more fun for me :)

@polysiens u gotta think about it as there's always tons of beginners looking to get started buy the amount of people who stick with it to the point this kinda stuff is useful is probably around 20% percent or less of people who start. :)

Yea definitely not hitting majority of Unreal Devs. Unfortunately the saturation of generic, mediocre and sometimes bad information related to unreal is pretty high, probably in part because they are targeting beginners that don't know any better.

@@polysiens just joining the thread to say this meticulous content is appreciated. There is probably a niche for software engineers that are migrating to unreal and are shocked by the general bad practice and organisation in 99% of videos about it. You are a notable exception and a much welcome one :)

I'd be curious about this too. I have no formal education in coding, so all I've learned from youtubers is just that, avoid most casting, although there are some times when it is perfectly fine with things that are always loaded to begin with.

Hi, Yes there is a lot of dogmatic advice that is given and need to be unlearned when it comes to blueprints in Unreal. For me, interfaces are mainly for decoupling(when creating open world games you should always think about who is going to pull who into memory and why) and sometimes I use them to talk to different classes. For example, I might have an item, door and NPC that I can talk to. I can create Interactable interface there, so I can more easily interact with them. There are more unique uses that you can think of, but I would say these two cover 95% of them. For open world games, you want to live in the second category I spoke about in the video and be very conscious and critical about what exists in your game and what does what, because everything matters. If we ignore the obvious like, texture size, polygon count, lighting quality,..., you also need to worry about asset variation(more unique objects means less optimization that the engine can do or you can do), batching and instancing, performance of behavior trees(if you use them) and so on.

Essentially, the end goal of asserts is to catch errors, so they help a lot with debugging, especially for multiplayer. A difference between catching a bug can be a 5 min fix or hours of testing, trying to figure out what even the issue is. If you are not ensuring your code is doing what you want it to do or creating fail safes when stuff goes wrong, you are hurting yourself down the road, the project becomes a buggy mess after some time and usually people just give up on their project.

@@polysiens Got it. Thanks again.