Видео

It was a long time ago, but I simply just use multiple line traces shooting in random patterns along a forward vector. The lights are seperate blueprints, which on overlap with the growth, will pop/disable. This was avery inefficient method that would not work well in a large map as hundreds, maybe thousands of these growth blobs would spawn, and I would do it very differently now with the knowledge I have. I should also mention that this was networked too, so I would imagine significant network performance issues unless using OnRepNotify's and network relevancy to lower network lag.

No, sorry. It was merely just me experimenting with render target painting, it had lots of problems. I believe there are a lot of tutorials out there now, explaining how to do this.

@@otreum25 what has alot of problems, your work or unreals render targets? Yeah i found some tutorials, they are mostly flat plane / board drawing I like how your drawing has depth into it , do you suggest a documentation or tutorial to get into this? Thanks

@@engimo5253 My work had problems. For example on the sphere, I couldn't paint any depth past half the sphere I think. I haven't worked on anything like that material in years so sadly I don't really have any recommendations. There's a link in the description of a much better implementation, and the video that inspired me to try this, I can't remember where I went from there to get information, it was so many years ago now.

@@otreum25 thanks for responds , I usually think it’s impossible for old videos channel to respond but surprisingly you responded quick , thanks alot , good luck with your unique creations

​@@engimo5253 That's ok, I tend to be very quick with my responses in general. I am currently working on a party game called Metal Heads, if you would like to check it out, there's a demo available on Steam with 12 mini-games (probably too many for a demo haha), and I'm aiming to launch into early access by end of March. store.steampowered.com/app/1132670/Metal_Heads/

It's been a while, but I believe I used Blueprint Interfaces. Either that, or on BeginPlay, I tell the Generator what the line is that it should be interacting with is, and then reference the generator in the generators UI. The UI would call functions within the generator object, and then the generator would call a simple On/Off function within the powerline. I'm not gonna go into detail, but it's really very simple logic really if you break down how things operate. The computer is connected to a generator, the generator is connected to a powerline, the powerlines are connected to other devices that have on/off functionality. Interfaces really help with this as they contain global functions that can be used across a multitude of devices. Now that I have far more experience in UE4, I would do my entire power system completely different in this game, but I would absolutely still use blueprint interfaces as they are very powerful.

Serj Serg Hi mate, its been a very long time and the project and both backups were lost due to a surge while backing up. And the other backup was on a dead hdd unfortunately. :( That was the main reason I stopped.

otreum25 it's very very sad :( I know what is it. I have the same bad story with losing sources of one old project. Be careful for the future :) P.S. Have you published some drafts, sources or parts of it on forums or website during development? Is it possible to find some parts of your project on the first stages of your job? It's better to find something than nothing :)

Haha yeah, I was very disheartened, as at that stage, I actually had it working cross platform, I was able to play multiplayer on a phone (Samsung Galaxy S2), tablet (Samsung Galaxy Tab 2) and Windows PC, which was pretty awesome even though it was really super early prototype networking. I never really posted anything anywhere or shared anything anywhere, only my progress on youtube. I since have worked on other much larger projects and am currently using Unreal Engine 4 for a party game, along side my main project which is a much larger multiplayer game, but too ambitious to get done in a short time frame while working a full time day job, having a girlfriend, looking after ferrets etc. If I were you, I would actually have a look into using Unreal Engine 4 for creating a similar game, I made a very quick prototype a while ago to try and remake what I had made here, and it worked really well within a VERY short time frame, so definitely worth looking into, and you don't have to be a programmer to do it either, as they have a visual scripting system called Blueprint which is a logic based Node/flowgraph system, it's super powerful and awesome :)

Hi Benny. Thanks for the encouragement and interest. At some point, I would love to do tutorials, because afterall, it is with the help of tutorials that has taught me so much about UE4. But between working a full time day job, having a relationship, working on my own game project, it doesn't leave any time for anything really :( I'm hoping that once I release a game, that it sells well, and then after that, I don't have to work my full time day job, which will give me more time to do the things I love doing :)

Any chance you could link some of the tutorials you might have used to teach you this particular event?

To be honest, it is simply logic based, and could really even be improved upon now that I know a hell of a lot more about UE4. I started out by just drawing on paper how I thought It would work, like a top down type thing, then I messed around in UE4 to see if I could make it happen, first I started by looking how I could move actors in a direction, then I looked at physical materials to help actors like the player slide along the floor (as the vacuum can suck the player out, I just disabled it for this video. And I did this as one of the first things I did for the project once I learned the basics of UE4 (by following all of the official tutorials at the time), so it's a bit janky and hacky, so could be more efficient and realistic. The zero gravity thing was purely just to show zero gravity, and is by no means realistic, but the gravity was controlled based on air pressure going lower than a threshold value, the lower the air pressure, the lower the gravity for individual objects. I haven't worked on this in a while, so I can't really remember specifics of how it was all done, but to start you off, it was done with a series of blueprint actors that are simply trigger boxes that store actor arrays. I believe that on begin play, the actors add connecting actors of the same time (think of a corridoor, that has all of these volumes in it, and they each have a slight overlap) to an array so that they can update one another. These same actors are used as child actor components for the doors, so that when a door is opened, the volume that is on the outside, updates the one on the inside, and thus a domino effect occurs throughout the level, with each one of these actors updating the next one. There is a fall off effect of the vacuum (as you can see by objects far away being pulled slowly at first), to avoid the entire level getting sucked out the vacuum, but as far as air pressure, oxygen value etc, that eventually gets sucked out of the entire level over time, which is handy for if there was a fire in a section of the level, and player opens an airlock, it sucks the oxygen out, thus the fire stops burning, depending on the type of fire. As far as working out direction of pull, that is simply done by having an A-B-C system, where A is the origin of the vacuum, it then tells B to pull objects towards A, and B then tells C to pull C's objects towards B. I'm probably not explaining it well, the there are problems with such a system, but it is functional. I will eventually expand it to be node based, so each section has several nodes, effectively creating a vector field so that if there were 2 airlocks opened near to one another, that objects won't only get pulled in one direction, but I imagine this solution being way more expensive if I were to do it. Sorry I can't help further, but hopefully that gives you something.

Haha, it really does, doesn't it, haha. One day when I get back into the project i'm using this with, hopefully the gaming world will see it's gruesome but potentially helpful in-game purpose. (in game medical/surgery/wounds)

I got it working a while ago, although, it felt real hacky. The computer object itself is a blueprint, inside that blueprint is a 3d Widget Component. The Widget component and computer blueprint communicate with one another directly, thus being able to do replicate certain things, such as button presses, screen changes etc. It's not perfect, but it works :) I haven't done a huge amount of work on the project in a while as it's scope is massive, and i'm running out of funds, so just working on a small party game for now to try and gain an income to power my main project, so as far as how to do the replicated 3d Widgets, I couldn't really go in detail.

otreum25 Ok thanks for reply. Its a strange thing with 3d widgets and replications. If i start with ded server one client can use it but the sec not. I have the 3d widget as component in a actor bp with other components. Ive made a simple replication function but it seems the fail is in the 3d W self cuz also hover dont working on client side its like the component isnt there for the client.

I think what I did was used the main computer BP collision to tell the server who should get control over the 3D widget, and it works fine. I can't remember if the hover functions show up for other players or not though. I imagine that I would need to create an event from the 3D widget to tell the main computer to replicate the specific button within the 3D widget being hovered. It's a silly way of going about it, but probably the only way without diving into C++ to create a replicated widget class (which is what I think I will eventually do as my main game project has computer interfaces all over the place which need to be replicated.

otreum25 Ok, so you gave the collision the set owner function and then cast to custom player bp? Or controller or you made it other way? I think that is what i would try but dont know if this works.

I honestly don't remember, but one way to find out, give it a try, that's what I do :)

Haha! Yeah, it always bugs me when in games, you can't destroy lights. There are quite a few games out there that let you destroy lights though, and those games are generally pretty awesome :) It's the little details that have the biggest impact I think. The only drawback to this, is that without proper global illumination and decent performance under dynamic lighting, it is difficult to make the lighting look good. But I think for the type of game I am making, people will be completely fine with that.

I can't remember specifics, but I looked at a render to target tutorial which showed how to paint a mask onto a blank render target. I think I used that blend target's value to drive the alpha of a lerp between the flesh and skin textures. The other way to do it I suppose is to have a flesh mesh underneath the skin mesh, and just modify the vertex normal so that the skin sinks beneath the flesh. I did this a few months ago and unfortunately due to a dead hard drive I lost the blueprint for it, so off top of my head, I can't really remember sorry :( This video is what I used to get me going though. ruclips.net/video/1Z-V1Buk6z8/видео.html

Also, a much cooler version of what I did, which inspired me to try and figure this out, is found here. It's way better than what I have done :) ruclips.net/video/8D_XIhGAfXs/видео.html

linear intrepolation node.

Haha, thank you.

I haven't worked on it in a while, and I had to create a new version of the power system to cater to my game's requirements, plus making it more optimized for network performance (it already was, but had way too many instructions than it needed). I just viewed the network like actual power. I'm not really an electrician but I just made use of inputs and outputs, logic gates etc. Basically the logic is this: TURNING POWER ON: 1. Send line trace from power source node to next node and check if it has power, or not, if it has power, stop looking for more nodes (this would not be the case anyway, but this could be used if two power sources worked on the same power grid for example, in my game, back up power is a thing, so I have this in place). 2. If no power is found in next node, check to see if it is broken, if it's not broken, give power to next node (this is represented by changing materials on node to be emissive for testing purposes, if it's broken, stop sending power through nodes. IF POWER IS CUT SOMEWHERE IN THE POWER GRID: 1. In the example of a power node that has 2 outputs (A straight section with ends A and B), we send a checking signal first, this doesn't do anything other than to look and see if we can find a power source in a particular direction, if we can find a signal, we don't do anything in that direction, then we change over to send a checking signal from the opposite point, and if we can't find a source, obviously we need to cut power in that direction, it's pretty simple. NOTES: There are other systems in place to avoid infinite checking loops, where if there were an actual loop in the power grid, there are fail safes that I have in place to avoid any sort of looping or incorrect behavior, but some of them get a bit complicated, perhaps unnecessarily so, but that's how I have it right now for prototyping purposes. I should also note that when doing a checking signal, the node that is broken sends a reference of itself through to the next node, which sends that same reference through to the next, and so on, so forth, each time, a counter is either added to, or subtracted from. In the case above, the count started with 1, subtracts when the signal leaves the node, it finds a valid node, adds 1, then does a check to see if the new value is 0, if the value is 0, we have pretty much exhausted the amount of nodes to check, thus we know it's time to do an action such as cutting power. If the checking signal were to come across a 3 way node however, it would add 2, as it needs to branch in 2 directions rather than 1. If it's 4 way, we add 3 etc.... None of that probably made any sense, because I was trying to summarize without going too in detail, but hopefully that gives an idea. No tutorials will be done, as it's something for my own game, and it's something I spent way too much time on, but hopefully the above gives you a rough idea.

That actually makes sense and it helps a lot! Thanks so much for the explanation

Thank you for the kind words. For my first Unreal Engine and Chivalry map, I was pretty happy with it :) But I did go a bit overboard trying to make features within a map that weren't seen in Chivalry, like the Captain feature, and as a result, it just confused players. I love the close quarters chaotic combat feel, I think chivalry is missing that, most chivalry maps look like they are designed for huge armies, not small warbands, and thus the maps feel big and vast, but empty. I think I was trying to fill that gap by making just full on, crazy gameplay, and it worked. Most people loved it, but there were a few who complained...probably because they couldn't do their spinny whirly backwards, upside-down attacks without dying, they had to kind of gather with the other team mates and work together to succeed. Thanks again for the kind words, I appreciate it :)

otreum25 hahahhahahaha Yes great observation on the gameplay I noticed the Same thing. most players want duels, but hey team objective is not like that. it's team tactics. I was wondering if you are making any other maps? I think it would be interesting to make a map or two that reflect battles from the era of the crusades. the map belmez is actually loosely based on a real historical event where the king of France Belmez ordered the extermination of the Templar knights. there are many to choose from. I totally agree with your take on the layout. I am sick of open spaces that create more running and Archer dodging than is fun.

I won't be making any more maps unfortunately. I made the Colosseum map, which was great, but then I sold the map to Chivalry developers and I dunno what happened, they changed things, broke the map, and I couldn't enjoy even my own map anymore, not only that, but I think it's one of the most hated "official" maps now, which is sad :( I am currently working on my own full fledged game, have been since I released the Colosseum map, but I'm keeping that under the radar/close to my chest until it's ready to be revealed ;)

otreum25 very interesting. the colosseum map has a lot of problems haha, ontheedge is an excellent map though, you clearly know what you are doing. well if you can't tell me what game you are developing can you at least tell me what type of game it is?

Hahah, at least I gave the illusion that I knew what I was doing XD The map did take over 1000 hour of work.... I'm still working on how to sum the game up, but it's a multiplayer/singleplayer space sci-fi, space station survival simulation game with a procedural/dynamic storyteller, and mixes a whole bunch of sub-genre's together.... I don't like to relate it to anything, because the project wasn't originally inspired by any games, just sci-fi horror movies, but Space Station 13 (SS13) is the only game that exists that is similar, but not the same if you want to take a look at that. Mind you, mine is first person, and done on unreal engine 4, so faaaaar better engine than SS13 :P

Hi Portl_ Not using physics volumes, just normal volumes, and adding physics velocity to objects within those volumes. Basically each volume interconnects with the next, and I can't remember, but either in the construction script or begin game, the volumes get their neighboring volumes and add them to an array. If one volume is updated, it sends the information to it's neighbors, including direction, strength of vacuum etc. And yeah, the hand crank for generator was inspired by Alien: Isolation haha, machinery in my game should be highly interactive, and I want players to have to memorize how the machinery works, rather than run up to something and hit an on/off button. There is a reason for the complexity, but I won't divulge just yet. I look back at this video and think "How far have I come!?" because since this, the game looks very different, feels very different, and has a plethora of features and new things. But once it's ready to show off, I'll make more videos ;)

In the latest version of UE4 (4.13) the screen interaction is done completely different, and is very frustrating that they changed it. Now 3d Widget interaction is designed for VR, but not for keyboard and mouse, so if you want to type in a textfield for example, you can't, you need to create an in game virtual keyboard...so in my game, the player will need to type in computer commands (ie "-rprog <x>") and the player would sometimes need to type this fast...but now, I need to fit a virtual keyboard on the screen, and if the player was typing, it would be like he was typing with one finger, which is stupid. I'll eventually try to re-implement the old system back into the newer engine versions to work around this, but doesn't seem like epic games are going to go back to the older (better) system, or at least give us the choice to go back to it :(

Yeah, I love Unknown Worlds and their work. They have done wonders with Natural Selection 2, and definitely serve as an inspiration for the alien growth here. However NS2 growth is technically a lot better than my current implementation, and I need to do a lot of work, on how this growth actually spreads. I have implemented a flame-thrower now which removes the growth which is pretty cool :) Also, that would be really interesting, seeing the vaccuum suck it all out....That is assuming that the airlock doors could open with the growth stuck to them ;)

Haha, thought so, it seams we play many of the same games, NS2, Overgrwoth and chivalry! NS2 is my fave of the bunch, although I think recent updates have not been all good, back in its prime I found this moded server with friendly fire and where gorges and commanders had much higher limits on structures they could build. As a gorge you could build 100 clogs, 30 hydras 30 babblers etc! It really felt like you were going into a proper alien base when you had ginormous clog structures bristling with hydras! Duion's Mod I think it was called, not seen it around in forever now though. Anyways sidetracked there! Hmm, yeah love the idea of doors stuck shut with goo, just thinking, if you got trapped in a section of corridor by the goo, would the doors work once cleaned up with a flamer? The option I prefer is that they would be permanently damaged and you would have to cut a hole through them once the goo reaches but maybe that's over complicating things, also might limit the things you could do with airlocks would be limited in that area later on so maybe not, haha.

Sure is. I recently had to re-implement this, as UE4 updated and unfortunately some of the blueprints were corrupted when transitioning from version to version. But thankfully only took a few hours to work out why it wasn't working. The current version looks at more variables and doesn't make objects float in zero gravity until they aren't in a gravity field (if the objects are in space or if grav generators are disabled/disfunctional/sabotaged).

Interesting, Would love it if you floated around when in zero gravity too! If there was air in the bit you were in you could just swim, and jump from the sides, in a vacuum this would be a bit trickier, but there would be no drag so you would never get stuck away from the walls, so could wall jump anywhere, could be very fun! Especially if the open door or wherever the air was going out of sucked you towards it too so you had to so some impressive parkour to stay inside, unless there is a handle to cling too or something!

+Alex Schmidt I didn't quite get right into the optimization side of things in the end, I mainly just relied on camera occlusion culling. And for the zooming side of things, well, I didn't have mip-maps for the tile textures, but that certainly would help in a big way when rendering so many tiles at once. If I got to entity and AI creation, I more than likely would have had basic AI for when the AI wasn't on screen, just to track movement and maybe if the AI was killed/attacked, but once in view, the AI would become more advanced to save on resources both on the logical side of the game, but also on the server, since it was going to be multiplayer. I was going to eventually do chunk loading (for the server), but I hadn't quite worked out what method I would go for as I hadn't got to that stage yet. I stopped working on the project after getting into the networking side of things, but I got it working cross platform on android, IOS and PC thanks to LibGDX. Unfortunately, to my dismay, the reason I stopped working on the project was my laptop that I had the latest version on got fried, alongside with my backup hard drive that contained the game. My main PC had a much much older version and to be honest, I became incredibly disheartened to work on it, so I saw a Chivalry: Medieval Warfare mapping contest using Unreal Engine 3 (which I had no idea how to use) and decided to dedicate myself 100% to that to try and win $10000. As a result of doing that contest, I worked on my Colosseum map, which is now part of the official game (albeit, it worked fine until the developers got their hands on it, and f'd it up, so now it has a bad name :( ). After that, I decided to work on one of the many project ideas I have, and have been working with Unreal Engine 4 since it was released to the public, and loving it's flexibility and ease of use. I was even considering creating this same very game using unreal engine 4.....personally, I think it is entirely feasible and probably very easy, but my problem is I always aim too big in my projects haha, and I am too stubborn to let go of the big projects and do something smaller. Unfortunately, I won't be releasing any source code (if I could even find it...I think it's actually completely gone now, as the hard drive that it was on, on my main PC pooped itself and I didn't recover it....unless I have the game on a usb somewhere, I don't think I have it anymore...all that work for nothing :( ...)

+ModBodDolphin Hi, thanks :) There won't be any tutorials. But I will say that it is a lot simpler than you might think ;)

+Александр Комаров It's not meant to make sense...just experimenting with dynamic elements of gameplay simulations.

RulesRedefined Hi RuledRedefined. Unfortunately I am not willing. But not due to selfishness, more so due to the fact that it's not completed, and I consider it a bit of a flop in it's design. (I approached the design incorrectly). I was working on a much better version, but have since stopped work on it. Also, the model is now part of the "Colosseum" map I made for Chivalry: Medieval warfare, which was purchased by the developers of Chivalry: Medieval warfare, along with the Colosseum asset itself, since it was part of the map.

otreum25 No worries. Thanks for your reply. You do great work :)

Brother Gaming This was from 2009 man, I couldn't remember if I tried :P

ah... :( okey

HolmsGaming Thank you for the kind words. :) Epic games live up to their name, they really are epic. Without them, things like this wouldn't happen...at least not from me.

Ohh I found the solution: There's a bug with pivot and the experimental WidgetBlueprint, so you have to set pivot to 0,0 and then it worked! I'm so happy

Glad you got it working! :) I had the same problem at first, but the mouse had an offset position to the screen and I figured the pivot had something to do with it, I changed them to 0,0 and it was fixed. Only to then find out heaps of people were having the problem, and the fix was already found haha. Sometimes it pays to experiment with settings, even if you think your idea is crazy.

I wasn't a big fan of the tmx map loader, so figured i'd learn how to do it myself without that. I feel I have more control this way as well. Unfortunately though. I haven't worked on this for a while now due to other work, life, etc. But i'm sure one of these days i'll continue on with development lol

Hey could you tell me what algorithm did you use for the islands generation. They llok very good!

Pedro Rodriguez It's a modified Diamond Square algorithm. Link in description. javagamexyz.blogspot.com.au/2013/03/terrain-generation.html

Also, this video is pretty early days on rendering. Here is where I got up to last. Auto-tile generation testing. ruclips.net/video/nYEB2TrtViE/видео.html

Thanks!!!

I won't be sharing any source code, the project is aimed at being a retail game on steam, not just a project. Otherwise I would :)

otreum25 Ok, thanks for the reply! I wish you good luck with your game =)

Thanks for understanding, best of wishes for YOUR game development. The project is on hold right now while I do some Chivalry: Medieval Warfare mapping contest stuff, but shouldn't be much longer before I get back to work on this game.

Hi Andre, I am not too sure. TBS said they will be testing over the next few weeks...that was a bit over a week ago. I don't think my map will win or become official however. While people want to see it as an official map, and love it, people also love maps such as Belmez, Cove and Castle Assault. I even had a dream last night that those 3 maps would win haha.