Unreal Engine 4 Thread

[Joewintergreen]

I still have no idea how my ladder system sold/is selling, but yesterday I went ahead and did a very similar system for handrails. I even modelled these ones myself! I'm no modeler so I'm pretty happy with how they turned out.

 

 

This and a bunch of other stuff I've done in ue4 recently are on me blog

[abregado]

I have been considering UE4 for simple 3d gameplay and level design prototyping. Couple of questions:

 

Is anyone using the UE4 Editor under Linux (Ubuntu)?

Is C++ really as much of a pain in the ass as every says it is?

I'm designing for low-power PCs and tablets, does UE4 support this "out of the box"?

 

I create game prototypes under my studio, Binary Gambit

http://bit.ly/binarygambit

[SuperBiasedMan]

Particularly in light of the GDC news, people might be interested in this document to help transition from Unity to Unreal:

https://docs.unrealengine.com/latest/INT/GettingStarted/FromUnity/index.html

 

I was thinking about learning C++ to expand my repertoire anyway so I'm thinking I might get back into making my weekly games (which I've neglected far too long) with UE, but do experienced people here think it's overly ambitious to try that? Just because, my general impression of UE is that it's bigger and higher level than Unity so I'm unsure if I'd actually make worthwhile progress on a project in my free time over a week.

 

EDIT: Also is it tough to run? My laptop is a mess and my desktop is dieing. They run Unity fine because I make games with primitive geommetry (basically either simple 3D primitives or sprites) but if Unreal was a big difference it could be a problem. I would just download it and check straight away but I'm on hotel Wifi.

[zerofiftyone]

Particularly in light of the GDC news, people might be interested in this document to help transition from Unity to Unreal:

https://docs.unrealengine.com/latest/INT/GettingStarted/FromUnity/index.html

 

I was thinking about learning C++ to expand my repertoire anyway so I'm thinking I might get back into making my weekly games (which I've neglected far too long) with UE, but do experienced people here think it's overly ambitious to try that? Just because, my general impression of UE is that it's bigger and higher level than Unity so I'm unsure if I'd actually make worthwhile progress on a project in my free time over a week.

 

My friends and I like to get together and make 48 hour jam games using Unreal on a semi regular basis, so I'd say its totally possible to make smaller scoped stuff in a short amount of time without getting bogged down by the toolset.

[itsamoose]

How is the coding/scripting side of Unreal lately?  I haven't taken a look at Unreal since Unrealscript was around, which I understand has been removed in favor of C++ and a visual system.  I've never used a visual scripting program to write actual game logic like Input, and the programmer part of me is a little uneasy about doing that.  For those that have used it, how does it compare to the previous method/traditional scripting?

[bakelite]

C++ and blueprints are good at different things, you'll probably want to combine them. I have taken to using blueprints with a lot of custom nodes written in C++, it's kind of the best of both worlds.

[Joewintergreen]

Submitted my fire propagation system to the marketplace!

[root]

I am also noodling with ue4! I am 'dawnchapel' in their instant messenger thing and I added everyone else upthread who listed theirs.  I haven't done a whooole lot of work with previous iterations of Unreal editors but I used to teach UE3/UDK to teenagers at a summer camp and am looking forward to find out how much fun UE4 is in comparison.

[SuperBiasedMan]

I'm signed up for UE4 now, gonna see how retaining my username as SuperBiasedMan and I added usernames I found in this thread.
I'll see what mess I can throw together in this!

[unc0nnected]

We've been working with UE4 for 2 years or so and I have to say that in the past 4 months we've seen some incredible direction with how they are letting the community shape the engine through bug fix submissions and involvement.  One of the things I would caution anyone about though is buying too much into the idea that blueprints are the way forward to make your game.  Epic has this way of creating these amazing user friendly tools like kismet or blueprints and then overselling just how much they can/should be used for the development of a product.  They are fantastic prototyping tools, no doubt, but if any of you are getting into the engine watch out for games getting crushed under the weight of relying too heavily on blueprints.  C++ is still the best way to go for creating and implementing complex system so be wary of becoming too dependent on blueprints because they offer the path of least resistance in terms of an entry barrier.

 

Anyways, here's a little something we cooked up for the Identity project that just got kickstarted: 

 

[Joewintergreen]

I've been working with UE4 for about the same amount of time, and while I don't disagree? per se and I certainly agree that that was the case with Kismet (and they shouldn't have been pushing it as a "make games without writing code!" thing the way they did), Blueprint is much more powerful and for a lot of people, their work isn't going to benefit from using C++ over blueprint. For me (and I suspect most people), well-organised blueprints will probably always be more legible and explicable than code, and the benefits of accessibility, speed, and workflow that you get with Blueprint are going to outweigh the considerable performance benefits of C++ in most practical situations.

 

I guess I just would caution people similarly in the other direction, away from the "oh, you need C++ for anything serious after all" direction, and as usual the most useful strategy is probably somewhere in the middle.

[Joewintergreen]

Ya'll should totes mcgroates vote for my fire propagation system on the UE4 marketplace https://trello.com/c/SaMJxkjk/300-impromptu-fire-propagation

[unc0nnected]

I'll admit that blueprints are a leaner and better system for heavier usage for sure, but out on the projects we've worked on that were of any weight the developers all came to the same solution and that's that a large complex project composed entirely of blueprints will eventually get crushed under their cumulative weight.  Little systems here and there aren't going to make a difference on their own but when we get into 100's of blueprinted systems or 1000's they all add up to a noticeable hit on performance and stability.  Simple projects or proof of concept demos I would encourage people to use them whole heartedly but if you are making something complex like a full fledged FPS or particularly an RPG then I would steer people away from it.  

 

I imagine it's something that epic is working hard to improve and may have already gotten them up to the point but I remember one of our clients getting about 9 months in of heavy development just to realize they would have to go back and rip out all of the major blueprinted systems and redo them with C++ as the project reached a certain size, the only thing  would want to say here would be to mitigate that risk and be selective on what you blueprint and what you do in C++

 

I've been working with UE4 for about the same amount of time, and while I don't disagree? per se and I certainly agree that that was the case with Kismet (and they shouldn't have been pushing it as a "make games without writing code!" thing the way they did), Blueprint is much more powerful and for a lot of people, their work isn't going to benefit from using C++ over blueprint. For me (and I suspect most people), well-organised blueprints will probably always be more legible and explicable than code, and the benefits of accessibility, speed, and workflow that you get with Blueprint are going to outweigh the considerable performance benefits of C++ in most practical situations.

 

I guess I just would caution people similarly in the other direction, away from the "oh, you need C++ for anything serious after all" direction, and as usual the most useful strategy is probably somewhere in the middle.

[unc0nnected]

totes mcgroates voted, lucky #50!

 

Ya'll should totes mcgroates vote for my fire propagation system on the UE4 marketplace https://trello.com/c/SaMJxkjk/300-impromptu-fire-propagation

[Joewintergreen]

Thanks! Submitting the handrails today

[bakelite]

Fun with post processing. I really like this style, but the aggressive contrast makes it less than ideal for an actual game.

 

[Joewintergreen]

That does look really rad. I should look into PP more one o these days, all I've done is some distortion stuff 

 

Here is a cool thing I did the other day with AI for my stealth game. I made it so that if a guard hears a sound that he doesn't know what it was, and can't see the player, he'll check for torches nearby and if there is one, he'll go grab it, then investigate the noise (move to the location of the sound). They don't bother with a torch if the torch has been doused, or if they already have fire arrows, because they can see by that light. The awesome thing is, this accidentally made it so that if a guard doesn't have a torch, and hears a noise he wants to investigate, he can go steal a torch from a guard who has one. It is lol.

 

[zerofiftyone]

Fun with post processing. I really like this style, but the aggressive contrast makes it less than ideal for an actual game.

 

 

Nice! Was this done through colour grading? My immediate thought was that you could use something like that to recreate the pause screen effect from Arkham Asylum.

[zerofiftyone]

That does look really rad. I should look into PP more one o these days, all I've done is some distortion stuff 

 

Here is a cool thing I did the other day with AI for my stealth game. I made it so that if a guard hears a sound that he doesn't know what it was, and can't see the player, he'll check for torches nearby and if there is one, he'll go grab it, then investigate the noise (move to the location of the sound). They don't bother with a torch if the torch has been doused, or if they already have fire arrows, because they can see by that light. The awesome thing is, this accidentally made it so that if a guard doesn't have a torch, and hears a noise he wants to investigate, he can go steal a torch from a guard who has one. It is lol.

 

 

This is awesome. Good to see you're still working on that stealth project. Also, those bots are so much more menacing looking when the footage is sped up like that.

 

Edit:

Congrats on getting your fire propagation stuff on the Marketplace btw!

[bakelite]

Nice! Was this done through colour grading? My immediate thought was that you could use something like that to recreate the pause screen effect from Arkham Asylum.

 

I tried color grading at first, but it looked really strange. Some materials would look fine while others looked like a black and white negative, if that makes sense. My solution ended up being super simple. Separate the red channel, compare the resulting float to another float and set the color to black or white depending on the result. You could repeat the process with the green and blue channels and add them all together at the end, but it looks basically the same except slightly brighter.

 

 

 

A bunch of nested if statements with different cutoff values and colors, and some pixelization:

 

[bakelite]

That does look really rad. I should look into PP more one o these days, all I've done is some distortion stuff 

 

Here is a cool thing I did the other day with AI for my stealth game. I made it so that if a guard hears a sound that he doesn't know what it was, and can't see the player, he'll check for torches nearby and if there is one, he'll go grab it, then investigate the noise (move to the location of the sound). They don't bother with a torch if the torch has been doused, or if they already have fire arrows, because they can see by that light. The awesome thing is, this accidentally made it so that if a guard doesn't have a torch, and hears a noise he wants to investigate, he can go steal a torch from a guard who has one. It is lol.

 

 

What happens if you fire a water arrow at a guard holding a torch? Will he freak out and start flailing around in the dark, potentially resulting in unfortunate guard-on-guard violence?

[Joewintergreen]

That looks really really rad. Might steal it for something

 

In other news my fourth marketplace item went up! Procedural handrail system, twenty bucks.

 

[Joewintergreen]

What happens if you fire a water arrow at a guard holding a torch? Will he freak out and start flailing around in the dark, potentially resulting in unfortunate guard-on-guard violence?

Haha, good idea. Right now his torch just goes out and he acts normal

[zerofiftyone]

Hey, do any of you folks have experience with HTML5 builds? My project will build successfully, but I'm not able to run it in browser. The documentation is pretty sparse too  . Anyone here had any success with it?

[Joewintergreen]

For a long time it was only working with 64-bit Chromium or Firefox builds, might still be the case?