Support NeoGAF

[missile]

... It'll never end. But that's cool, because I like making games! And if I end up old and poor, well, I'll have spent my time making rad games ...

That's an attitude I wanna see from people making games. Thumbs up, razu!


@Roquentin: The comment about the millionaire shouldn't be taken literally. I
just wanted to contrast the status of being unemployed. But alas, some people
can't deal with having lots of time at their disposal, likewise with lots of
money. Speaking about habits. Habits are quite a good excuse for not stopping
something, eh? Well mate, how many iterations you need is up to you. But if
you really start thinking about it, it can be done in just one. Anyhow, you
are the judge. Always.

 

[Ranger X]

As mentioned, writing BMPs to disk or something may actually be one of the better options if nothing works for you, especially if your game has a small screen resolution.

If you haven't tried it yet, you might also try CamStudio (http://sourceforge.net/projects/camstudio/). It's free, open-source, and you can record screen regions for making tutorial videos...but I don't know how well it works for a 60 fps game.

*edit* Bandicam (http://www.bandicam.com/downloads/) seems like another option. It's shareware, but I don't know anything about how reputable the author(s) are.

I am curious how to record image by image. I don't know how to do that. Is it like a window plugin that basically is doing multiple "printscreen" in succession?

The Elgato Game Capture HD might work for you. This is a box where you plug in HDMI from a source (your video card), then you plug another HDMI cable going from the box to your monitor, and USB to your PC. Using its software you can record whatever comes in from HDMI, so in theory you could set it to start recording everything your PC does, then minimize it and run your game, and when you're done running your game maximize it and hit stop. Its software then lets you edit the movie you made (no replacing audio or anything cool like that though), and you can even upload it directly to YouTube if you want. You just have to make sure your game is running in one of the standard TV resolutions.

I've never tried recording PC from it, but it works great for tablets and game consoles, and there's no reason I can think of that it wouldn't work for PC, as long as your video card has HDMI. The box costs like $160.

That's interesting. I will look on that.

 

[BomberMouse]

I am curious how to record image by image. I don't know how to do that. Is it like a window plugin that basically is doing multiple "printscreen" in succession?

You are using GM? there's a function that copies the entire screen, then just make an object that is rendered last and in every step takes a screenshot and writes it to disk with incremental names. ie. ss001.png, ss002.png, etc.

The downside is that you won't get any sound, that's how I made my trailer a long time ago.

 

[Blizzard]

I am curious how to record image by image. I don't know how to do that. Is it like a window plugin that basically is doing multiple "printscreen" in succession?

I don't know if GameMaker supports it, and/or if GameMaker would be able to do it quickly enough, so maybe it was a bad suggestion. What people are saying is that if you have your own engine (or the ability to write a screenshot file to disk), and you have a fast enough disk that you can create a whole bunch of files quickly, you could actually flip a switch that dumps the game buffer to a BMP/PNG file, so you end up with frame 1, frame 2, frame 3, frame 4...frame 1000. *edit* BomberMouse already beat me to it and it sounds like it's possible. And in theory with the right editing tools you could add sound after the fact, or possibly even record the sound at the same time with a different tool.

Naturally that uses a huge amount of disk space, but then there should be tools (not sure if there are free ones) that can take all the BMP/PNG files and encode them into a video.

I'd suggest trying CamStudio (make sure you play with the quality settings) and/or the shareware version of Bandicam before you try to go that route, however.

 

[JulianImp]

You are using GM? there's a function that copies the entire screen, then just make an object that is rendered last and in every stop takes a screenshot and writes it to disk with incremental names. ie. ss001.png, ss002.png, etc.

The downside is that you won't get any sound, that's how I made my trailer a long time ago.

However, the audio issue can be fixed easily. All you have to do is use Audacity to record your computer's audio output (set the input channel to "Line Out" or something along those lines).

You should probably add something to your game as it boots up to sync audio and video (just like in the movies), an event that is very short (at least from the video side) and easy to recognize, like how the classic Game Boy's jingle played just as the logo stopped moving downwards, or how the Nintendo logo instantly appeared onscreen as the coin pick-up sound played on Super Mario World.

With that done, you should be able to easily sync both channels by making sure the sound starts in the same frame the visual cue happens. You might even do that before the game boots up and remove it afterwards while you're editing the video.

However, I guess capturing the screen so often while recording sound could be a bit rough on an older computer, so you should do a test recording before doing it for real.

On completely unrelated news, I've coded a new shader for the player's cell. Before, it was over one hundred lines of CGProgram code that required Shader level 3 (which means it wouldn't work on OpenGL 1.0 devices); now, it's this:

Shader "Custom/ShaderEsfera" {
	Properties {
		_MainTex ("Base (RGBA)", 2D) = "white" {}
		_MainTex2 ("Borde 2 (RGBA)", 2d) = "white" {}
		_Color ("Color (RGBA)", Color) = (1, 1, 1, 1)
	}
	SubShader {
		Tags { "Queue" = "Transparent"  }
		Blend SrcAlpha OneMinusSrcAlpha
		Pass {
			Color [_Color]
			SetTexture[_MainTex]
			{
				Combine Previous * Texture QUAD
			}
			SetTexture[_MainTex2]
			{
				Combine Previous * Texture
			}
		}
	} 
	FallBack "Custom/ShaderEsferaSimple"
}

Here's a comparison of both shaders taken from inside the Unity editor (oh, and please don't mind the slightly smoother [half] sphere on the right):



I think I shouldn't be doing three operations in a single pass in order to maximize compatibility on mobile devices, but I'll talk about that with my shaders teacher next Monday. It still needs some more work before I can get it to look more like the older shader, but lowering the graphic card reqirements is a good start.

The game runs fine on a PC with an on-board video card which was so old it even refused to load the Unity editor, so I think it should work even on old hardware, which could give it wider compatibility on smartphones (...probably).

I'm also creating a simple script to let the game recognize touch events as if they were mouse clicks, since it seems touching a collider doesn't trigger its OnMouse<X> events, but so far I've only been able to make the OnMouseDown work.

 

[Tashi]

So I have a question about collision detection. What would be the best way to differentiate collisions on the side of an enemy vs the top of an enemy (think walking into one vs jumping on top of it) I was thinking having 2 collision rectangles for the enemy. One that would cover the sides and then a very thing rectangle that would cover just the very top layer of the enemy. I'm not exactly sure how to go about doing this to make sure it makes sense in the end to the user and is best for gameplay.

 

[BomberMouse]

So I have a question about collision detection. What would be the best way to differentiate collisions on the side of an enemy vs the top of an enemy (think walking into one vs jumping on top of it) I was thinking having 2 collision rectangles for the enemy. One that would cover the sides and then a very thing rectangle that would cover just the very top layer of the enemy. I'm not exactly sure how to go about doing this to make sure it makes sense in the end to the user and is best for gameplay.

It's difficult but probably you might want to change your logic based on intentions instead of just collisions:
- If I'm in an attack state (ie, jumping), any collision on the head counts a a hit, disregard any collisions with the body.
- If I'm in a non attacking state, any collisions with the body counts as a hit towards the player, disregard any collisions with the head.

 

[Tashi]

It's difficult but probably you might want to change your logic based on intentions instead of just collisions:
- If I'm in an attack state (ie, jumping), any collision on the head counts a a hit, disregard any collisions with the body.
- If I'm in a non attacking state, any collisions with the body counts as a hit towards the player, disregard any collisions with the head.

That actually doesn't sound very difficult at all because I already have states for when the player is walking, jumping and attacking.

I'm just worried about those rare instances where may the player jumps, attacks but lands like 1 pixel off the ground and to the side of the enemy and the behavior doesn't make sense to the end user. And right now my enemy is just a rectangle (250, 60). I haven't actually designed any real characters or enemies. I'm just trying to see if the base gameplay is fun. I guess I could do it your way for my purposes right now since I'll be the one playing and I know exactly how to play.

Thanks!

 

[BlueMagic]

A simple way that probably works well is checking to see if vertical speed is such that the character is going down (I would say vertical speed < 0 but not every engine/framework/library uses the same orientation). That way you don't need extra collision rectangles. But yeah as you said it is not the best for tall enemies. Additionally you could check to see if it's in the upper part of the collision rectangle, but it's almost the same as having two rectangles.

 

[Tashi]

A simple way that probably works well is checking to see if vertical speed is such that the character is going down (I would say vertical speed < 0 but not every engine/framework/library uses the same orientation). That way you don't need extra collision rectangles. But yeah as you said it is not the best for tall enemies. Additionally you could check to see if it's in the upper part of the collision rectangle, but it's almost the same as having two rectangles.

That could work as well. Right now I'm going to just cheese it for prototyping purposes.

Got another question/issue. So what I want to do is if the character jumps on the enemy without attacking, I want the player to be pushed back. So I figured the players motion would be parabolic. Just a little rainbow from the point of impact and to the ground. How I determine which side it lands on, I haven't figured out yet in the design. As far as the math is concerned I need a quadratic equation right?

So the Y position is equal to (x position ^2) + (x position) + Y axis intercept

I don't know if I want the Y axis intercept to be the ground or the top of the enemy(point of collision) Not sure if that will have an impact on anything later on down the road.

Also the way I have it set up is that depending the type of collision (walk into, jump on, or attack) it calls a method in a different class. I'm going to have to increment X each time but I'm just discovering that as I write this post it's not going to work. The different collision methods are only called when there's a collision which is only once right? I need the collision method to be called until the player's motion is complete. Should I setup a boolean? Maybe a thread?

Damn. Sorry if this post is just a ramble and me thinking out loud.

 

[Ashodin]

Some updates; my RPG is on the backburner as I want to work out of smaller projects first. Start small as you can, I figure.

To this end I bought a Construct2 License and am now crafting a kirby-clone-esque platformer for PC/Android/iOS.

It's called Mr. Wave, and I've been working on the basics so far:

It's totally going to be retro. As I have shitty graphic skills. But I think looking like an NES game will do the game wonders (as it's what I was going for anyway, yay!)

 

[backdrifter]

Something I'm often intrigued by is what all you talented chaps do for QA. Do you rely purely on family/friends to bash the heck out of your game and hope to clear everything up that way? Put trust in your own testing skills?

I do QA day in day out for work, so it's something I'm obviously well and truly interested in. And if anyone fancies a tester to help them out, feel free to send a PM!

 

[Noogy]

Some updates; my RPG is on the backburner as I want to work out of smaller projects first. Start small as you can, I figure.

To this end I bought a Construct2 License and am now crafting a kirby-clone-esque platformer for PC/Android/iOS.

It's called Mr. Wave, and I've been working on the basics so far:

It's totally going to be retro. As I have shitty graphic skills. But I think looking like an NES game will do the game wonders (as it's what I was going for anyway, yay!)

A Kirby inspired game where you play as a microwave oven. There's genius in that.

 

[JulianImp]

Something I'm often intrigued by is what all you talented chaps do for QA. Do you rely purely on family/friends to bash the heck out of your game and hope to clear everything up that way? Put trust in your own testing skills?

I do QA day in day out for work, so it's something I'm obviously well and truly interested in. And if anyone fancies a tester to help them out, feel free to send a PM!

I've learned a few things from playtesting my game (both with friends/family and strangers at an argentine games convention). Hopefully, you'll find some of these tips useful:

• Take verbal or written advice with a grain of salt unless the palytester is quite professional or has an extermely honest personality. Strangers are likely to sugar-coat any criticisms they have about your game to avoid being unpolite, and friends/family even more so
• Most importantly, watch people as they play. Look at what they do both in and out of the game (vocalizations, gestures, etcetera)
• I think taking mental notes is far better than writing stuff down. That way, playtesters are less likely to think of the situation as some kind of scientific test
• Try no to intervene during a playtest. You won't be around to help people when you release your game, so the best you can do if a player gets stuck or frustrated is take notes about stuff you should tweak, explain, or train the player in
• You should always prioritize playtesting with people who haven't played the game before rather than ones who have experience with it. While you can certainly find the latter's feedback to be useful, most of the work you'll have to do relates to making sure newer players can get into the game
• Try to get into the player's head as you watch her play to understand what choices she makes and why, and then comparing them with what you had originally envisioned

 

[Ashodin]

A Kirby inspired game where you play as a microwave oven. There's genius in that.

Wait until you see the enemies, gameplay, and powers!

 

[Platy]

That "create a game in a genre that you hate" jam starts today !

Sign Up if you haven't

 

[embalm]

So I've been arguing with my artist half about the terrain textures of our game looking too busy. I was using Heroes of Newerth as an example of how too much detail can look bad, especially when we have special attacks going off and units all over. So I thought I would share a screen shot and see if you guys think it's too busy.

He took the units and GUI from HoN as a joke, but the screen works for showing off our terrain. The big bushes and grass pieces are still very early concepts, so try to ignore those in your judgment.

So do you think the terrain tiles are too busy?

 

[PhiLonius]

I think they're fine. As long as there is some differentiation between the backgrounds and characters/attacks/etc. and one thing isn't blending into something else, it shouldn't be a problem.

 

[razu]

Something I'm often intrigued by is what all you talented chaps do for QA. Do you rely purely on family/friends to bash the heck out of your game and hope to clear everything up that way? Put trust in your own testing skills?

I do QA day in day out for work, so it's something I'm obviously well and truly interested in. And if anyone fancies a tester to help them out, feel free to send a PM!

My game is so simple, there haven't been all that many bugs. If I was making a bigger game then I think I'd have to do some kind of beta and get data that way.

 

[Noogy]

One of my wonderful voice talent, Kimlinh Tran, posted this behind-the-scenes video of a recording session. Even as the developer I find this stuff absolutely fascinating, and it gives me an incredible appreciation for good direction and talent. I can't give Deven Mack, my casting director, enough praise. And the highest recommendation, if you are looking for a casting director for your game.

Featuring Kimlinh as Fidget, and Lucien Dodge as Dust. Please note that there are some spoilery bits from the game.

 

[Mr_Appleby]

Some updates; my RPG is on the backburner as I want to work out of smaller projects first. Start small as you can, I figure.

..../snip/...

It's totally going to be retro. As I have shitty graphic skills. But I think looking like an NES game will do the game wonders (as it's what I was going for anyway, yay!)

Had a similar realisation earlier this year, stop trying to start with sprawling, generative RPG's and go for something simple. (even if one or two of those RPG's were actually playable!)

Hence, the 2D platformer:

I'm doing the whole thing in Java, inefficient I know but I'm having a great time solving a variety of problems on my own. The collision system is the best one I've personally created so far and allows for walking up slopes, riding moving platforms and other neat (basic heh) stuff. The goal eventually is an adventure/puzzle platformer, something like the Dizzy games.

 

[Blizzard]

I'm trying to at least do a tiny bit of work on my project each day or so, though I'm making practically no progress and have a ton of other unrelated-to-this-project things that need caught up on.

I did create a GUI editor project. Tonight I got it compiling, and next I need to actually create a script and generate the resource library file that will contain GUI editor resources, allowing it to run. The first task is implementing a simple main menu -> editor transition, and then I will probably test GUI editor stuff (which may require creating more widgets), and after that I will want to create some sort of new/load/copy/save logic for GUI designs.

 

[Nezabyte]

Gotta say that Unity rocks. We've been fidding around with this as we learn how to use Unity and various tools from the asset store --

 

[Alts]

It's about halfway finished, but Github is hosting a game competition that some of you might be interested in. https://github.com/blog/1303-github-game-off

Basically, create a web based game based (potentially loosely) on some aspect(s) or features of git. It's a good opportunity for me to play around with canvas. I just started, but i think my idea might work out pretty well. If some of you, like me, work better with a deadline: here you go.

 

[OccamsBlender]

Crazy good info depot here:
http://www.mangatutorials.com/forum...-Ultimate-Indie-Game-Developer-Resource-List&

 

[missile]

Not a game quite yet, but still!

I have now finished my rasterizer. And it works so well. Fuck you! xD
Using an edge-tracing algorithm, everything depends on how good you can step
along the edges of a polygon. As one can see, most homebrew polygonal
rasterizer suffer from bad edges and wrong endpoints. It's a bit tricky to get
it right and becomes even more difficult if you allow for the rasterization of
an n-gon, since the slope of a line (being part of the boundary of the
polygon) can change upon arriving at a new vertex.

I develop this rasterizer from scratch based on my line rasterizer using a
very solid and sound mathematical foundation, and have rederived the edge-
tracing procedure for rastering a polygon but with an ansatz I have never seen
in any book or paper. Can't understand why the literature makes it so
difficult. Anyways, the result can be seen above. The edges are perfect (best
approximation on the given raster), and the endpoints are matched exactly.
This situation can only be improved with sub-pixel rendering. Oh, yeah!

The animation above shows a triangle rotating in three-dimensional space
around the z-axis while marching along. The triangle gets clipped at the
screen boundary producing an 5-gon shape at worst in this case. But this poses
no problem, no triangulation needed. The given rasterizer can render n-gons
as well.

 

[JulianImp]

Crazy good info depot here:
http://www.mangatutorials.com/forum...-Ultimate-Indie-Game-Developer-Resource-List&

Wow, that's an awful lot of stuff. Thanks a lot for the link! I just I shared it on Twitter as well (with all of my fourteen followers, heh).

Oh, that reminds me... I had to set up a Twitter account when I submitted the game to the IGF, so I guess I'll share it over here as well, for whoever wants to follow the game's development. It's @Julian_Imp

 

[Nemo]

So you are a time millionaire. People say money = time, but on my cards it
reads time > money. One can always earn money, but time is running out for
everyone ... no matter what.

Oh man, you just gave me some very much needed motivation as well. Haven't been doing anything since yesterday, motivation can really be a bitch if you don't have it. Probably by far the most important thing for making games

 

[missile]

@SanaeAkatsuki: Thx for the link(s), mate!

Something I'm often intrigued by is what all you ... chaps do for QA. ...

That's easy. I call an ape and let him slam the inputs long enough. If the
game still runs, it's going on sale. Btw; Someone once told me that you can
have Shakespeare that way! xD

... So do you think the terrain tiles are too busy?

The tiles are too busy in my book. Can't help, but they even look out of place
with respect to the characters. But that's perhaps just me.

Had a similar realisation earlier this year, stop trying to start with sprawling, generative RPG's and go for something simple. (even if one or two of those RPG's were actually playable!)

Hence, the 2D platformer: ...

This looks really crappy.

Spoiler

But in a good way! That dude! xD

Oh man, you just gave me some very much needed motivation as well. Haven't been doing anything since yesterday, motivation can really be a bitch if you don't have it. Probably by far the most important thing for making games

Indeed. Keep going! That's the only thing that really helps.

 

[OccamsBlender]

Something I'm often intrigued by is what all you talented chaps do for QA. Do you rely purely on family/friends to bash the heck out of your game and hope to clear everything up that way? Put trust in your own testing skills?

I do QA day in day out for work, so it's something I'm obviously well and truly interested in. And if anyone fancies a tester to help them out, feel free to send a PM!

I personally plan to go with Tofu Testing(http://tofu-testing.com/), but if you're willing to do iOS testing, I'll definitely send a PM your way.

You're awesome for offering QA!

 

[Blizzard]

It was late but I made myself work a little more today, to get the editor skeleton up and running. Ultra impressive screenshot:

Next up, transitioning to an actual editor user interface, and providing the ability to create/move/resize/delete/copy etc. widgets. Of course I'm laying out this stuff by hand since I don't have a GUI editor yet...and I just realized the label text isn't centered properly in that screenshot, lol. I fixed it but didn't upload another picture.

 

[missile]

@Blizz: Seems you invest a lot of time into. What's the GUI for, for an
immediate game of yours?

 

[Blizzard]

@Blizz: Seems you invest a lot of time into. What's the GUI for, for an
immediate game of yours?

Yep. I'm working on a turn-based strategy game that should heavily require a robust, responsive, GUI/menu/etc. system. It's probably not a matter of me investing so much time into the GUI as much as I'm not investing much time at ALL, comparatively, into the whole project.

I'm also trying, when feasible, to make the work I'm doing now useful for future projects. Once I get a way to save GUI designs, for instance, I should be able to quickly use the GUI editor to lay out stuff on future projects, and reuse all of the GUI widget/event stuff, without having to redo any of the work I've been building. I don't think I've had to work on the content-loading pipeline in probably 6-8 months, after I took the time to make that fairly robust and scripted, and that will continue to be useful for any future games I make with this engine. I may have to revisit loading performance, but for now I'm not focusing on that as much.

In general I'm working towards a game design I have laid out, but also trying to spend a bit of time laying robust future groundwork as I go along. Maybe I will regret it in the future, but maybe some of it will pay off in future game projects and experiments. I suppose it's a gamble like most things.

 

[survivor]

Anyone have any experience with some of these HTML5/JS engines: Crafty, EntityJS, or Akihabara. I'm trying to look for a JS engine that is mostly programming focused and has good documentation/support and these were the 3 that I found while searching. If you guys know any other, please recommend them.

 

[embalm]

I think they're fine. As long as there is some differentiation between the backgrounds and characters/attacks/etc. and one thing isn't blending into something else, it shouldn't be a problem.

Thanks for the feedback. I think we'll be ok with the characters against that background. We are going to move forward with tiles very similar to the ones I showed above.

@Nezabyte
Unity does rock. Keep up the work on the awesome looking game. Can't wait to hear or see more of it.

 

[missile]

... I'm also trying, when feasible, to make the work I'm doing now useful for future projects. Once I get a way to save GUI designs, for instance, I should be able to quickly use the GUI editor to lay out stuff on future projects, and reuse all of the GUI widget/event stuff, without having to redo any of the work I've been building. ...

What are the dependencies of your GUI, i.e. rendering backend or additional
libraries you may use, or is it built entirely from scratch?

In general I'm working towards a game design I have laid out, but also trying to spend a bit of time laying robust future groundwork as I go along. Maybe I will regret it in the future, but maybe some of it will pay off in future game projects and experiments. I suppose it's a gamble like most things.

Even if it doesn't pay off you will make an experience nevertheless.


GUI programming can be really cool, indeed. And I think we'ven't reached any
limits here. I for one have a very awesome GUI in my mind which is based on
pure physics (partial differential equations). This may change how we see/feel
a graphical user interface. Perhaps used in a game of mine for the first time.
I already work on parts of the theory behind to make this and many other
things a reality some day. But until that day I better grow some wings to
survive my indie freefall.
My little 3d engine needs some more breath. Next up; backface culling....

 

[FormallyMegadrive]

So I installed Unity3D 4 just before and now whenever I boot up I get this error:

Error Loading Page

SSL peer certificate or SSH remote key was not OK

Googling gets a bunch of Ubuntu Unity forum threads. Anyone got any idea as to why?

Edit: NEVERMIND, figured it out. It wasn't reaching the licence server, a router reboot fixed it.

 

[Teknopathetic]

For you Unity upstarts, a free version of Shade3D (A Japanese 3D package) was announced for Unity and the Mac version was released a few days ago with a Windows version coming shortly.


http://mirye.net/shade-12-overview

http://www.shadenews.com/homepage/n...ree-game-content-development-system-announced

https://itunes.apple.com/us/app/id569369787?mt=12


A quick glance at the official website and its features would imply that this is a fairly full featured program (it is on its 13th version, of course, and has been developed for over 20 years?), so this should be a alternative for people looking to make 3D art and animation for their games but fear/loathe Blender. Most importantly, IMO, is that it appears to have pretty decent character animation features. Blender is the only package under 800-1000~$ that has a full set of character animation features. There's Messiah Studio (which is only for Animation/Rendering), I suppose, but that interface is somehow even worse than Blender's.

Something to keep an eye on.


Edit: After, a deeper glance at some of the tutorials for this and it would appear that this will export FBXs without having to go through Unity, so theoretically, this should be usable for people who use *any* engine, not just Unity.

 

[bomer]

Does anyone have some good Scripting tutorials / references for Unity? I had a little play with it a while ago and found available resources pretty dismal. I have a C++/C#/Python/Java/PHP background.

 

[Teknopathetic]

"Does anyone have some good Scripting tutorials / references for Unity? I had a little play with it a while ago and found available resources pretty dismal. I have a C++/C#/Python/Java/PHP background."


I assume you're aware of the scripting reference?

Since you already have extensive programming knowledge, I can't imagine you'd really need more than that if you're one of those types that like to learn via scouring API references.

 

[Kritz]

Does anyone have some good Scripting tutorials / references for Unity? I had a little play with it a while ago and found available resources pretty dismal. I have a C++/C#/Python/Java/PHP background.

I don't have an answer to your current problem, but I agree that the unity homepage doesn't really have any nice tutorials to go through to learn the software / language intricacies. It does have one or two things for it, they're just... not very good.

But I will say that pretty much every unity issue I've ever had has been resolved by typing "Unity3D [problem]" into google. To be clear, I am not saying you should google the issue. I'm saying answers.unity3d.com, docs.unity3d.com and stackoverflow resources are fucking amazing for Unity.

So you should probably keep looking for help, and I'm sure people in this thread will supply some, but keep that in mind. Unity support is perhaps the best support of any engine / framework / language I've seen for making games. The docs stuff is also incredibly useful and maaaaaay actually have some answers you're looking for. I like documentation that actually describes usecases for things, and not just a super low description of what a certain function or class does.

 

[bomer]

@Teknopathetic, the barebones language reference is not overly useful and a lot of the tutorials I've found are not overly either.

@Kritz, I'll dive into a StackOverflow whole over the weekend and see what I can come up with. At this point I'm interested in doing some pure scripting creation, trying to not use much of the in actual editor to do much and just get a feel for how some of the systems work. I'll look into answers and docs too. Cheers!

 

[BlueMagic]

That could work as well. Right now I'm going to just cheese it for prototyping purposes.

Got another question/issue. So what I want to do is if the character jumps on the enemy without attacking, I want the player to be pushed back. So I figured the players motion would be parabolic. Just a little rainbow from the point of impact and to the ground. How I determine which side it lands on, I haven't figured out yet in the design. As far as the math is concerned I need a quadratic equation right?

So the Y position is equal to (x position ^2) + (x position) + Y axis intercept

I don't know if I want the Y axis intercept to be the ground or the top of the enemy(point of collision) Not sure if that will have an impact on anything later on down the road.

Also the way I have it set up is that depending the type of collision (walk into, jump on, or attack) it calls a method in a different class. I'm going to have to increment X each time but I'm just discovering that as I write this post it's not going to work. The different collision methods are only called when there's a collision which is only once right? I need the collision method to be called until the player's motion is complete. Should I setup a boolean? Maybe a thread?

Damn. Sorry if this post is just a ramble and me thinking out loud.

Well, theoretically, yes, but if you are already simulating gravity by giving the character a vertical acceleration you would just need to give it an upwards initial velocity (say, they collide, and the character gets vspeed = 10), and some horizontal velocity (which depends on the side of the enemy he's hitting, I suppose). The gravity is going to take care of giving the movement that parabollic feel since that is what causes it anyway.
If you want to know what direction the character should be pushed back to, one way would be to take the difference of the x position of the center of the enemy and the x position of the center of the character. If it's positive, he should go right. If it's negative, left. Additionally, you can also use the difference of the y position to change the feel of the pushback, but that depends on how you want it to feel (more 'organic' I guess vs more hardcoded).

 

[Uthred]

I know the question is a little nebulous, but is it worth learning AS just for use with Scaleform in UDK (assuming that your game is going to be very menu heavy)?

 

[RelentlessRolento]

Just wondering if there is an easy export option for making an animated texture in After Effects and export for use in Unity? or something similar?

 

[Dascu]

Just wondering if there is an easy export option for making an animated texture in After Effects and export for use in Unity? or something similar?

Somewhat related: In Unity, how would I best approach an animated texture on an object? More specifically, I want to show an animation on a TV screen object. Make an animated image in an image editing program and then use that .gif? Or do I have to turn it into some kind of raster/sequence in one big single image?

 

[Noogy]

Just wondering if there is an easy export option for making an animated texture in After Effects and export for use in Unity? or something similar?

I use After Effects to create my animated textures, but have never found a one-click solution to do it. I create the animation, create a new sheet for the texture to reside on, and then manually cascade the animated texture over time across the sheet. I then render out the frames of this sequence, and then reimport the images into AE and then export it as a single sheet. Then I merge all visible in Photoshop.

It's inelegant, but it works.

 

[Blizzard]

I know the question is a little nebulous, but is it worth learning AS just for use with Scaleform in UDK (assuming that your game is going to be very menu heavy)?

I'm not sure you have any other options, really, unless there's some third-party menu system (there may be fan-made ones for all I know). I suppose if you knew Flash you could do it, but I think you would still need bits of ActionScript to actually link things together. If you go the full Flash route there are tutorials around...doing it in pure free scripting and still making it look nice may be tough but I think it could be possible. :/

What are the dependencies of your GUI, i.e. rendering backend or additional
libraries you may use, or is it built entirely from scratch?

I started the overall engine with SDL, then switched to SFML recently. I started using SFGUI for the GUI after that point, then realized it looked like it might be a huge pain and require my own tailoring and implementation to do actual bitmapped/graphical GUI widgets...and since then I have removed all SFGUI code and implemented my own from scratch, on top of my mostly generic 2D engine (which still uses SFML/OpenGL for rendering).

Even if it doesn't pay off you will make an experience nevertheless.


GUI programming can be really cool, indeed. And I think we'ven't reached any
limits here. I for one have a very awesome GUI in my mind which is based on
pure physics (partial differential equations). This may change how we see/feel
a graphical user interface. Perhaps used in a game of mine for the first time.

If you like unusual menus, I seem to recall hearing that Endless Space has an innovative and very effective GUI. You might check it out on Steam or maybe a quick look on YouTube, if you like that sort of thing.

 

[JulianImp]

Just wondering if there is an easy export option for making an animated texture in After Effects and export for use in Unity? or something similar?

Animation textures is something I'm still inexperienced in but, based on what I know, I think you could use the first approach from my response to Dascu (below).

Somewhat related: In Unity, how would I best approach an animated texture on an object? More specifically, I want to show an animation on a TV screen object. Make an animated image in an image editing program and then use that .gif? Or do I have to turn it into some kind of raster/sequence in one big single image?

Here're two methods I thought up with my limited knowledge in texture animation (so they are probably not the optimal solutions):

Option 1: Texture altas

Start with a large texture (however, I believe Unity won't load anything larger than 4096x4096 pixels), map the object's UV coordinates to a small fraction of that texture (say, a 64x64 "tile"), and then move the texture's offset within Unity so that it jumps from one tile to the next. For example, if the texture consists of 2x2 tiles, the UV offset for each of them would be:
Tile #1: 0,0
#2: 0, 0.5
#3:0.5, 0
#4: 0.5, 0.5

That means you'd render four frames of animation and place them in a 2x2 pattern inside the texture atlas. It's probably the easiest thing to do, but you'll have to limit frame sizes to make sure they all fit inside a texture, and you should try to keep texture sizes low if you're aiming for compatibility. Individual frames don't have to be square (you could have a 7x3 grid if you wanted), but you should keep the whole texture's size as powers of two to prevent Unity from wrecking the image by automatically stretching or reducing it to a Po2 size.

Option 2: Render to texture

I think this is a Unity Pro only feature, but if you have it or intend to purchase it for your game, you should look into it. I don't know much about it since I've never used Unity Pro myself, but I think it uses a on-scene camera to capture an image and save it to a texture you can assign to your TV. This method should result in smoother animations, but it'll probably be more expensive (since Unity will need to continously render the camera's input) and forces you to rely on in-game real-time animations, so you'll have to take the rendering budget into account.

I'm not sure if there's some kind of video-to-texture support or something along those lines, but I guess you could look into it.

 

[Tashi]

Well, theoretically, yes, but if you are already simulating gravity by giving the character a vertical acceleration you would just need to give it an upwards initial velocity (say, they collide, and the character gets vspeed = 10), and some horizontal velocity (which depends on the side of the enemy he's hitting, I suppose). The gravity is going to take care of giving the movement that parabollic feel since that is what causes it anyway.
If you want to know what direction the character should be pushed back to, one way would be to take the difference of the x position of the center of the enemy and the x position of the center of the character. If it's positive, he should go right. If it's negative, left. Additionally, you can also use the difference of the y position to change the feel of the pushback, but that depends on how you want it to feel (more 'organic' I guess vs more hardcoded).

Holy shit you're so right. That would be so much easier. I mean, what I have for the character gravity is super simple but I could totally implement it for that. Thank you!