Support NeoGAF

[timetokill]

WIP
Guys, I am making some nice Sprite Shaders for Unity which gives the following goodies besides Diffuse Lightning:

I am making those Shaders with the Intent to sell them on the AssetStore but I am honestly developing this kinda blind because I don't know if 2D Devs are really interested in this.
Scene Lightning and handling Materials is something most 2D devs aren't familiar with and it changes kinda the workflow.

Awesome!

I'd be interested. I'm making a (third-person) game in a voxel environment and I'm trying to decide on voxel characters or 2D sprites. One of the reasons I'm leaning voxel is because of the lighting stuff. My game is heavily based around fire effects so it's pretty crucial.

 

[Jacksinthe]

WIP
Guys, I am making some nice Sprite Shaders for Unity which gives the following goodies besides Diffuse Lightning:

• Curvature (aka Normal Mapping)
• Surface manipulation (eg. rough or glossy)
• Light Emission with Masking and Tint (eg. eyes glow Yellow in darkness)
• Reflection with Masking (use Cubemaps, Reflection Probes or even normal Sprites)
• X-Ray with Density Settings (a second sprite gets revealed by light from behind)
• Nice Material Inspector

Gifs

I am making those Shaders with the Intent to sell them on the AssetStore but I am honestly developing this kinda blind because I don't know if 2D Devs are really interested in this.
Scene Lightning and handling Materials is something most 2D devs aren't familiar with and it changes kinda the workflow.

Depends on how much extra work is involved with maps for that lighting but there are a LOT of amazing things you can do with 2D lighting for presentation.

 

[mStudios]

Seem like I should start to sell Unity effect assets. Aren't there many
already? However. Well, the code isn't that hard. It was just meant as an
exercise for the 1d scaler I've given away recently. It goes something like
this;

If you're interested in selling the source code, I'm willing to buy a few. I want to apply them on my game.

Just let me know!

 

[Mental Atrophy]

Oops. So awhile ago I made my characters composite sprites. Characters can change class/job like Final Fantasy Tactics or similar, but I didn't want to have giant sprite sheets with every possible class they can be. So I made it so characters only have a head sprite, which is essentially pasted onto the class body at rendering. Each character also has a primary color to differentiate them as well, so whether Johnny is a knight or a wizard, he'll have red armor or a red robe or whatever. What I didn't consider is skin color, so even though Johnny's head implies he's a black dude, his hands are light skinned. I guess I need to make skin color palettized as well.

 

[gundalf]

Depends on how much extra work is involved with maps for that lighting but there are a LOT of amazing things you can do with 2D lighting for presentation.

You only need a single Normal Map and the optional Masks are just Gray-scale Sprites.
And doing Normal Maps is not difficult anymore because the latest Photoshop CC release comes with a Normal Map Generator build in (good timing adobe!).

 

[Noogy]

okay, yeah, that is a lot going on. I figured it was very busy, and I guess that's why I was having trouble identifying exactly what all was going on with my eyes.

you were limited by memory constraints when originally making this, and this reduced the number of graphics you used, right? I wnat to see what you come up with next..

Yeah, memory was always a huge issue. Being inexperienced, I look back and realize that where I wasted most of my memory was actually on audio, but that wasn't where I did my trimming since I didn't know any better.

As for my next couple projects, I'm probably not going to try to do anything as crazy as Dust. That's mainly because I want to fiddle with different genres.

 

[Poetic.Injustice]

I'm just now starting to realise how much work animators have to put in a 3D model after animating a single quadruped creature for the past few weeks. :/

 

[MrNyarlathotep]

I am honestly developing this kinda blind because I don't know if 2D Devs are really interested in this.

Have you looked at Sprite Lamp?

Other than competing on Sprite Lamps feature set, things I would imagine would be useful / desirable to add at a shader level
- mobile support for as low end devices as possible (a majority of Unity products are in mobile right now)
- HSL From -> To recolouring (ie select two HSL values with a colour picker and it will recolour option 1 into option 2 - this would solve Mental Atrophys problem described above, for example)
- Animated textures, either from sheet or via texture offset, with configuration options for direction and speed (ie add a water tile, configure shader, auto waterfall)
- Texture transitions (ie "dissolves" from one spritesheet to another or to nothing)

 

[gundalf]

Have you looked at Sprite Lamp?

Other than competing on Sprite Lamps feature set, things I would imagine would be useful / desirable to add at a shader level
- mobile support for as low end devices as possible (a majority of Unity products are in mobile right now)
- HSL From -> To recolouring (ie select two HSL values with a colour picker and it will recolour option 1 into option 2 - this would solve Mental Atrophys problem described above, for example)
- Animated textures, either from sheet or via texture offset, with configuration options for direction and speed (ie add a water tile, configure shader, auto waterfall)
- Texture transitions (ie "dissolves" from one spritesheet to another or to nothing)

Thanks for your suggestions, this helps me a lot to find some concrete directions and the last two options is something I am already looking at!

But I must say that I don't see me competing with Sprite Lamp, since that is a Normal Map maker tool while mine is Shader which is feed with said Normal Maps

Edit
Regarding HSL Coloring, this is something I am aware about but it's something I really don't want to touch because there is a huge amount of Assets already on the Market which do only this one task, and they do it quite well!

 

[MrNyarlathotep]

But I must say that I don't see me competing with Sprite Lamp, since that is a Normal Map maker tool while mine is Shader which is feed with said Normal Maps

You're right, "compete" is probably the wrong word... as SpriteLamp is at the production level and a shader would be at the implementation level, it was more pointing it out to you if you were unaware of it that there is obviously enough desire for this for SpriteLamp to get kickstarted in the first place, and I guess it would be worth your while to see what features SpriteLamp offers to see what things people are looking for (and potentially offer a direct pipeline from SL through to your shader)

 

[gundalf]

You're right, "compete" is probably the wrong word... as SpriteLamp is at the production level and a shader would be at the implementation level, it was more pointing it out to you if you were unaware of it that there is obviously enough desire for this for SpriteLamp to get kickstarted in the first place, and I guess it would be worth your while to see what features SpriteLamp offers to see what things people are looking for (and potentially offer a direct pipeline from SL through to your shader)

Ah now I get it! Thanks for pointing it out and I will toy with Spritelamp ASAP and compare it with my other tools (Quixel Suite, Bitmap2Material and Substance Painter).

 

[Jacksinthe]

You only need a single Normal Map and the optional Masks are just Gray-scale Sprites.
And doing Normal Maps is not difficult anymore because the latest Photoshop CC release comes with a Normal Map Generator build in (good timing adobe!).

I'm in.

Let me know if you need me to test anything.

 

[MrNyarlathotep]

So I missed the news that Adobe have bought out Mixamo but of potential interest to indie devs;
- There are two high quality anim packs on the Unity asset store right now for free (as theyre not taking any payments while they upgrade to being part of the BorgAdobe Collective), and all their other animations have been removed from the store
- Mixamo Fuse (character creation app thats compatible with Substances) is straight up free now AFAICS.
- On the Mixamo website there are a fucking ton of free high quality rigged models available, and although they're rigged for selling you Mixamo animations, the actual rigged models can be imported into Unity and UE4 very easily (I gave both a quick try; they work flawlessly with Units Mecanim system by importing as a humanoid type mesh, and they are easily converted to UEs skeleton by rotating 270 degrees on import)
- Anyone signing up for a Mixamo account gets 10 free autorigs of their own models and 20 free anims of your choice from their pretty damn extensive selection. You can't actually buy anything there right now.

 

[missile]

I bet you could make a bunch of interesting unity assets.

If you're interested in selling the source code, I'm willing to buy a few. I want to apply them on my game.

Just let me know!

Let's me explain a bit where I'm heading with all of this filter stuff
before I will roll some new ones.

In a couple of month, hopefully, I'm going to release some analog video
filters being part of the CRT TV simulator which will be a different release.
These analog video filters will be much more interesting than those little
gems I posted over here. With just a few parameters you will be able to get
some pretty cool video distortions, transitions, tearing, ghosting, folding,
etc. which can all be made dependent on the action of a game. At the end of
the day I'm going to port it all over to Unity, I guess. But for the time
being it's all C/C++ code. C/C++ will be the primary platform/implementation,
since I want to target many different platforms, i.e. I want to also target
some graphic tools like Photoshop, AfterEffects, etc. in building plugins to
make the filters available for artists and animators.

One of my goals within this regard is to make the video and CRT filters to
overlay a canvas in one of these tools such that an artist can draw his/her
graphics straight under the influence of said filters, in realtime. Hence, you
will set you parameters for the filters which will fit a certain aesthetic of
your game etc. and start drawing your stuff with the filters online. This is
quite interesting for doing retro graphics and other sort of graphics (say;
video/analog graphics). But it also serves another purpose I have in mind,
i.e. in controlling the video filters themselves -- as you draw. I hinted
about it while talking about the over-modulation characteristics and how it
can influence the video signal. So when you draw a picture with the filters
online you can immediately see how your colors may manipulate the filters
outcome/effect. For example, if you use lots of saturated colors in
combination, then this may lead to video distortions if a specific limiter
(de-saturator, or clamping -> producing higher harmonics) in a given filter
isn't active. The interesting aspect is that you can see it straight on the
screen while you draw. Speaking from another perspective; you can actually
design video distortions by drawing pictures (or sequences thereof) which may
have nothing to do with your actual game graphics but can if you so choose.
This will actually give an artist control over some video filter aspects
without needing to know any technical knowledge. Those pictures can later be
saved as a parameter-set to be imported into the game. When the video filters
get presented with such pictures (given the same filter parameters during the
design stage), it will exactly reproduce the distortion in your game the way
you designed them on the drawing board. Cool, no?

I want to use it for my own games. But there is some more work to be done
here. Am still in the design phase so to speak. Another problem I'm confronted
with is performance. Many of the video filters will put your system under a
heavy load. Therefor, I'm looking for ways to cut down on this load. Signal
resolution and the resolution of the discrete/analog components and their
digital approximations etc. need to be addressed right from the get-go. Hence,
currently I do experiment with some image-based algorithms to see if I can cut
down on the load of some of the stages to either replace or mix them together.
Within this regard I'm going to implement and experiment with another
new distortion algorithm I have in my mind up next which will allow me to
shear an image along multiple arbitrary-oriented lines. I want to see if I can
lessen the burden on the deflection beam of a TV (to a given degree) to draw
the image using parts of such an algorithm, not saying this algorithm won't be
quite computation expensive as well. However, I'm pretty sure that this
algorithm will likewise be very useful in its own right, for some more cool
screen transitions to come.

Anyhow. Regarding the source code and stuff you asked for; you can get these
effects for a good beer! I just need to get the hang of Unity somehow.
Yeah, would be cool rip one or two algorithms out and port them over to Unity.
Problem is, I currently don't have the time to learn all of Unity's caveats
and read their docs etc.. But if someone may point me to a good clean shader
setup procedure with the least overhead and stuff, I may give it a try. But
don't quote me on this! xD

 

[gundalf]

I'm in.

Let me know if you need me to test anything.

Sure, I am glad to have some Beta Testers!

 

[Jobbs]

Yeah, memory was always a huge issue. Being inexperienced, I look back and realize that where I wasted most of my memory was actually on audio, but that wasn't where I did my trimming since I didn't know any better.

As for my next couple projects, I'm probably not going to try to do anything as crazy as Dust. That's mainly because I want to fiddle with different genres.

I feel buried under Ghost Song, and it's not as elaborate overall as Dust, I wouldn't say -- but it's also apples and oranges.

But even still, I'm not sure I want to do something quite like it immediately after, so I understand how you feel. I have an idea to do a game that would be more impressive but not have quite as large of a scope (in most ways). I want to make a more living world, with day night cycle, seasons, dynamic plant life, and other stuff -- and let the game design live on top of that a bit more, so it wouldn't be quite as tedious designing the entire game. It would also have a notably smaller game world. Anyway, hard to explain exactly what my idea is without writing a novel, but that's kinda a clue as to where my head is at for next time.

Btw, you should participate here by showing clips now and then of whatever you might be working on.

 

[Lautaro]

I think I'll leave the next week only for QA but I'm pretty much ready for the release of Nomad Fleet. I was thinking July 14 but if someone has a space or strategy game for that date maybe we can discuss a change privately.

Also I have space for more testers, I can't pay you but I can give you a key and a mention in the credits. If anyone is interested, please send me a message.

 

[Jacksinthe]

Sure, I am glad to have some Beta Testers!

PM sent

-

Pretty sure we just hit our final cameo confirmation :O

 

[mStudios]

.

Oh, if you're going to use them for your game, that's fine then! I'm asking for them because my programmer does not know anything about shaders so far and we could use someone who can make a few.



A quick sketch of a new concept art of both Alices!

Spoiler

Dat murderer face.

 

[Unain]

Been having some fun with generating roads for my randomized city.

- Dark blue = Corner piece
- Light blue = T-Split piece
- Gray = Straight piece
- Rest is empty space (will populate those places with buildings/etc later).
Sorry for the weird color scheme.

All done in UE4 blueprints so far, but will convert it to C++ a little later. I want to have a proper working prototype first.

Here is an insight on how I'm doing this:
- First populate an array with the edge cases (buildings on outside and 1 road on the inside of those buildings). This is always the same.
- Generate multiple straight roads (randomly placed, can never touch each other) on the inside lane of the outer edge road.
- Then I loop through those roads and place roads in front (or right/left, never back) of it until it hits another road.
- There might be some tiles (undefined tiles) that are in big clumps, so I get rid of those by looping through the undefined tiles.
- At the end I do one more loop to add the T-Split/X-crossing tiles and fix corners/straight roads if they have been adjusted by the undefined tile loop.

There are still some roads that could end in a dead end, but I might keep that. Or it might annoy me enough to fix it .

 

[Jumplion]

Whew, back from dealing with schoolwork. Working on some small game with a few others, might put up the prototype of that when we get the basics all laid down.

Been having some fun with generating roads for my randomized city.

- Dark blue = Corner piece
- Light blue = T-Split piece
- Gray = Straight piece
- Rest is empty space (will populate those places with buildings/etc later).
Sorry for the weird color scheme.

All done in UE4 blueprints so far, but will convert it to C++ a little later. I want to have a proper working prototype first.

Here is an insight on how I'm doing this:
- First populate an array with the edge cases (buildings on outside and 1 road on the inside of those buildings). This is always the same.
- Generate multiple straight roads (randomly placed, can never touch each other) on the inside lane of the outer edge road.
- Then I loop through those roads and place roads in front (or right/left, never back) of it until it hits another road.
- There might be some tiles (undefined tiles) that are in big clumps, so I get rid of those by looping through the undefined tiles.
- At the end I do one more loop to add the T-Split/X-crossing tiles and fix corners/straight roads if they have been adjusted by the undefined tile loop.

There are still some roads that could end in a dead end, but I might keep that. Or it might annoy me enough to fix it .

Procedural generated stuff is facinating to me. I'll need to experiment with that stuff some time. Looks cool though!

 

[Raide]

Heya Indie GAF,

After years and years of coming up with ideas with a programmer friend. We are toying with the idea of looking for an artist. We managed to grab another programmer to float around some thoughts and ideas so we can actually pick something and move forward.

We seem to be doing most things in Unity, so either 3d or 2d would be great. We are all super amateur, so we are not expecting AAA stuff here. PM me if you want a bit more info.

 

[ken_matthews]

So I decided to download the Substance Painter demo because it's rather difficult to create materials for complex objects in Quixel.

So this is what I managed to come up with in about an hour (I think i looks fairly good for my first go around, or at least I am happy with it).

So I like the workflow much better, and I think it is easier to get better results (especially with more complex objects). I think I am going to buy it, but now my question is, how important is Substance Designer? Do I really need it? I've seen some tutorials for baking in mesh curves, and world space normals, for use in Substance Painter, but is this really necessary? If I need those maps, can I do it a different (free) way? If I hadn't just purchased Quixel for $250, I'd just pick up both, but now I think I made a mistake. I should have listened to MrNyarlathotep... :/

 

[Huw_Dawson]

Checking out Adventure Creator for Unity this weekend. It's pretty neat! Especially given the prior best-in-class indie engine was the cumbersome AGS.

I think I have a neat little project I can hack together this weekend.

 

[MrNyarlathotep]

I've seen some tutorials for baking in mesh curves, and world space normals, for use in Substance Painter, but is this really necessary? If I need those maps, can I do it a different (free) way?

I have some really good news for you

Substance Designer is used for creating Substances, and Painter is used for applying them - as you've probably found out in the demo, the stuff you can do just in Painter is really versatile.
Whether or not Designer is worth your money is probably going to depend on your workflow - if you think you're likely to be doing a lot of things manually repeatedly that it would be better to just have as a substance you can just apply (for example if you're constantly making your own tile textures and making minor adjustments to avoid repetition, where a designed substance can just have a random seed to do that for you) it will probably be worth it, if you're happy with the workflow of using and combining the premade substances provided in Painter (because honestly, there's not that many different types of metal) you're fine with Painter.

As I think I posted before, they have a cool rent-to-buy program if you're in doubt that might be a better alternative to just dropping the cash right now too.

 

[ken_matthews]

Do you think the subscription to the Material Library is worth it?

 

[MrNyarlathotep]

Do you think the subscription to the Material Library is worth it?

I honestly don't know - like most subscription type things its a yes if the thing they give away is something you happen to need right now, and a no if its not and you need to go ahead and make your own.

If you're planning on picking up textures as needed I would guess a GameTextures subscription would be much more useful as you can select exactly what you need when you need it.

As a bonus for trying out Substance Live its probably a pretty neat bonus to stash away stuff you might use in future, but I wouldn't want to be reliant on it if you see what I mean.

 

[shacklesmcgee]

Substance Painter and Substance Designer are life changing, you will not regret the purchase. I've fallen in love with Designer a tad more for the ability to experiment with procedural textures, but Painter is so easy for single assets I'll never use Photoshop again.

 

[Davision]

Some progress with my first person painting game:

 

[ken_matthews]

^^^ wow, that's looks really good!

 

[Unain]

Procedural generated stuff is facinating to me. I'll need to experiment with that stuff some time. Looks cool though!

Thanks, procedural generated stuff is pretty easy until you want specific rules. It gets pretty crazy in the end.

Like yesterday, I was creating a function to get rid of clumps (tiles of 4 roads clumped together). I made a whole lot of branches (if-statements) and got it to work. It only worked on some seeds.
Today I had an epiphany where I could compress all those branches in 4 branches by just looking at the 3 neighbouring tiles (x4) and making my decision based on that. Made it a whole lot easier and more readable.
Part of that blueprint: http://i.imgur.com/0GPD2TV.png
Used to be 4x as big.

My function list looks like this:
http://i.imgur.com/RrK6iuI.png
(Might add some categories tomorrow and do some refactoring, because it's getting a bit big)

The hardest part, debugging. Luckily with UE4 you can make a "Random Stream" variable that you can use as a seed in your Randomization.
This way your randomization will always be the same with the same seed. But it's still a lot of loops and is annoying to debug sometimes.

 

[planckpixels]

Some progress with my first person painting game:

Thats pretty dam neat

 

[Huw_Dawson]

Hey folks,

Like I said above, I found Adventure Creator and it's really inspired me to make something. In addition, I recently found out a really cool thing about my favourite adventure game - Discworld Noir - in that all of the game's audio (but sadly not its FMVs as nobody has written an FFMPEG module that can read them ) is unencrypted and ready for anyone to do anything they like with.

I reckoned getting the really early parts of the game built would be a cool project that'd challenge me to learn more Unity, and do some level creation in 3DS Max which I have not done in a couple of years.

Tonight I rigged up some scripting inside Adventure Creator that covers the original narration of the opening cutscenes. No 3D graphics at all yet, but this is a learning process even after reading and watching a bunch of tutorials. Currently the project boots, plays the correct dialogue with subtitles in the correct place on the screen, loads through the correct scenes to progress and cheerily informs me that all my scripting works.

Just to see how utterly unimpressive this is: video!

https://www.youtube.com/watch?v=Q7hIqhgeXUQ&feature=youtu.be

Every weekend and evening I spend on this, I'll try and upload a video showing what I did.

 

[Jumplion]

Some progress with my first person painting game:

Gorgeous. This might be on purpose, but it reminds me of that one Robin Williams movie What Dreams May Come. It has some scenes that has practically the same aesthetic and it's equally as purty to look at.

Thanks, procedural generated stuff is pretty easy until you want specific rules. It gets pretty crazy in the end.

Like yesterday, I was creating a function to get rid of clumps (tiles of 4 roads clumped together). I made a whole lot of branches (if-statements) and got it to work. It only worked on some seeds.
Today I had an epiphany where I could compress all those branches in 4 branches by just looking at the 3 neighbouring tiles (x4) and making my decision based on that. Made it a whole lot easier and more readable.
Part of that blueprint: http://i.imgur.com/0GPD2TV.png
Used to be 4x as big.

My function list looks like this:
http://i.imgur.com/RrK6iuI.png
(Might add some categories tomorrow and do some refactoring, because it's getting a bit big)

The hardest part, debugging. Luckily with UE4 you can make a "Random Stream" variable that you can use as a seed in your Randomization.
This way your randomization will always be the same with the same seed. But it's still a lot of loops and is annoying to debug sometimes.

I keep telling myself I'll get around to learning Unreal but I always get distracted by other things/schoolwork. Need to get around to that pronto before I start slacking off even more.

 

[gundalf]

Awesome!

I'd be interested. I'm making a (third-person) game in a voxel environment and I'm trying to decide on voxel characters or 2D sprites. One of the reasons I'm leaning voxel is because of the lighting stuff. My game is heavily based around fire effects so it's pretty crucial.

Sorry for the late Answer but I didn't got it at first what you where talking about, because when I hear Voxels, then I always think about 3D Games but now I remembered Terraria which is a 2D Voxel game! So yeah, if you are making this kind of Game, you can get a Beta Version if you like!!
Just write me a PM

@Davision
https://www.youtube.com/watch?v=UskxgxxJWIA&feature=youtu.be

 

[_machine]

Some progress with my first person painting game:

That looks gorgeous! I also think the nature theme and painting mechanic are a lovely match

 

[SpacePirate Ridley]

Oh, if you're going to use them for your game, that's fine then! I'm asking for them because my programmer does not know anything about shaders so far and we could use someone who can make a few.



A quick sketch of a new concept art of both Alices!

Spoiler

Dat murderer face.

Lovely sketch!

Some progress with my first person painting game:

Damn Davision, your game looks absolutely fascinating every time you show it.

 

[Feep]

[snip]

Pretty amazing stuff. Is this just a post-process filter on a normally rendered scene, or do you do some crazy texture stuff too?

 

[Jacksinthe]

Some progress with my first person painting game:

:OOOOOOOOOOOOOOOOOOO

 

[mStudios]

.

Thanks!

Click for video.

Everytime a character switch there is a Transition. When the character changes poses, it fades out and in. Does it look good?
Also, I found a glitch when there is a character switch.

 

[Davision]

Pretty amazing stuff. Is this just a post-process filter on a normally rendered scene, or do you do some crazy texture stuff too?

Just a post processing shader that I'm still working on, the goal is that I have one shader which can turn a scene in lots of different styles by changing parameters and shader textures. Obviously it will also require different artwork then for more abstract/simplified painting styles. That scene is just a realistic looking country side from the UE4 marketplace: https://www.unrealengine.com/content/dc872c9def304b4291f221154f0225c4

Gorgeous. This might be on purpose, but it reminds me of that one Robin Williams movie What Dreams May Come. It has some scenes that has practically the same aesthetic and it's equally as purty to look at.

Yea that scene is very much being inside a painting. I'm also aiming for more abstract styles though, so there should be lots of variations of different painting/worlds in the game.

 

[missile]

How is selling shaders on the Unreal Marketplace, anyone has some experience?

Oh, if you're going to use them for your game, that's fine then! I'm asking for them because my programmer does not know anything about shaders so far and we could use someone who can make a few. ...

What sort of effects do you need regarding your game? I may work towards some
of them while working on some new stuff of mine, and may start to port some of
my filters over to Unity and Unreal. Your are on Unity, right?

 

[Blizzard]

:OOOOOOOOOOOOOOOOOOO

It must be the Monet! </painter jokes>

 

[slash3584]

So GAF what would you recommend me to use to create a fully procedural (towns,biomes,etc) 2D top down action rpg?

From what I've gathered this kind of stuff is better done from the ground up without using Game Maker or the like.

 

[Huw_Dawson]

Spent most of today getting used once again to model creation in 3DS Max. Blocked out Lewton's office and had a pop at a very basic desk.

Everything is taking a long time at the moment because I'm so rusty.

Screenshot is from Unity.

There's tonnes of awesome stuff in this thread, and I'm literally just getting my toes back into the water. :O

BTW, reference from the original game:

Probably going to focus on completing the desk next to have a clear idea of the whole modeling and texturing workflow I'll need for the project.

Out of interest: What do people think about definitions of done in game development? By that, I mean - for everything in this project, my philosophy is to finish things to a shippable quality before moving on. Is it realistic to develop a normal game via lots and lots of little individual vertical slices?

 

[soapOD]

Finally got some explosions and camera shake implemented! Shitty GIF ahead:

 

[Jobbs]

.

Mind if I ask what the object of the game is? How does the gameplay work?

 

[missile]

... for everything in this project, my philosophy is to finish things to a shippable quality before moving on. Is it realistic to develop a normal game via lots and lots of little individual vertical slices?

Depends on the complexity of the game. If the game is complex (has many
interdependencies, quality of components do depend on each other), it will be
better to distribute the weights/work more equally on all components of the
game and using said components to build an approximate vertical slice of the
game/prototype, instead of finishing one component after another to a
"shippable quality" which may require you to change it later again when having
done some of the other components.

Instead of trying to split the game into independent components (which may
become rather hard or impossible if the game is more complex) and trying to
build each component to a shippable quality, you start out with a very rough/
approximate solution to your game with each component just doing a few basic
things (while getting a vertical slice of the game right there). And then you
start to refine each component slightly (not putting too much weight on just
one component) to arrive at another (closer) approximate solution to your
final game. You basically iterate though the loop until you reach the intended
quality of the game.

Some of the advantages are;
- you always have a running (approximate) version of your game
- you won't run the risk of putting too much weight onto just one component
which may later turn out to be a wast of time (most likely for complex games)
- you can judge the improvements towards the final game by comparing the
current iteration to the specs, and you can also estimate the rate of progress
by comparing each iteration with the previous one, which may also tell you
when to stop (diminishing return)
- your game will be of uniform quality

Some of the disadvantages are;
- you need a good overview of the whole project right from the beginning
- you need to understand a good portion of the theory behind all the
components you want to realize and also how they may influence each other to
a given degree
- not easy to parallelize (workers need to look at the progress of others)

However. If the game is of low complexity you may try to divide the game into
independent components and produce each component to the maximum quality. This
is the divide-and-conquer method -- which strictly assumes the independencies
of the components. Yet the game can also be complex as long as it can be split
into independent components. The main advantage of this method is that you can
assign a worker to a component with him/her working at a speed independent
of the progress of all others. But there are disadvantages as well; you will
never really have a working game up until the end. Further, the final quality
of the game needs to be kept upheld during the whole development cycle.
Reallocating resource during the project may lead to a game not having the
property of uniform quality, i.e. some components of the game will be superior
while others will not due to special events in the development cycle leading
to a rescheduling of resources and as such may cut on resources to produce some
of the components to maximum quality.

It is perhaps a good strategy to combine both approaches together by using the
first one as the main one with the second one applied to sub-components of
components of the first one, if possible.


Edit:
Historically, the method of iterative refinement has lead to the first
airplane*. It is impossible to build an airplane with the divide-and-conquer
method. Contrary to all who came before, the Wright Brothers have put equal
importance to each component, i.e. they iteratively optimized each component
with respect to each other and have solved as such one of the most complex
systems of all time (measured on the centuries it crossed). This approach was
later adopted by the whole industry (full-system approach) and as such it
became possible to solve rather complex interdependent problems. The airplane
is perhaps one of the greatest achievements, but the method the Wright
Brothers exercised is without any doubt one of the greatest of all time.

* An airplane which could sustain itself in the air.

 

[missile]

.

How are things, Feep? And what are you prototyping?

 

[Feep]

How are things, Feep? And what are you prototyping?

I'm good! Can't quite say, unfortunately, but I will tell you that there are cones and even a few lines in it.