next »

[JobLeonard]

Oooh, like how the Homeworld games did sky boxes?

That's amazing!  I had no idea.  What a fascinating solution to a design problem.  I love vertex colors.  They're useful for so much - beyond colors, too, of course.
My stars are just disconnected tris projected onto a sphere, facing its center, with vertex colors just to keep it simple - so, not much like the Homeworld skyboxes, right now.
Almost everything you see in the game, though, is purely vertex colors.  I kind of find texturing tedious, and I do genuinely like the look of flat-shaded and some low-poly art.  So, I thought I'd give it a try myself.  It hasn't saved me much time so far, but once I get into the art production, it really should help. 
The only thing that makes major use of textures is the background sky gradient.  I tried it with just vertex colors but it was slow.  The current setup uses a set of four, two-pixel textures and blends between them.  I rather like it and I don't think it taxes the GPU too much.

(that whole anecdote could be a narration in the game hahaha)

I'm starting to come round to your idea of having some narration - perhaps for a trailer, as you once noted.  Perhaps in the game itself.  I do like to tell stories!

I love it!

Thanks, I love it, too!

Wow, a lot of updates since I last saw this thread. Really cool stuff!!

Thanks for checking-in. 
LOVE the time bubbles in AO, by the way.  I'm curious to see how you're going to deal with different geometry in different bubbles. 

It hasn't saved me much time so far, but once I get into the art production, it really should help.

Should have the side-benefit of keeping asset sizes really tiny (and graphics card memory requirements a bit lower, although I doubt texture sizes would be that significant for a game like this). Interesting how it becomes really taxing for the GPU though, maybe some vertex culling tricks could fix that?

EDIT: Pagination

[oahda]

Thanks for checking-in. 
LOVE the time bubbles in AO, by the way.  I'm curious to see how you're going to deal with different geometry in different bubbles.

Aw, thank you too!

[amasinton]

It hasn't saved me much time so far, but once I get into the art production, it really should help.

Should have the side-benefit of keeping asset sizes really tiny (and graphics card memory requirements a bit lower, although I doubt texture sizes would be that significant for a game like this). Interesting how it becomes really taxing for the GPU though, maybe some vertex culling tricks could fix that?

Yes, I'm very pleased to save some disc space with having very minimal textures - but I am very interested to see how the GPU handles doing everything with vertex streams (because I also only have three key materials in the game, applied to pretty much everything, and variation and effects are done via vertex colors and manipulating/abusing UV channels.  My expert plan is to just keep piling more and more on the GPU until my lower-spec dev machine (a laptop with integrated graphics) falls over.  What a plan!

[amasinton]

Procrastination #2: Planting Daisies

I didn't have a lot of time this last week to make much progress.  Such is life.  I've been focusing on my little roadmap, and the first leg of that journey is the save/load system.  I just couldn't get it to load.  Turns out, that's because I didn't know that you couldn't just compare byte arrays for equality with a simple:

Code:

if (byteArrayA == byteArrayB)
{
   DoSomething();
}

All of the core building blocks of the game have unique GUIDs and the loading system relies on those to reconstruct the game state from the last save.  The GUIDs are byte arrays.  I was being a little simple in my GUID checking.  So, having figured that I actually need to compare arrays using a for loop, I am now able to load my saved game state.  Success!  But nothing pretty to show for it.

However, I also distracted myself over the past couple of weeks with finding a way to add some variety to my grassy topsoil by scattering daisies.  This is fun!

Last year Unity bought the PolyBrush asset and has recently made it available to everyone.  Awesome, because, among many other things, it has a scattering system to place meshes on top of other meshes.  

Only, because my topsoil meshes are composed of hundreds of separate, watertight fragments, the PolyBrush system wants to paint on every face of every chunk of topsoil - even the faces hidden under/behind other faces.  

So, I came up with a two-step workflow.  Step 1 is to scatter my daisies onto a plane hovering above my topsoil.  Then, I disable the plane and use a "Drop to Surface" script (found here) to drop those daisies onto the topsoil beneath.  The drop script also makes the daisies children of the topsoil chunks they land on.  This is nice because the daisies will de-activate when their parent does, and I don't have to worry about pulling up the daisies myself on every chunk of topsoil I remove during gameplay.  Yay!

The next thing I want to do with this system is to make the daisies instances using Unity's Graphics.DrawMeshInstancedIndirect.  But that's not on the roadmap yet, so I'll set that one aside.

Here's a blurry video of planting daisies.  Enjoy!

I'm going to be a bit scarce for the next couple of weeks because I have to shift gears and finish a project I was commissioned to do ages ago.  Hopefully by the end of the month I'll be fully back to the swing of this.

[kyliebriggsmusic]

Whoah, such a cool concept!! Really inventive and different than a lot of what's out there, and so cool that it's inspired by your own experience.

Do you have a composer involved, or are you looking for one at all?

[TheElementalist]

Really love the style and the concept. I know it's hard to stick through to the end (and where even is the end) but I really hope you do. I love the play on time and location.

[JobLeonard]

I am very interested to see how the GPU handles doing everything with vertex streams (because I also only have three key materials in the game, applied to pretty much everything, and variation and effects are done via vertex colors and manipulating/abusing UV channels. My expert plan is to just keep piling more and more on the GPU until my lower-spec dev machine (a laptop with integrated graphics) falls over.  What a plan!

I basically have no experience with graphics programming, but isn't part of it also the art of knowing how to "batch" draw calls in the most efficient way and with the least amount of recalculations and lookups?

[amasinton]

I basically have no experience with graphics programming, but isn't part of it also the art of knowing how to "batch" draw calls in the most efficient way and with the least amount of recalculations and lookups?

Yes, that's a big part of it.  That's why I only three or four materials and vertex data do 90% of the work - allows batching and minimizes draw calls while still allowing for some variation in the appearance and behavior of the objects.  But I still wonder, in the end, what the trade-offs really are in terms of rendering efficiency, once everything is in-place and all said and done. So, I will let the question be academic until it begins to make a real impact on the game.  Just putting my fingers in my ears and closing my eyes, you know. 

It IS very nice that I don't have to set specific materials and textures for individual objects in the scene, though.   

[amasinton]

Really love the style and the concept. I know it's hard to stick through to the end (and where even is the end) but I really hope you do. I love the play on time and location.

Thank you!  The project keeps moving along - slowly but pretty steadily.  Sometimes there's not much visual progress, but that's okay.  Sometimes you just have to build stuff behind-the-scenes.

I think a lot about time and location - kind of force-of-habit as an archaeologist.  I'm always in two or three time periods at once.

[amasinton]

Whoah, such a cool concept!! Really inventive and different than a lot of what's out there, and so cool that it's inspired by your own experience.

Do you have a composer involved, or are you looking for one at all?

Cheers!  Thanks for the encouragement.  It means a lot.

I'm thinking about music very much - but that's about all I can do right now.  Part of my personal "meta-game" for this project is that it must pull itself up by its own bootstraps in terms of funding.  I do development on this right now as a side project, which means the resources are extremely limited.  I don't want to ask anyone to do sound or music because I don't have funds yet to pay them and I don't want to ask anyone to work based on the potential for profit-sharing down the road  - I don't think that's fair to the artist because who knows if anyone will buy this game when it's released?  That's the reality of indie development and I don't want to take advantage of people.  So, for now, I can dream about the music for the game, but not much else.  For now. 

Sigh ... and thank you!

[amasinton]

Time Travel Update!
The time travel system is now functional!!!

I haven’t posted for three weeks.  While a large portion of that time was consumed by teaching a class and working on a long-overdue project for a client, I was also hard at work on the game’s time travel logic.  And, as of this morning, I have a functioning system.

Behold!





Impressive, no?  Well … no.  YouTube compression issues aside (sorry so blurry!) it doesn’t look like much.  And a part of me really feels like it shouldn’t feel like so much, personally.  This was an easy problem to solve, after all.  Right?

Hmmm.  Not so much.  Not for me.

The Basics
As I’ve said before, time travel in this game lets you toggle between two ends of a wormhole.  One end is a time you set based on the creation dates of artifacts you choose from your inventory and which you find while playing – the other end is the time you are currently in.  So, find a silver penny of Edward I – open up a wormhole to 1280 and step through.  The other end of the wormhole is the date you just left.  You can jump back and forth at will.

Lots of games feature this kind of time travel.  Zelda games do this lots and lots, in fact – sometimes toggling between three or more time periods.  This is an old trick and how it’s done should be part of the standard game dev box of tricks, like normal map baking and making players harder to hit when their health is low.

Where it Gets Tricky
The tricky part is that in this game the time travel has to be affected by the player’s ‘conversation’ choices with the ruins they discover – choices which almost always have a backwards influence.

In the video the player has a choice between two arch shapes: a pointed arch (Gothic) or a round arch (Romanesque).  The player chooses Gothic and then jumps back to a point in the past when that part of the building had a Romanesque arch.  If the player had chosen a Romanesque arch, there would have been no change to that portion of the building when traveling through time.

What the player chooses in the present has consequences in the past.

So, I have to make a system that accounts for player choice – choices which I can’t predict, but which will have meaningful effects.  Also, their choices should exist within a system which allows maximum flexibility for me as a designer and which will not require me to think of all the branching paths before-hand.

Also, adding to the tricky, is that I’m not a terribly logical thinker.

How I ‘Solved’ It
I know the building’s history.  Well, a history of the building.  Behind the scenes the player’s choices set off sequences which lock and unlock portions of the building’s past.  This isn’t a proper combinatorial system (I’m not that clever), but it does mean that a building’s past can differ in unexpected ways across different playthroughs (if it’s a complex building with many player-reconstructed parts).  I set up the past, and the player knocks holes in it.

This is what I experience when I work with a building as an archaeologist.  The building has an established past – but it’s one I don’t know and can’t know completely.  Based on the clues that survive, I can construct a hypothetical past which will share some things in common with its actual past, but which will also differ in some areas.  I will never know.  My choices in the present effectively change the building in the past.

The systems and code behind all this in the game is where I got stumped for three weeks.  I even read James Gleick’s Time Travel: A History to help straighten out my thinking.  I created system after system to make sure that I could toggle between ends of the wormhole without ignoring the player’s choices.  Each system was more Baroque than the last.  And none of them worked.  I didn’t really understand what it was I was trying to do on the object-by-object level.  

I thought – a lot, I drew diagrams, it got to the point where thoughts of time travel put me to sleep almost immediately – as though I was trying to steer myself away from the problem entirely.

Then, when I was trying to work out the staff scheduling on the calendar at work the solution appeared: simply compare the date now and the travel date against the lifespan of every object, regardless of that object’s current state in the world.  Player-placed objects replace designer-placed objects when their lifetimes overlap.  Set the object’s visibility according to whether one, both, or none of the dates is within its lifespan.

Simple for someone who thinks a little more logically than I do.

But, in the process, I have learned a lot about how the systems in the game as a whole work together.  And which ones are redundant.  I’m understanding better and better what I’m trying to express.  This will serve me well when it comes to rebuilding the game to make it lean and more efficient.

Still on the Map
So, I’m now passing through the suburbs of the second destination on my roadmap: time travel.  I’m on my way out of town heading toward stabilizing the building system and, eventually, a vertical-slice demo.

Thanks for reading!

[JobLeonard]

Wow, looks great! I'm watching this without sound but can still somehow hear the "falling brickwork" sound effect as the thing reconstructs itself.

RE: YouTube blurriness, it's worth noting that YT tends to create better quality videos for lower resolutions from higher resolution uploads, even if those videos are nothing but upscaled low-res videos. It's weird like that.

[ChrisLSound]

It's been a bit since I checked in, but this is still looking awesome. Really cool to see the time travel in action. Great work as always.

[amasinton]

Wow, looks great! I'm watching this without sound but can still somehow hear the "falling brickwork" sound effect as the thing reconstructs itself.

RE: YouTube blurriness, it's worth noting that YT tends to create better quality videos for lower resolutions from higher resolution uploads, even if those videos are nothing but upscaled low-res videos. It's weird like that.

Yes exactly on the sound effect!  I've fiddled around a bit with creating a sound effect just like that.  I took it too literally, of course, and matched a variation of a single falling stone to each fragment, thinking it would give a unique, custom sound effect every time part of a building reassembled.  But it just made a cacophony and was slow as well.  A single, pre-rendered sound works better.  I've got a lot to learn in that department.  I'm looking forward to it.

And thanks for the tip on YouTube blurriness.  I've given it a try.

It's been a bit since I checked in, but this is still looking awesome. Really cool to see the time travel in action. Great work as always.

Thanks for checking-in and posting!  I've left things a long time, but there's a longer post incoming now!

[amasinton]

Cutscenes(ish)

After building the time travel system I headed into a very busy season at work and in family life.  I was exhausted and I wasn’t looking forward to another round of bug fixing as I brushed-up the building system.  However, there was another system I had not yet worked on, but which I knew I would have to come to grips with sooner or later: cutscenes.

“This should only take a couple of days,” I thought.

Six weeks later …





This is what amounts to a cutscene in the game.  It’s a little Obra Dinn, a little Recompile, a dash of Heaven’s Vault.  And boy, have I learned a lot!

While explicit narrative is something I’m trying to avoid in the game – I want the players to build that themselves, because narrative is the point of the game, just not my narrative – there are some moments of punctuation, which, like all of the other clues the player uncovers, will help the player to weave a narrative around what they think happened on the island of Archangel.  These moments are delivered via cutscenes, the recitative of video games.

Well, maybe less cutscene and more tableaux.  (I love cutscenes in games – it’s just there’s only so much time and there is still so very much to do.)  At certain points, ghostly figures emerge.  The player is free to explore these scenes, once triggered, as often as they like.  They are wordless moments with insubstantial human figures caught in time.  The scenes ask, “What’s going on here?”

So, why six weeks to implement this?

As usual, I didn’t quite know what I wanted and I didn’t know how to make it.  Also, I’m terrified of human figures in my own art.

In archaeology the one thing we never get are the people themselves.  We never meet the ones responsible for the sites we study, the monuments we survey, the artifacts we uncover.  We only know them through what they left behind, their bones and beads and buildings.  (The archaeology of the contemporary past is sometimes the exception.)  So, in my professional work, I almost never include people.  I don’t have a lot of practice in representing the human figure.  I’m an environment artist, not a character artist.  The Uncanny Valley is too haunted for me.

One thing I’ve always dreamed about is finding an artifact and being able to see how it got there.  Who lost this gold ring?  Were they upset?  What did it mean to them?  How long did they search?  

I want a lantern like Star Lord’s on Morag where he shone a light into the past, illuminating the people there going about their lives, unaware of him in their future.  Whenever I hold a find on-site, I feel so close to having this lantern in my hands.

So, since this game is an archaeological fantasy, we get to have that lantern.

So, really, why six weeks?

I don’t draw people.  (I really want to, but it has always seemed like such a distant skill to acquire.)  And I’m not great at animation.  These two things really hold me back.

However, Return of the Obra Dinn with its freeze-frame vignettes has given me some hope.





If I could simply pose some figures, maybe I could suggest a story, too.  I just need some figures.  But the details!  The game covers more than 10,000 years.  That’s a lot of fashion to recreate.  And that’s not the point of the game.  Perhaps, if the figures were stylized.  Just on the cusp of recognizable as human.  Everybody’s Gone to the Rapture did this with its energy ghosts.  

Recompile takes the energy ghost to a new level using an aesthetic and technique I like very, very much.  (Go look at their site.  Go on, I’ll wait.)  Perhaps if I borrowed, just a little bit?  




(Just look at it! … Look at it!!!)

Finally, some of my cutscenes (not this one) will have several stages of pose, like extremely low framerate animation.  How to make the transition between poses to help the player link the two together?  Fortunately, the way Aliya and Six move through the nebula of Heaven’s Vault one frame blurred to the next, suggest a way forward.  

https://www.youtube.com/watch?v=RUgtMlkwXRA

So, I have settled on Obra Dinn-style tableaux rendered a la Recompile using compute shaders (because I want to learn the dark art of shaders, too), and with Heaven’s Vault movement!

And the past six weeks have been the journey of me trying to figure all of this out.  Finding some simple human base meshes with as few recognizable details as possible (thanks Talon!: http://wiki.polycount.com/wiki/BaseMesh), rigging them (thanks Mixamo for the skeletons and the weights - thanks C4D for letting me blunder toward making them useable), converting them to voxel volumes (thanks MagicaVoxel!), and then getting them into Unity where I could apply my newfound basic compute shader skills (thanks, Bgolus).  

THAT’S what took so long.  When you have maybe 45 minutes four or five times a week and you’re learning several new skills at once, sometimes it takes a while.  At least, that’s the way it works for me.

So, here it is.  I could tweak the shader forever, but for now, for a demo, I’m satisfied.  I imagine that in a year I’ll be horrified, but by then I’ll have gotten better at all of these things and will be able to build on the foundation I’ve created.

I’m sorry I’ve been so silent for so long.  I’m chipping away at the project.  I’m energized to get back to the bug fixing in the building system.  There’s a lot to do, but almost all of the systems are now in-place and I’m back on my roadmap moving forward toward a demo.

Now, compared to all that, the audio system should be relatively simple.  It’ll just take me a couple of days (wink!) …

[nova++]

compute shaders

You don't know the power of the dark side

[JobLeonard]

However, Return of the Obra Dinn with its freeze-frame vignettes has given me some hope. If I could simply pose some figures, maybe I could suggest a story, too.

I get where you are coming from, but everyone focuses on the dithering graphics with Obra Dinn, but I think that in practice its narrative is actually more like 90% sound design, 10% visual presentation. It's more like a hyper-radio game.

The narrative intent of Obra Dinn is also almost the exact opposite of yours: it's a whodunnit where you are trying to unveil what really happened. The aim is to reconstruct one true narrative. The world of Obra Dinn is knowable with one objective external truth. Basically, Rashomon it ain't. And if I understand you correctly your intent is even more open than that murder mystery!

Which brings me back to the "hyper-radio" remark: I don't know if you intend to have fragments of things people say, but it sounds like something as precise as that would contradict the open narrative intent you have for the game.

Not saying this to discourage you - I think your intended use of vignettes will probably work the way you want them to, but it might be useful to the design process to realise that they do so in a very different way than what Obra Dinn does.

[amasinton]

You don't know the power of the dark side

I don't, I totally don't.  But I really, really want to!   

[amasinton]

I get where you are coming from, but everyone focuses on the dithering graphics with Obra Dinn, but I think that in practice its narrative is actually more like 90% sound design, 10% visual presentation. It's more like a hyper-radio game.

You know ... you're right!  I hadn't thought of it that way, but that sound design at the start of each vignette, that sets up so much of how you interpret what you see.  It's a radio play with a digital notebook to help you keep track of your own thoughts.

Speaking of digital notebooks and dithering, I've been kicking around an idea for one in The Resurrection.  I'm on the fence (dithering), though.  I want it to be something you build up without really noticing - so it doesn't overwhelm the narrative you're building up in your head as you play.  Metadata, I guess.  The notebook in Obra Dinn is such a key mechanic to the game but it is also heavily dictates the player's own interpretation.  The notebook is its own mini-game.  I will explicitly encourage the player to keep their own notes with pen and paper, too - because I like doing that sort of thing and I frequently hear that others like it, too.  Many games now are concerned with doing that for the player and it kind of distances the player from the material elements of the experience of playing.  But, I digress..

The narrative intent of Obra Dinn is also almost the exact opposite of yours: it's a whodunnit where you are trying to unveil what really happened. The aim is to reconstruct one true narrative. The world of Obra Dinn is knowable with one objective external truth. Basically, Rashomon it ain't. And if I understand you correctly your intent is even more open than that murder mystery!

(Awkward confession: I've never seen Rashomon. I will fix that.). You make another well-observed point.  In my game, something absolutely did happen on the island in the past.  But it's not knowable - not completely.  I think in this game I'm exploring the interplay of what you know for certain with what you cannot ever know.  So, these 'cutscenes' are like building foundations or artifacts: they are knowable parts of the past - but their relationships, their temporal order, and what they mean are up to you to build a plausible narrative around.  Hopefully these vignettes aren't going to be too heavy-handed.

Which brings me back to the "hyper-radio" remark: I don't know if you intend to have fragments of things people say, but it sounds like something as precise as that would contradict the open narrative intent you have for the game.

No, I am not planning on including fragments of what people say - no audiologs or diaries.  I think voice clips could be used carefully, like the vignettes, but there's so much more to control about them: gender, age, nationality, emotion of the speaker.  (Also, localization issues.). I think it would be best to avoid them.  Even written fragments would be too heavy, I think.

Not saying this to discourage you - I think your intended use of vignettes will probably work the way you want them to, but it might be useful to the design process to realise that they do so in a very different way than what Obra Dinn does.

Thank you for your thoughts.  They have been so helpful.  I have been thinking about them all week.  The vignettes look superficially like Obra Dinn, but, you're right, they are in the service of different goals.

[amasinton]

It’s snowing hard outside, it’s very cold, the morning light is dim, the local school is cancelled. Here at work the lights are all off, we aren’t open yet, the Library is lit only by the greywhite light from the windows, and I am the only person in the building. I have a few moments entirely to myself. It’s cozy.

I have this very rare moment to compose an update.

Still on the Map

Yes, I’m still on the roadmap toward a demo. Despite a sustained period of increased family and work commitments, development grinds steadily onward.  The basic form of the first site, the parish church, is compete and functioning in-game (you might have seen some posts on Twitter’s #screenshotsaturday).  I’m working on the finishing details of the church now – the roof structures and some basic furniture.

Here’s some shots of the progress:


The shell of the church.


Two forms of tracery for the east window.




The east end being assembled in-game.

I was getting kind-of worn-down on development because there didn’t seem to be much progress or time for progress.  However, a couple of weeks ago, when I had the basic building blocks done and setup in the game, I was able to see the form of the complete building for the first time and I caught a glimpse of how the game might look and feel for a first-time player.  And it was magic.  

So, with renewed encouragement, I’m back to work.

What’s Left for the Demo?

Details.  Lots of details.  I think there are always details that could go in.  But they don’t have to go in, so I’m trying very hard to take note of them and then set them aside for now.  

Audio.  All of it.  Again, I want to thank everyone who has offered to help on this.  I’m flattered and humbled (oxymoron?) by your interest and your talent.  I’m also very sorry that I don’t have the funding in-hand to be able to accept your generous offers.  I don’t want to ask anyone to work for promises on later return.  This is such a niche game, I have set my expectations very, very low in terms of sales.  I simply want to create an experience that will delight and entertain.  As such, I’m not sure there will be any income at all from this – and I think asking anyone to spend the most valuable thing they have – time – for free is just not right, in this case.  So, for now, the audio is in my very amateur hands.  Hopefully my placeholder audio will be just that – placeholder, and, later, if I can find some modest funding, we can use it to build something beautiful!

Time Travel. I think the wobbliest of my systems is time travel.  It’s so fragile that I’ve held off further work on it while the main structure of the church comes together.  But I’m going to have to get that system working reliably, so this is a point where work will slow down a bit, but it is essential to make it robust enough that I can move forward on creating content.

’Cutscenes’. Those odd ‘cutscenes’/vignettes need to be triggered in some way.  This will be a combination of triggering based on other events in-game, the player’s current physical location, and their current temporal location.  I need to work out the rules for this and build them.  This is a lot like the other rules working behind-the-scenes, so I think it is a matter of adapting work that is in-place and functioning, rather than forging something new.

Menus. Well, UI generally.  A year ago I created a working menu system and starting screen.  But I need to get that better integrated into the game now.  And I need to add a way to have multiple save files that the play can access.  (Multiple save files will also help development because I can load different points or branches within the game at-will.)  Again, I’ve got the basics of this already working – but there is some user-facing work that needs to be done.

So, it’s all still progressing.  One of my major time commitments just came to an end, so, for a couple of months, I won’t be dealing with that.  I might be able to convert some of that space to dev time, which would make me very happy.

And, of course, I want to get back into more regular posting here because this, too, is an important place of encouragement, learning, and motivation.

Thanks for reading.  I’ve got to go turn on the lights and shovel some snow!