My first experience in video game / game jam / demo release

(ok, so, people can complain about Smile Game Builder, still, when your writing a whole feedback on itch.io about your gamemaking experiement and save it as unpublished and itch.io answers a "body: expected text between 1 and 204800 characters" then crashes and you lose everything, it hurts, like, a lot, and everything is lost like tears in the rain and nobody complains)

https://media.tenor.co/videos/a93351265e708b9085be9b331ea3e8fc/mp4

So, here we are.

AGAIN.

In the end this feedback might be angrier  & shorter than expected because the first version was totally lost thanks to itch.io.
Still I'll do my best to share with everyone.

So, let's do it again, and shorter this time.

On saturday june 19th I've release my first public playable demo version of a videogame, that you can find here.
It's called "Manteca - Beyond Mississippi".
It was made mostly during the Amalgam Ash 2021 Summer SGB Jam.
And it's from the universe I've developped earlier on the pen & paper RPG univers of Mississippi - Blues, Prohibition, Vaudou.

1/ Why am I making this devlog note ?

Because it was the first time I went into gamemaking.
In fact, not really the first time, but as you'll see earlier, it was the second time, but the first time as "serious business" and a bit more of experience.
During the development of my game, I took daily video & pictures to check major milestones, and now that my demo is released, I wanted to make some retrospective about my work, my challenge, my successes & my fails.
If that can help anybody, then it's a win.
At worse, it'll be some old story I'll talk my son about in a few years.

2/ Who am I ?

I'm Christophe, 35 ans, a fresh dad (my son is so far eight months old), from "Franche-Comté", France.
I work as a pro Senior QA / tester / sometime project manager for SaaS web app providers, so I have good experience with software management / scheduling / development.
I'm a gamer since my 4th birthday. At this time, my parents were not really really rich, so when they bought me my first NES, they told me they wanted to see the end credit screen before buying a new game. Sadly my first console was NES and my first game was TMNT, and it took me a year to finish. Then, after, Mario Bros, 3 days. Then, after, Zelda II, another year. Sadly for them, after that, it was waaaaaaaaay easier to finish like every other games on Megadrive / Genesis or Super Nes after that hardcore training.
First time I went for game dev was when I was 15 years old, like 20 years ago, on RPG Maker Vanilla. Obviously, I was young, I wanted to make my FF7 Sequel, I didn't have any scheduling skill, so I didn't do a concrete thing in the end. But after that, I went on writing & game design (I now write novels, pen & papers RPG, escape book, tabletop game) & tech jobs (first as a software support then as a QA / QA Analyst / tester / transversal project manager for big french e-commerce platforms & SaaS providers).
So, just to be clear :
- I know how to code (but I don't know & don't feel at ease to code with every langage, being most comfortable with JS, Java, PHP, CSS, html & Python), and it's not something I really like in a daily routine
- I know how to manage a software developement  lifecycle
- as a QA I'm very, veeeeeeery critical, even with myself

3/ How I went to game making (again)

20 years ago, the idea of coding something, taking the gamepad, push the "right" button and seeing the character going right was mesmerizing to me, like "wow, I'm a god".
Since, I did a lot, in both gamedesign / writing & tech cursus.
Last year, after I bought some heroforge models for a tabletop game prototype, during the first pandemy lockdown in France, I wanted to import my .stl models in Blender to decimate them in order to import then again in Tabletopia. Then, I was like "oof, I really want to dig in deeper, potential seems infinite", and I forget that.
In march 2021, thanks being drunk at night, I bought some kind of 3D Lego craft game called Trailmakers on Steam. And, again, it was mesmerizing. Doing my craft in 3D with basic bricks and the engine processing all of that in real-time made my brain explodes.
At the same time, with my partner in publishing (we own a book publishing company between France & Swizterland, PVH Éditions), we wanted to dig into Visual Novel lands, so as my friend was adapting one of our novels with a dedicated team, I wanted to benchmark some game engine.

4/ Why Smile Game Builder ?

Thanks to my (old) previous experience with RPG Maker, I knew a bit about stuff you don't expect when you know nothing about gameking (map creation, teleporters, events, etc), and way more thanks to my job. So, I have a precise idea of what I wanted to do, and what I wanted to find in my game engine.

I benched engines like Unreal Engine, Unity, Godot, RPG Maker MZ, and Smile Game Builder.
In the end, Smile Game Builder took my attention for this project.
But, to understand why, let's talk about my phisolophy earned from my civil job experience when you benchmark a tool (for me, mostly automation test framework ) :
- what are your need ?
- what can the tool do ?
- for needs that aren't covered, what is the investment needed ? (time, money, plugins, etc)

That's why, once that done, I went for Smile Game Builder.

5/ What were my needs ?

As I was going for a Visual Novel engine at first, obviously, I went for :
- dynamic dialog (still, SGB has some lacks for complex dynamic paths)
- display of custom HD characters
- pictures displaying & music playing

But after I saw what Smile Game Builder could do, I went for more gameplay desires.

To be able to easilly create a map with buildings & NPC & playable character was a big plus. I didn't wanted any kind of battle in my game, still, this made me think a lot. So, from a VN project, I went to some kind of time-loop narrative exploration game, something inspired by Hades (for the die & retry roguelite style gameplay -I played something like 100+ hours in January, personal highscores are heat 32 & no heat run in 14:35), Disco Elysium (for the narrative / exploration part), PT - Silent Hill 5 demo & Majora's Mask (for the caught-in-a-timeloop part), Little Big Adventure (for the overall gameplay outside battles) & some others.

So, I've made my mind : Manteca - Beyond Mississippi would be a spin-off of my pen & paper RPG, and my protagonist would be caught in a timeloop in 1940s New Orleans, and will have to explore, talk & think in order to get off this loop.

But, in fact, why going to videogame ?

To be honest, since the pandemy and first lockdown in France, my writing motivation went quite to zero. Because no public event, because moral issues, because wife's pregnancy (then childbirth then time to spend with child), but also because when you works on a book, you have to work & wait with a drawing artist, a printer, a published, etc, etc, and it can go from months to years between the time you finish your writing & the time someone can read your book & play your game. Indie game seemed, on the moment, way fasted to me. But, in fact, a lot of challenges were present.

6/ Step by step (ouh ouh ouh)

(yeah, sorry, the more I have to type this devlog entry, the more I'm tried, and the more I'm tired, the more my musical taste & cultural references go down)

So, in order to know if Smile Game Builder was my thing, I had some things to check for my kind of game :
- usual things (dynamic dialogs -even if I have some thinks to blame SGB about-, map teleporter, event switchs, global events triggered on variable, etc)
- some countdown decreased on each map change
- events when the map countdown goes to 0 (with pseudo game over cutscene)
- easy 3D asset import
-  etc

So, I went with this first technical demo / benchmark, with basic & some imported CC assets  from Sketchfab. It can seem dumb at first, but I was really proud to make that, and it answered most of my needs. Plus, I managed to bring into the game engine a mechanic from the pen & paper RPG (when you have 50% chance on some actions but by betting soul points to the devil you can have 100% chance).

And so, after that, the technical benchmak was validated.

I had to go further on two huge topics : characters & building assets.

7/ Make it nice, make it beautiful

As I am quite stubborn, I always want the stuff I write about Louisiana & Mississippi in the 1850s-1950s to be historically accurate. And I wanted to be the same for the buildings.

For the technical demo, I did import some CC assets I downloaded from Sketchtfab, but for the rest, I wanted buildings which really look like New Orleans French Quarter.
And I found some !

But... they were "New Orleans inspired". Not historically accurate.
So, as stubborn as I was... I wanted to make my own stuff.

For that, I went with AssetForge.
It was really interesting and I think I was able to have quite decent results, but :
- as it was it gave me the desire to remake all of the buildings of New Orleans French Quarter, which would have take me, like.... 6, 8 months at least ?
- at this time, I was quite noob on Blender but I already identifed some issues with AssetForge for SGB, like a bad management of materials, some textures issues an lack of faces optimizations with the designed-by-brick way of building things

So, in the end, after this, I finally decided to from AssetForge for now.

...still I was really happy to walk a bit in a pseudo virtual New Orleans.

And I tried some stuff to easily use same buildings with differents textures if I wanted to separate game phases without doing everything again, like an alternate post-apo timeline, a Sin City graphic stage or whatever.

But in the end I had to little time to make this myself so I went for the Daz 3D New Orleans style building.

Of course objects where damn to HD so I had to eyebleed while using Decimate on Blender.

Result was OK, still not perfect, and some buildings were truly to complex to be imported under the 5000 faces limit of Smile Game Builder, but... at least I had something (preview here with building outside the 5000 faces limit of SGB, later models were more fluid but less pretty).

https://www.youtube.com/watch?v=5c9lepe1NB8

After that went the question of characters.

I did bench some character maker tools, but most of time, I've felt like I was playing with glorified virtual dolls, and most missed custom options I wanted. As I wanted to go for deep, unique characters, I wanted to have something more than that, and that's why I wasn't happy with stuff like Human Pack v2.60 for Smile Game Builder nor Smile Game Builder 3D Character Editor (I wanted to go outside the "bighead" 3D Super-SD Character stereotype).

So, I went for Autodesk Character Generator. This one was quite funny because, without trial version, I had to send a request to be called by their commercial service, so they called me back like "do you want 5, 50, 500 licence ?" "Eeeeeeer... one in fact, but as I'm not sture, is refund possible ?". Then I went for it, something like 80$ a year. My fears were about miss of assets, but this one seemed highly compatible with Mixamo for animations, so, I tried...

At first, I was like "wouhouhouhouhou", but 2 things happened :
- 3d import to Mixamo / Blender / Smile Game Builder was a bit messy
- I found that, despite the annual licence, I had to buy additional cloud extra credits to export my character (if I remember well, it was like something like 5$ a character). I had at least 25 for my demo, so, *gulp*

At the same time, Character Creator 3 from Reallusion caught my attention too, and created character were perfects for me :

Still, I had some fears :
- there was like tons of assets, but mostly paying ones, and know that a virtual haircut can go more expansive than a real one, I feared that going to Reallusion marketplace was like opening a Pandora's box
- characters were damn too HD and felt poorly under Blender Decimate tool

Still, for characters portraits during dialogs, it was perfect, and quickly, I chosed this solution at least for that (plus, it allowed me to mix the same character in 2 variations for the pre-post timeline switch on the game). I was really pleaed by the PS3 vibe from these ones.

Yeah, sorry AmalgamAsh & Shar for the spoil, but she won't have a nice life in the new timeline...

8/ Give them face & movements

So, the idea was to use Autodesk Character Generator for 3D models on map & Reallusion Character Creator 3 for high-res portraits. I was afraid of having twice the work on character definition on both tools but... Still, I tried. Of course, screenshoting from CC3 and using remove.bg was quite simple. And really soon, I had all of these buddies. Biggest challenge was to not just go & instabuy everything on asset marketplace, define a max budget, choose useful assets for my game, optimize the selection then going for it.

For inspiration & exercice material, I asked some buddies in order to convert them to 3D, with more or less happy results :

Then I did the same with other inspired character as references.

For Autodesk Character Generator to Mixamo to Blender to Smile Game Builder, it was... something else... My goal here was to be able to create a character similar to another one created on CC3, export it in 3D via Mixamo, then import it on Smile Game Builder via Blender to display it on the map.

Ealier, I did have some "results" with Mixamo to SGB attempts :

Character shape & animation were here, texture was missing.

Then after some texture skill ramp-up & mumbo-jumbo on Blender, some went well, some weren't. And, still, I had my issues with some animations (which if fixed since, and don't be like me, read that AND that).

Mixamo version was quite decent, I now had to do the same with Autodesk Character Generator. And it... didn't go well.

At the same time, I tried some ne stuff, like having a car item for your character to move on the map (sadly I wasn't able to manage the landing case of the event once going out of the vehicle).

Some scaling trial & error for characters vs buildings vs Smile Game Builder grid :

Some improvement of movement fluidity in NPC speed & frequency :

And same for the main character to walk outside the grid and make her walk smoother.

Still, I didn't have the feeling to go anywhere with Autodesk Character Generator, and assets were limited, so I refunded if on 29th day (the customer service was top notch 👌) and I went back to CC3 in order to try to export HD characters from CC3 to Mixamo then Blender then SGB. Then, after 18 attempts...

IT WORKED 🎉

All of this thanks to CC3 InstaLOD module and a bit of black magic.

Sadly, sooner of later, some characters had an issue with the "all black except hairs & facial hairs", as you can see here :

So, right now, I still have the issue, but for the demo, I decided to drop the animation for these ones and export / import them as simple .obj then .fbx to SGB, which kinda worked for a MVP :

After that, some more improvement, new features (like "Souvenirs", a Hades codex like using the skill mechanics from SGB)

And finally, a middle-step demo with Daz3D SW French Quarter Building and CC3 x Mixamo characters.

(yeah, it deeply feels like a printed solution / walkthrough on a vintage Console Mag from 1990s)

It was nearly at this time that Seliana Filby earned her name & expressions.

9/ All aboard the GameJam Train

As long as I was able to have building, dialog & map representation of my characters, and as long I as was able to do it by myself, most stuff was done. At this point, I was more or less considering my "game framework" was ok and allowed me to work, except for one this. Which one ? A basic one, the really really really small one on a videogame in the 1940s in New Orleans : MUSIC.

I did schedule 2 whole days of work on it, but the answer went way more quickly than expected, thanks to Incompetech. Hardboiled crime noire movie, blues, jazz, swing, funk, fantastic & mystery soundtrack it was perfect to me. So, 38$ later, I was happy to support him by buying the whole collection. I just had, now, to search soundtrack for each scene, place & eventually characters.

(just a small part of assets used, mostly from Sketchfab)

("How to train your drag... map limitations on a demo")

Then, being French, but contesting in an american game jam, I had to release my demo in english, so I worked on a multi-langage management for the game :

(you cannot hear me over the sound of my non-captured debug console with my "langage" variable)

And, obviously, as I was short of time, I've had something to do about petting the dog, because I have values & convictions.

(the debug dog on his debug dog house...)

(... and here's what happen when you pet a dog 666 or more time - even the Hellhounds are rooting for you)

Just because it was fun and there was time to waste (in fact not but still did it anyway) I added another funny feature where your character can be painted on a park because if felt cool :

And I added arts from my last published book (with the same lore as the videogame) :

Then, I did my scenes, cutscenes, endscenes, compiled it and sent it for the game jam after a whole from friday 8pm to saturday 10:30am session for final tests.

Here was the result :

10/ What's new since ?

As explained earlier,  some improvement were already identified.

Some are already done :

By importing complex assets as "accessory" on CC3 with a transparent / non-displayed character, I can now use CC3 InstaLOD to export them as FBX with 5000 max faces (or less if required) and a monotexture in .png. That will allow me to have smoother buildings on SGB, and some assets I won't be able to decimate properly to this kind of lowpoly definition with just Blender decimate tool.

Plus, even if it's not perfected yet, Selenia can run, and even crawl under a table / small tunnel ! She's no Lara Croft yet, but she's improving.

Sadly, for some other actions, it's a bit more complex, and I'm still working on it.

And I even went a bit on Unity for a comeback benchmark, but so far, it seems overkill in term of required coding work for remake vs new possibilities vs global rendering.

And she has new place to visit right now (that and the doll house) :

https://www.youtube.com/watch?v=RdTwmN1s_X0&feature=youtu.be

11/ And now ?

Right now I'm experiencing some healt issues with my arms & hands, and I cannot type for more than 2, best 3 hours a day. I'm thinking about what to do next, and now that my intro is something that can be called complete / playable, I'll need to thing about the core game, the die-&-retry exploration game. I already have a lot of WIP maps, still it'll be a huge game design / level design challenge, some puzzles / quests to do, a lot of test, etc. But right now, as I am, I cannot do that, so I'm leaving it for later. I'll probably go for the french translation first, for the release of a public en + fr fixed demo with improved graphics & less bugs.

We'll see soon ! 'til thanks, thanks for your reading :)