Primordial Soup Society Week 2

Another week on PSS (it’s been 2 weeks as a whole) and Paul aka Bowl the Expert did a crash test.

Gameplay :

– Bowl is more experienced than me with card games, so he pointed out all the problems. He oriented me with a rock-paper-scissor gameplay, that I adapted with my favorite stats : karma, sex-appeal and sanity. It looks a lot more promising. In the end I think I got the right system. Now it will be about creating interesting cards, with balancing and shit …
He also said that it’s cool to create cards, but remember that collecting cards is the core of a game like this. So we thought about a lootbox system to get ingredients to craft your cards. So it’s collecting THEN crafting.
Then we also discussed about not having too much numbers on the screen, and since cards are bringing a bunch of them (3 x attack, 3 x defense) I agree to drop the 3 base stats numbers of the player, and this way it will be all about the cards.

Server : 
– finished a complete basic fight VS I.A. . Very basic … As Bowl mentionned, there’s only simple blaster cards for now. But it works super fluently and I’m still amazed by the speed of nodejs !
– also did a session system so if you disconnect, you reconnect in the right place of your game at any time.

Client : 
– enhanced the UI. Now things aim to look like cards : places, people / creatures, and of course cards.
– enhanced editors. But still a lot to do.
– started something like a tutorial.

Questions : 
– I’m doing the game in french, because it’s a literature game. But I know the scene is all about english. So I really should prepare it for a english translation. Like a english version of the game, synchronised over git and stuff … Yeah. And I have to look about how translations are done with Symfony.

What’s next : 

– PvP
– Card manager, Card Editor, starting with card attributes & lootboxes. I agree with Bowl saying I should make a solid card game before going further.
– Creature manager
– Map as dungeon with guardians blocking the way. Walls.
– Realms (a player can create his own map).
– Make design great again

Estimation ? 8 weeks. Since I get around 1 week every 2 months (I got kids and I’m doing music besides), I should expect a beta in 16 months. That’s okay !

Then I will have a game. And it will be community management, balancing and shit.


Primordial Soup Society

Almost 1 year after my node.js serveur tests with the Cehiho prototype, I’m getting back to it with a new game. I’m not calling it prototype since I aim to create a real game in online production.


It’s a roleplay-game with a card system as fighting, looking like a visual novel. Here’s the twist : it’s a dynamic content game where the players are creating everything : monsters, places, cards … Including visuals !

Technically, it’s a web application, based on Symfony 3 for the front and DB and a node server since it’s multiplayer. Canvas for display but also as editor for user-created content.

So far, I got the server running, user registration, and content creation is work in progress. Can’t tell about a beta release yet though.





New “prototype” working on. Inspired by my games “Blektre” and “Ours en Slip Simulator”, but also a 10 years old MMORPG UO-inspired concept called “CIO”, and also game like “Space Station 13”.

Features so far :

  • nodejs realtime multiplayer server
  • persistant world (mysql based), randomly generated map
  • construct 2 client
  • chat with players and create persistant items

Cehiho is a MORPG (does not pretend to be “massive”)

Gameplay : WIP


In the darkness : SELBST RPG

selbstss2Selbst RPG is a prototype I’ve been working on lately and now I’m looking for testers and feedback ! I don’t know how far I will go with this yet but for once, I feel good with the prototype.

Presentation :

Selbst RPG is a top down shooter / ARPG where you have to build your weapons with a modular grid system. It has been inspired by modular music softwares I’ve been using in my musician life.

Special mechanics are easy to explain : in the “engine” grid, you have generators, spawning “energy particles” that will power your “weapon” modules such as : front fire, side fire, bender (allowing you to bend the trajectory of your fire), rocket launcher (rockets can be bend too), shield, mines, teleportation, speed up bonus, health regeneration and so on … you can build everything you like, like focus on one specific firepower or make something balanced with lots of defense and various weapon. You also have to manage the controls of it with Hotkeys (modules connected to your mouse/keyboard, releasing a signal to trigger the modules that requires it) and “effects” modules such as tank (accumulating energy to release bigger particles), angles, speed up energy particles etc…

So as you play, you build an engine that will allow you to defeat ennemies but also resolve puzzles, you can save your builds to “disks” allowing you to reload any build at any time (because you will need to make special builds for special situations) and you will find new modules as loot on your path.


As today, mechanics are the most advanced thing of the game. As you will notice, graphic design is quite minimalistic (320×200 resolution pixalated squares style is easier to deal with for me 😉 ) , background story is an abstract joke, and I’ve made efforts with the level design but I still need to work on it to get the prototype I’m thinking about. For now it just asks the player to use various techniques but it still needs “big” situations and extreme cases where you really need to get the maximum of your engine with various ennemies at the same time, and I would also like to get a more extreme feeling with tons of sprites on the screen like in Bangai-o game, for those who know.


Dev thoughts:

My personnal questions so far are mostly : will Construct 2 (or should I say : javascript) allow me to make that big “sprite explosion” game ? I already dealt with a lot of performances issues (because I require a rock solid 60fps on an average gamer machine) and I still have plans to improve it. For example I noticed that ennemies pathfinding itself was using 20% CPU, so I know my next move will be to make a mix of custom movement and using pathfinding only when needed (10% of the time). I’m also using a tiny resolution so I don’t have to care about sprite animations size. A grid / tilemap system is also helping a lot. Then so far things are not doing bad, maybe I’ll be able to do this.

One other question is about the mashup between arpg and shooter. Is it a shooter or something else ? I’ve been thinking about making it simpler as a shoot’em’up (one path, automatic scrolling and stuff) in order to get more speed and action and less thinking beside the engine building. But I also like a bit of creative freedom so I’m thinking about levels and small challenges.

So far the demo is offering a tutorial to the basics and  small dungeon made of 4 levels with a end boss.


I need you !

You can try it the actual demo here :

Every feedback will be greatly appreciated 🙂



Presentation Video (contains a bit of spoilers so don’t watch it too carefully if you’re willing to play the demo)





Bel-Ocean is a point and click adventure game. You play Jean-Chien, a disenchanted dog who has been cursed by a giant fly-fairy that makes him living the same day over and over again.

Jean Chien lives in a seaside city over polluted. The sea became became a giant pool of poop and the sun is not helping the stink by striking big time. People are trying really hard to keep their dignity though, but that’s not going to last very long.

On top of that, there’s also his best friend, Jean-Dauphin, killing himself every night at 3AM.

The fly-fairy has been clear on one point. To break the curse, Jean Chien has to clean the ocean during this one very day. This is up to you to take advantage of that curse to change the course of other things, like saving your best friend.


Bel-Ocean is a “classic” point&click game. The gameplay twist lies in the looping day and the need to organize your actions in order to achieve all your goals in one day. For this, the player has special tools such as : a magic map able to track the NPCs, a magic watch to travel through the hours of the day, and a magic notepad to organize your roadmap.

So you get everything you need to do the fastest run you can !


Developped by Fromage Interactif, basically composed of two brothers : Charles Torris, writer,  developper and music composer, and Paul Torris, designer, illustrator and animator, both passionate with videogames. Construct 2 has been chosen for efficiency and convenience, combined with the powerful Spine software for the animation. Our own dialog management system in PHP / MySQL has been made from scratch, allowing us to deal with more than 4500 lines of dialog.


We started this project 3 years ago, and now the gameplay part is finally almost done, as a complete playable alpha version already runs and rocks. We need an extra 4 month to complete the graphic assets. A devlog is available here.

However, one main question remains : shall we do some voice over for the characters?

We would need 3000€ to make it the good way, and we don’t have the money. So we’re thinking about crowdfunding. Why not trying ? If it fails, we’ll release the game anyway, because it already works like that. And maybe that’s for the best ?

Stay tuned !


Bel Ocean devlog #14 – Alpha coming near





Bel Ocean devlog #13 – Playthrough, finally !


The game engine is now like complete and well-tested, and the playthrough has been massively written these last months. It means some full alpha version is about to be done for internal use, so Paul will test it for the first time very soon. It means tons and tons of debug action but that’s the funniest part isn’t it ?

And I also ordered a whip to get him working faster on the remaining assets.

We’re also starting to investigate about the voice over matter. This is the only part we can’t deal with by ourselves if we want quality english-speaking actors, and that means money. And money we, of course, don’t totally have yet. That’s why we’re thinking about doing some crowdfunding just to pay some nice voice over. Or would some editor interested ? We don’t really know how to sell our game, but for now we’re more likely to do it by ourselves : doing a good game is the first priority.

To get the voice over quotation, we had to do some nice statistics; we got 3000 lines of dialog and counting. Since there’s still a lot of secondary characters to design, we can expect 3500~4000 lines the end.

Then, here we are. We can start to see some schedule : a beta playable version should be done for the end of the year. We may need some beta-testers at this point, so, if you’re interested,get in touch !

Bel-Ocean Devlog #12 – Writing !

We did some nice updates before my daughters birth, and now it’s time to get back to work ! In order to celebrate, my brother made a nice joke, while leaving some pasted items in the last background update, so I spent like 2 hours to find where this “bug” was coming from. How funny !

Beside that, I think we’re getting close to a complete game engine. You can follow all the technical details in my technical devlog here .


Full Time Job …


Bel-Ocean Devlog #11 – One Year Later

Despite the sporadically blog entries, we did work this year !

The first good news is that we found a better main character than the horse : let me introduce Jean Chien.

The best thing about him is that he tolerates the shitty smell of the studio. Problem solved !


New episodes about Paul’s Wrath Against Cutting-Edge Softwares ! 


Paul has been working on Spriter for a few weeks more before he went berzerk about the shitty unprofessionnal quality of the unfinished software. He started to harass Spriter’s developpers to fix some basic features but they just ignored him and keep on releasing “art packs” instead. The Paul went completely mad ! (he’s a passionnate kind of person). Eventually he started to break everything in the studio.

It went to a point Paul turned is rage against me and, for my defense, I suggested to roll back to traditionnal animation with a FPS cap … That was very sad option but we had to move on.

HOPEFULLY we discovered Spine. Spine is a new software that does just exactly the same as Spriter, but with all the professionnal features that Paul needed !

ONE problem remained : no official plugin from Spine to Construct 2 was released. After a few days, we did find one on Github made by Flyover (nobody knows who is he) but the plugin was incomplete : no mirror function, scrollTo didn’t work, no interpolation disabling … Problems .. ALWAYS !

(George’s Character is totally inspired by Paul’s deep nature)

Then, since I know Javascript, we tried to fix the plugin ourselves .. Paul also bribed Flyover to push him to fix some of the most complicated features, and it worked ! I also managed to disable interpolation and other stuff, creating a fork of the Spine plugin for Construct 2. At this point, we knew the software was the good one. Paul started to re-make all the characters in the new software .. For like, the third or fourth time .. and I had to do the same in Construct 2 in order to use the new plugin.

One month later we got back to our Spriter developpement state. But this was for the best, for now we’re absolutely sure everything will rock .. almost …


Because we found out that Spine wasn’t supporting animation duplication in order to create the skins we needed (mostly to create the dynamical illusion of day and night lights on characters). Paul went berzerk again (he threatened to kill my hamster this time) but hopefully quickly found a solution to accomplish this skin duplication with hacking the export files with notepad. I eventually took a few hours to create a patch software to make this automatically ! Damn, another problem solved !


What else ?

  • A new NPC “engine” of our game, allowing to use simple functions to create their agenda in the world.
  • The Dialog Manager has been upgraded.


Approximative statistics :

– YEARS OF DEV : It’s been like 4 year of irregular developpement (beside our day jobs) since I wrote the plot of this game. 7-8 month of full time work estimated.
– POINT&CLICK ENGINE (made w/ Construct 2) : 80% complete
– GRAPHIC ASSETS : 69% complete
– STORY TELLING IN GAME : 7% (fully written on the paper)
– DIALOG LINES & TRANSLATION SOFTWARE : Written in PHP, 99% complete . I managed to fully connect it with C2.
– SPINE PLUGIN : 100% complete.

We hope to get a fully playable alpha around september 2016.

See you soon !



Newer posts »
« Older posts