Flare: The Game (Preproduction)

[clintbellanger]

We're at the point where Flare has at least the first version of MOST major features required to make a game. Much of the remaining work we have planned for the engine is refactoring for modders.

This means it's time to take a serious look at Flare: The Game. Taking the current assets, putting a bit of polish on them, and turning out a satisfying full game with it.

Engine Features

First thing to discuss: what missing engine features are absolutely necessary to making a complete game?

The main thing that comes to mind is a "Cinematic" game state to display cut scenes (most likely slides of images + text + music). But that's a feature that can be implemented really late into the game-making process.

Second: what current engine features absolutely must change to make an interesting game? The Powers menu is the most likely. On paper it seemed fun to unlock powers that way, but in practice it really isn't satisfying. Thus our planned Powers menu rewrite.

Data Features

What near-essential data needs to be created to make a good game? Maximum results for minimal effort (aka we're not creating 10 new tile sets).

Here are my thoughts:

I want to flesh out the current 3 tile sets. Each of them will get new tiles to add interesting options for map designs.

I want to create Teleport Circle art and add it to each tile set. Then I want to create a test "hub map" where you can warp to teleport circles you previously discovered. I'm pretty sure all this is completely doable with the current campaign status system.

I want to at least add 2 new shields and 2 new armors, so that at each tier there is a shield + armor to wear.

Features like More Equipment Slots, 2H Weapons, Dual Wielding might simply not make the cut. We might save those for Flare 2 aka Wandercall.

I want to create a few completely new Test/Dev maps FULL of interesting events.

• One map will be all about Traps (poison darts, floor spikes, blade pendulums, floor sawblades).
• Another map will be all about the mapmod feature, showing neat ways to alter the terrain through events.
• Another map will be for Creature Features: showing patrolling, summoning, and other wacky combos.
• The goal of these test maps will be to push the current features in interesting directions, and create a suite of tools/ideas so that all new maps are Very Interesting.

We'll also play with the numbers. Example: I want Defense to matter more, so I'm scaling up armor absorption. To do this I will need to scale all the numbers up (e.g. starting health might be 100 instead of 10).

Finally we'll create a game data "style guide". Things to keep in mind for map size and layout. How often to place teleport circles. How to place bosses and chests so they aren't cheaply farmable. etc.

This preproduction phase will happen in parallel to v0.17 development. In the meantime we'll kick around basic story/plot ideas.

Once we have decent progress, we'll start making final game maps in earnest. I expect that ALL of the previous maps will be thrown out. However, with the fun tools/templates from the dev maps and the data style guide, it should be straight-forward and fun.

Scope

If we end up making a solid 10 hour single player campaign, it would be a huge success.

[pennomi]

Probably for engine features, we'd want to finally refactor the buff/debuff system, so the game designer can create different bleed types, shield effects, etc. by defining them in the power config instead of having some power ids hardcoded in as effects.

Another engine feature: re-enable various power triggers, like on_death, on_half_health, on_hit, etc. I think it could make the boss fights much more interesting if, for instance, the boss healed himself the first time he hit half health.

But other than that, I think the engine is in a good state.

As far as data features are concerned, in addition to your dev maps, I suggest we use a dev mod. I think hero powers should be redone creatively in the dev mod, so we can see what kind of options and creative combos can be made there.

[clintbellanger]

Good points. The power triggers need to be rethought; e.g. many of them make more sense if we implement Instant powers (that don't put the entity in Cast animation state). That can be part of the planned v0.18, though we'll keep it in mind for boss/creature ideas.

We can create a dev/tutorial mod soon. We can drop the wipcore mod (not really necessary in the Flare repo) while we're at it.

[igorko]

About "Cinematic" game state. Maybe it would be better to do it using some easy script language? Should not be too hard to implement. Like:

start_sound=sample.ogg move=hero, 12,30 talk=hero, Hello knight,5 talk=someNPC, Hello Hero,5 stop_sound=sample.ogg move=some_creature, 12,30 execute_power=hero, 23

where : 12,30 is position 5 is time of text showing 23 is power id ...

Those events can be easily executed by current game. If not, it's not hard to add. Clint, correct me if I'm wrong.

[clintbellanger]

It's possible to do such in-game events but I'd rather not do that for the 1.0 engine.

My reasons: I don't like in-game cinematics that take over control, especially in action games. I don't really think cutscenes like that belong in games period.

If we really wanted support for that kind of event handling it'd be way easier if we used a full scripting language like Lua, I just think it's completely overkill for Action RPGs where story takes a distant back-seat to gameplay.

If I'm in the vast minority here, I'll roll with it and we'll pursue these kinds of scene scripting.

The cinematics I'm talking about would be for e.g. starting the game and ending the game.

[clintbellanger]

Ooh, the dev maps/mod gives me a good reason to reintroduce myself as an NPC :)

[stefanbeller]

I agree with Clint, that there should only a few scenes such as intro and extro scenes, the game in between would be hack and slay (killing many mobs in a short time... I really like the zombies in the aveguard temple... I hope the player will get ways to kill multiple mobs at the same time such as powerful area damage magic or such, reminds me of D2)

Clint, in the opinion of my girlfriend the player is way too small, i.e. a moving pixel line on screen, maybe we should think about rescaling a lot?

For the story line, maybe the evil forces are starting to attack the kingdom, and the hero is a prisoner sentenced to long jail, who has the choice to either take the punishment or to be sent against the raiding orcs. As it turns out a path can be found around the orcs , so the hero can train some levels on the already existing goblins and grasslands.

Later on the orcs would come in handy for the story line, so the hero will be sent to a town which is currently being raided by the evil. I am not sure about the orcs, but I think of really strong enemies which can be used to make up such a decision for the hero in the beginning. Later on these might even be fightable.

[igorko]

I think for the Game we need higher resolution art than we have now. Current art doesn't look very nice in context of the full game.

[clintbellanger]

The first Flare game won't get higher resolution art. Most of the current art would look awful at double resolution. So we'd need to create all new art. We will not discard the last 3 years of art efforts.

If the hero seems too small, try playing fullscreen at a lower resolution. This isn't a 3D game or vector game where higher resolution really matters.

Maybe we can make a mod that doubles the resolution of all the art (in a lazy way; every pixel becomes 2x2 pixels). It would require changing the base tile size to 128x64, updating all the tile set defs, all the animations, all the menu configs. Note that this would require 4x the video memory; computers with 256MB or less video ram would probably not handle it. I'm not sure this would be worth the effort.

The planned new art for Wandercall will be a bit better -- the hero gets about 1.5x larger, and would look good at double that size. But these new pieces take about 5x more effort to create than similar old art. So it won't be ready any time soon.

[igorko]

Correct me if I'm wrong. Most of current art is exported from Blender, and resolution is set during export. So we already have(I hope so) high resolution art for tiles and sprites, it just need to be reexported.

[clintbellanger]

Technically yes. But most of that old art has no textures. Which is fine at tiny resolutions, but becomes obvious and ugly at larger resolutions.

[clintbellanger]

Progress:

Lots of engine configurations have been added to the engine. We're getting closer to the feature-completeness that will let us make this game.

Teleport circles have been added, and on_load map events to help support them. I like this method -- anytime a game can use gameplay to represent something instead of a menu it's fun to me. But we could add a boring list-style teleport later if we want.

[dorkster]

Something I noticed when playing a Wizard build. Warriors look good in leather/steel armor, and Rangers look good in clothes/leather armor. Wizards don't really have anything that could fit their archetype.

Wizards need something like a robe, so with my limited Blender knowledge, I made some robes based on Clint's clothes model. I've created a branch for them here.

Stat-wise, I'd imagine them having stats similar to clothes, but with additional elemental resists. I think it would be a fair trade-off for not having the absorb bonus of heavier armors.

Here's a preview of how they look in-game:

Would there be a place for these in the final game?

[clintbellanger]

I probably will create something like robes. I'll take a look at the ones you've done to see how they'll hold up.

[igorko]

Maybe we could recruit some authors/writers/artists on OGA forum?

[dorkster]

I think we should have some form of storing/transferring gold with the stash. The way I've thought of would be to add gold items (gold bricks, gold bullions, etc) with a 1:1 buy/sell price. Make it so that they aren't part of random item drops/vendor inventories. Then make "banker" vendors that only have these gold items.

[dorkster]

Should slingshots/bows really be in the 'Off Hand' slot? Yes, it prevents equipping a bow and a shield at the same time (which makes little sense). But it also allows having a bow equipped with a sword/staff (which also makes little sense). It makes more sense to me to have all weapons go into the 'Main Hand' slot.

To handle bows and shields being simultaneously equipped, maybe add a slot option to MenuInventory called disables. Allow it to contain a list of slots to check for items, and prevent equipping things to those slots.

[clintbellanger]

One benefit to having the bow in the off-hand slot is that's how it's drawn on the hero. Carrying a bow in the left hand makes sense for right-handed people because that's how a bow is fired. If we put bows in the main hand, we'll probably still want to draw them in the left hand. That would require more complicated avatar sprite layer logic, as slots would no longer have a fixed draw order (as "main hand" is no longer "right hand").

I'm actually liking the current system. Opposing attributes represent the same slot, so there's a meaningful choice when equipping items. I have to decide if I want a physical or magical main hand item. I have to choose between an offense bow or a defense shield for the off hand.

Down the road we would probably introduce 2-handed weapons (not before Flare 1.0). After we do that, we could make Bows require both hands. But it's not really necessary. 2H weapons are usually just "huge" weapons, and why not just let +physical warriors hold them with one hand.

[dorkster]

Ah, I didn't think about draw order.

[igorko]

Migrated to clintbellanger/flare-game#37