Vapourware Codexian Game Development Thread

[Jaesun]

And just because I am slightly drunk, if you are going to be using QUEST variables in your game, using the Gold Box Game's examples will provide you with some really good basic RPG functions, just by checking a quests value:

QUEST# (VAR between 0 and 255) ALL Quest variables are initialized, and have a value of 0, meaning the player has NOT received that quest yet. Useful so you can see if they even have the quest yet.

Value 0 - Player has NOT received this quest yet

Value 1 - Player has accepted the quest, quest is STILL ACTIVE

Value 2 - 99 various stages of a quest, quest is STILL ACTIVE. Be sure to fully document EACH STAGE of a quest to see what each value means.

Value 100 - the quest has been COMPLETED

Value 200 - the quest has been REJECTED by the player

Value 255 - the quest has FAILED

 

[DavidBVal]

And just because I am slightly drunk, if you are going to be using QUEST variables in your game, using the Gold Box Game's examples will provide you with some really good basic RPG functions, just by checking a quests value:

QUEST# (VAR between 0 and 255) ALL Quest variables are initialized, and have a value of 0, meaning the player has NOT received that quest yet. Useful so you can see if they even have the quest yet.

Value 0 - Player has NOT received this quest yet

Value 1 - Player has accepted the quest, quest is STILL ACTIVE

Value 2 - 99 various stages of a quest, quest is STILL ACTIVE. Be sure to fully document EACH STAGE of a quest to see what each value means.

Value 100 - the quest has been COMPLETED

Value 200 - the quest has been REJECTED by the player

Value 255 - the quest has FAILED

Yeah, Gold Box was a pioneer in many RPG engine concepts, I think nearly every RPG followed a similar system from then on. My own variant of the above is, negative values for failed quest, and values of 100 and above for completed (back in the days of Gold Box, quests usually could only be failed or completed in one way, now you will probably want different outcomes). In the quest file, you add a description of each value, for the journal entry. From then, you have the choice to simply store the curent quest value, or if you want to have a proper journal, keeping all the states the quest went through, just store each game-timestamp and quest value, so you can show the player how the quest progressed over time.

 

[Zed]

fried my brain doing this shit but it's starting to come together. INI conversation editor

 

[Zed]

and an update on how it could look in-game. this is a test i sent to sensuki just to see that it works on computers that arent my own.

custom bitmap font

test your checks ftw

 

[Zed]

first draft character lmao nice hat

 

[zwanzig_zwoelf]

Zed, you know I love your stuff, but if you keep posting this kind of shit without turning them into finished games, I'm going to get seriously depressed.

 

[Zed]

Zed, you know I love your stuff, but if you keep posting this kind of shit without turning them into finished games, I'm going to get seriously depressed.

me too

 

[DavidBVal]

Finishing a game. In theory a possible thing to do.

Sprite looking good. Keep at it.

 

[Ninjerk]

Zed, you know I love your stuff, but if you keep posting this kind of shit without turning them into finished games, I'm going to get seriously depressed.

me too

shihonage brofists this.

Depressed mens unite

Spoiler

or just stay in bed today

 

[Zed]

Zed, you know I love your stuff, but if you keep posting this kind of shit without turning them into finished games, I'm going to get seriously depressed.

me too

shihonage brofists this.

Depressed mens unite

Spoiler

or just stay in bed today

hey I'm actually at work... thinking about combat systems. that's a good sign!.. I think.

 

[Castanova]

first draft character lmao nice hat

What is that software?

 

[Zed]

first draft character lmao nice hat

What is that software?

multimedia fusion. man it's all i know and all i use for everything

 

[Rhuantavan]

Keep 'em coming Zed!

 

[J1M]

Well, I'll make the editor eventually, but it'll work internally with these same file formats, so it won't suppose a problem nor I'll have to redo them. Sure if I had done the editor from the beginning, JSON would have been the perfect choice, directly serialized into conversation objects. But this will do too, and allows me to work easily on a smaller scale.

You're right, the scripting has caused bugs in the past, but there's already 100+ conversations and 30 quests, I'm fine so far. Luckily I have good testers.

You should be using some form of schema validation. Look up what .xsd does and find something similar for your format.

 

[r3jonwah85]

Added some functionality to the AI - they can now avoid drops/pits, jump over pits, jump down from ledges, strafe when attacking , roam around, fight eachother, navigate around obstacles, jump over small obstacles. So pretty much nothing that wasn't present in FPS-games 20 years ago...

 

[gaussgunner]

...Sure if I had done the editor from the beginning, JSON would have been the perfect choice, directly serialized into conversation objects.

I dunno, JSON is hell for hand editing, and only 4 datatypes. I stopped using it except for AJAX requests.

YAML is ok.

INI is nice and flat, better for C++. Handrolled is probably the way to go if you need more structure. Works for me.

 

[Zed]

quick n dirty char gen so I can set stats to try out my conversation editor.

I wanted a low value attribute system (1 is base), where background and gender (male +1str +1int ) decides starting attributes.
attributes are at this point only used for conversation but I want to use them for combat as well.

 

[r3jonwah85]

Urghhh, found a nasty performance killer in my aggro-code, but without that it works pretty ok with a relatively high enemy count (360 in this test).

 

[Zed]

Urghhh, found a nasty performance killer in my aggro-code, but without that it works pretty ok with a relatively high enemy count (360 in this test).

you should work on blood decals and splatter and shoot-off-limb physics- should be prio no 1!!

 

[r3jonwah85]

Urghhh, found a nasty performance killer in my aggro-code, but without that it works pretty ok with a relatively high enemy count (360 in this test).

you should work on blood decals and splatter and shoot-off-limb physics- should be prio no 1!!

That is probably the next step, but right now I want to have good performance in large numbers before proceeding. I have done two different pathfinding/avoidance algorithms depending on the needs, the more humanoid characters use the stuff in the second last video, the latest video use navigation straight out of Doom. But tonight I am going to make a test using physical bodies instead of character controllers since I think that will speed things up.

But a simple blood decal system does not have to be that hard, just spawn a quad and align to the surface with a small offset, collect them all in a instanced static mesh structure for perfromance. I think using the normal decal system will bog down performance too much, but will do some reading. Perhaps you or someone else has got some experience on the matter?

Shooting off limbs would be really satisfying, but I think it has to be reserved for enemies that appear in smaller numbers to be performant. But a quake-like gib-system is already in there, just need to make the chunks that fly around.

 

[Zed]

I have no experience on the matter.

But I do know this:
There's a big hole in the market after Soldier of Fortune II.

 

[r3jonwah85]

I have no experience on the matter.

But I do know this:
There's a big hole in the market after Soldier of Fortune II.

I can agree on that, the GHOUL system was really fun, the simpler system in SIN was also quite nice. GHOUL seems doable but quite labor intensive (https://www.youtube.com/watch?v=sd7fsY33L6Y), but perhaps something closer to Chasm: The Rift (https://www.youtube.com/watch?v=9m_iOHnv4Zk).

If anyone else wonders, after a quick read it seems that using a simple quad is most likely way faster than using the deferred decals in Unreal.

 

[Jaesun]

 

[Viata]

"open eyes for the first time in centuries"
You better include Lazy Motherfucker as background there.

 

[zwanzig_zwoelf]

quick n dirty char gen so I can set stats to try out my conversation editor.

I wanted a low value attribute system (1 is base), where background and gender (male +1str +1int ) decides starting attributes.
attributes are at this point only used for conversation but I want to use them for combat as well.

Kickstarter when?
Patreon when?
Release when?