The Month of the NPC - Roadplan in OP, creature-rewrite branch in repo

[GlyphGryph]

We won’t even be thinking about factions and stuff just yet. This is just getting the foundations stable.

Thanks for all your work, guys - your code certainly IS ugly, swwu, but that doesn’t stop it from being beautiful.

So now what we need is to create a base class for Player that is a child of Creature, but can also be a parent to the NPC.

So that we can collaborate more easily, would you guys want access to the CleverRaven repo, and then we can push up a dev branch for the NPC rewrite, but we can still all basically work with and stay up to date with the most recent branch?

Alternatively, we could just work off swwu’s repo, which would be fine.

[youtoo]

are you going to json the npcs? I belive the discussion was in part around

1. mob type monsters
2. npc types that wear equipment and have interactions.

is that correct?

[swwu]

Either sounds fine to me. It might make more sense to have the new branch be on main repo (just so it will be easier to find for anyone who wanders in later) but it’s not super important to me, whichever’s more convenient for you guys.

I should have shared effects/diseases for the base creature class in by tonight and then we can start planning out how to actually do NPC things after?

re: factions, yeah, from where I’m looking at the code from here they are probably going to be a long ways off

[CodeHero]

I want to help! I just need to know what there is to be done.

[GlyphGryph]

are you going to json the npcs?

i don’t even know what this would mean, honestly? But at least some NPC definitions will end up existing in the JSON so I can fill the kickstarter rewards.

[swwu]

Okay, all monster effects now use the new effects system (although in name only, since their processing/whatever is still done in exactly the same way, and we won’t be able to really change that until player diseases get fully ported). Both players and monsters now also use the “down” and “stun” effect (rather than disease) so some of the combat is back. The rest of the player diseases are still… being worked on - in general the handling of player diseases is a lot looser than the handling of monster effects, so they’re more difficult to track down (some of them also hit bodyparts, which means effects will have to be bodypart-aware, even if most monsters won’t actually have multiple bodyparts).

Should be up on my repo now (https://github.com/swwu/Cataclysm-DDA/tree/creature-rework), I cheated my skills to 10 and managed to kill a few hundred Zs with various weapons without any segfaults, so hopefully it’s sorta stable.

[GlyphGryph]

I’ll work on getting this on a main-repo branch so contributors only have to point at the one. I think swwu already has repo access there - anyone else who is interested, let me know so I can give you access.

[GlyphGryph]

And done.

https://github.com/CleverRaven/Cataclysm-DDA/tree/creature-rework

Note that I’m still not back home, so my fiddling around with it has been limited. But I like what I see so far.

[swwu]

GlyphGryph, I don’t actually have push access yet, if that’s what you meant by repo access.

[TheGrifter]

NPCs are coming back? Awesome. Are the factions going to work this time? Two way radio might have a use!

Ought to throw in some Walking Dead references. A walled off prison filled with survivors, things like that.

[GlyphGryph]

swwu, you should have access now. I am now home, and will be spending tonight helping. If you’re online and interested, we should probably coordinate.

[swwu]

Sounds good, I will be on IRC for another hour or two at least, and probably all day tomorrow.

FWIW I am just moving the remaining combat-debuff diseases over to the new effects system today and sorting out whatever minor bugs I find.

[GlyphGryph]

I’ll be on in about half an hour.

[Heruedhel]

@Swwu: Hi. I am sorry I did not reply to your post earlier on, but I got distracted by RL for a few days there. Sure, it would be much better if we both worked on the same branch rather than try to merge our efforts later on. I’ll discard my changes and look at what you guys have done, and work from there.

@GlyphGryph: When I played with the code I came across a few architectural problems that I would like to address. These changes do not add anything to game play or solve any bugs, but they make the code a bit messy. Should I try doing something about them (which will touch many files, and potentially introduce bugs) or are these the domain of the “core” development team? Examples of this include circular dependencies, unnecessary parameters being passed to functions (e.g. game object being passed into many functions that don’t need it), and certain functionality being handled inconsistently in code (e.g. different parts of code use player::IsNPC() and a comparison to ‘u’ global for the purpose of deciding whether the actor is the player or an NPC)

[GlyphGryph]

Architectural fixes are definitely appreciated moving forward here, but note that architectural fixes in the NPC class won’t do much since we are replacing it. Instead, help us work on the new architecture on the creature-rewrite branch.

[Heruedhel]

The issues I am talking about aren’t restricted to the creature classes (So this might not be the place to discuss them really) so I imagine these going into mainline.
For example, the game class seems to be a lot of things at once. It is:
A- The driver: it initializes, loads, saves and quits the game.
B- Game World: It keeps track of where zombies and players are, light, fungus …etc
C- Player and monster actions: throwing, firing, using computers …etc are for some reason game methods that some times take a player object as input.
D- UI: display messages, draw explosions and bullets …etc, handle mouse.

This means that a class which needs any subset of this functionality, has to be exposed to the whole thing. and since game is the super class that knows about all other classes, we end up with lots of unnecessary loops.

This is just one example of the kind of thing I would like to address. Maybe I should start a new thread so as not to pollute the current thread?

[swwu]

Heruedhel: a lot of the player/monster actions you mention in item C will probably get redone in the npc rewrite we’re doing. It may be easier if you did those rewrites on the creature-rework branch just because I imagine the merge between two extensive architectural changes would be a pain.

Also, no need to discard your changes. You can just git rebase it to the rework branch.

Glyphgryph: thoughts on adding a “rework roadmap” google doc to this thread? Just a list of all the things that still need to be fixed, might make it easier for new people to jump in.

[ClaytonRISING]

I’d like to help.

Also, I want to push forth the suggestion of making NPCs as modular as the enemies currently are. Meaning that if someone wanted to make a total conversion mod and have NPCs/humanoids that use their own inventory and are inherently agressive, it’s possible.

Uh…wink wink

[GlyphGryph]

Glyphgryph: thoughts on adding a “rework roadmap” google doc to this thread? Just a list of all the things that still need to be fixed, might make it easier for new people to jump in.

If you write it, I’ll link it.

[swwu]

Okay, here is a task outline/roadmap, should be publicly editable: https://docs.google.com/document/d/1pWQ1FG5GAO7lZ9uo2FSxQ9MRzISwKXloA2LVQ6Af2A4/edit

If you want to help with a specific task, just stick your name next to it and go for it. Make sure all PRs you make are against the creature-rework branch and not master. If you’re not sure what a specific task means or how to start it, feel free to ask either here (in this thread or via PM) or on IRC.

Glyphgryph: I’m less familiar with the totality of stuff that needs to be done at the Player/Nonplayer/Character levels so those parts are a little bit sparse, you may need to list (and claim) some of them yourself.