Progress

Graphical juice

An interesting way to make the graphics look cooler that I discovered accidentally; mixing the color channels (for example, swapping red and blue) will give it a cool effect. Perhaps if this could be gradually blended in and out, maybe even as an energy overload indicator or grayscale to indicate lack of energy, to communicate that energetic state can affect visual perception.

AI

Refactor to Engine

At this point, the AI has become pretty interesting and fairly advanced, so I want to create an API out of it. That way, the AI engine can be used to run agents for other simulations.

Perhaps name the engine DuckMind.

The engine would be a component-based system, that would likely depend on Nez for math and types. However, the component system would be independent of Nez, and instead would be a relationship of "Mind has Components" rather than using polymorphism with the mind.

Ideally, this should provide all the current functionality with room for extension for simulation, specific things, but should also maximize the amount of shared code that can be extracted to the library, as that would allow broader usage of the library without first creating a detailed scenario.

It is okay for init time to be higher, as long as per-loop hot performance of the mind remains good.

Currently, the AI is built in a tiered system. Since not every node of each tier requires a tier below it (this explodes complexity), the tiers should be as independent and decoupled as possible. However, the tools for each tier will still be available; just that using the broadest tier will not necessitate using more specific tiers.

Duck Intelligence and Socialization Engine (DISE)

Sor implements an AI system that allows it to model duck intelligence and socialization.

Lunchlib AI Continuation

imported and adapted the code from older agent logic research project ("Lunch Simulator") start working on the personality vector representation.

Architecture
Logic algorithm split

reasoner is generally reserved for choosing higher-level actions. the actual execution of these actions is left to simpler, more precise subroutines.

Hierarchy

Use the utility planner to select the most "useful" objective. it gives the planner its target world state and leaves it up to the planner to figure out the actions necessary to reach that goal state. the reasoner just decides what the best goal is, the planner figures out how to reach that goal.

based on my research so far of goap, i want to use the XGOAP library. since it looks like most of those are based on boolean preconditions and postconditions (for obvious reasons), we need to let goap do high-level planning of the steps/actions required. for example, goap could tell us to "eat beans."

actually carrying out that task of eating beans will mean constructing yet another plan on how to achieve that goal, though this will not be done by a planner, it will be done by handwritten code paths. for example, to eat beans, i will have code to eat the nearby beans in a particular order until a quota is reached. it will pursue goals in order. if an order becomes no longer valid, it will be canceled. if it was a critical order, the whole plan will be invalidated. then, a new plan will have to be formed. in addition, if utility decides that a new objective is more important, the current action plan will be discarded and a new one will be instituted.

Separating action planning from action execution

both the action planning and execution systems should be distinct and able to operate independently (minimally coupled). that way, the action execution only needs to care about how to carry out the immediate task in the world given very detailed instructions. all the plan generators need to create detailed plans to feed to these though. for example, if we are adding pathfinding, we shouldn't need to add any new support to the plan executor; all we need to do is add some more plans to our task queue.

Performance/Optimization

the lighter the computational load of each unit's AI is, the more AIs we can support. ideally we would aim to reduce the compute load by reducing unnecessary or unused computations, and decreasing the frequency of expensive ones. a lot of the sense/think can happen on worker threads. for example, vision is not something that needs to be recomputed every update frame. significant savings could be gained by simply updating vision every 0.2sec or such. one cool idea! make the ai update timers flexible, so for ai out of the scene the updates can be much slower and thus much performance can be gained. we can focus our compute resources on the ai nearby, as that will be more interesting. for now, use Tasks and manual task management, but in the future if optimization can be gained by using Nez primitives such as Nez Coroutine, look into that.

Threading

Use Interlocked for atomic variable operations across threads.

Personality

eating/feeding

receptiveness based on traits. this is used to scale the opinion bonus from being fed a capsule.

a revision that would be cool is to make birds easier to feed; this would be done

emotions

need to add emotions soon.

Considerations

Social

Social objective aims to make friends with other birds, usually by feeding them. In the future perhaps it could do things like build rapport by proximity support to a friendly bird. The consideration will only be selected if there's a bird with a plausible chance of making a friend with. Thus, we're guaranteed to have at least one nearby bird that is a candidate. Select the first candidate (this represents that candidate being the highest expected value for feeding) and try to make friends with them. In the future perhaps use multiple heuristics and a GOAP solver to figure out what birds to feed and for what costs, taking into account things like speed discrepancies between the birds or distance.

Explore

to-do

Defend

to-do

Pathfinding

A* heuristics

excellent reading on use of the heuristic: http://theory.stanford.edu/~amitp/GameProgramming/Heuristics.html map representations: http://theory.stanford.edu/~amitp/GameProgramming/MapRepresentations.html

Structural navigation graph

The map analysis stage can build a proper room graph representing connections between the rooms, but this may not be particularly useful for actually understanding how to navigate the structure of the world. Instead, the room graph should be used to compute a "structural navigation graph" that contains nodes with positions corresponding to world positions; each room will translate into multiple nodes on this graph. For example, a room will have a node at the center, representing "being inside" the room, and two additional nodes per door, one slightly inside the door and one slightly outside. This way, a bird can path from any room to any other room by simply following the open lanes from doorway to doorway.

To build the structural navigation graph, start with the room graph. Then, we add four "nodes of indirection" between each room-to-room link; first the inner-door of the local room, then the outer-door of the local room, then the outer-door of the far room, then the inner-door of the far room. We use breadth-first-traversal on the room graph to discover all the nodes. Then, we map each room node to a "structural navigation room center" node, each of which start off completely unconnected. Then, we iterate through all the center rooms, iterating through each of their door-links. For each door-link, we then generate the four nodes of indirection (we only do the local two, and each of these are mapped to the "Door" edge so that we can look it up. when the other side of the link is processed, we do the other two.) The end result of this processing will be the structural navigation graph, which can then be traversed by A*.

Persistence

Saving generated maps

Possible ways:

  • save the rng seed used for mapgen

    • this may be bad because rng seed is used for plenty of other generation than just mapgen

    • if this is the option chosen, a separate rng should be created for mapgen

  • serialize the grid and recreate the roomrects from that

Mechanics

Fruit trees

In rooms, fruit seeds can be planted that grow into trees that release capsules. trees can be generated. perhaps seeds can be spread and new trees can start to grow. Later stages (7-10) release fruit.

Pips

pips represent 40 opinion (red, orange, yellow, blue, green) <-500 to -300: red -300 to -100: orange -100 to 100: yellow 100 to 300: blue 300 to >500: green

Boosting mechanics

Boosting now drains energy at a per-second rate. While boosting, the bird passes through any walls.

Evolution

Make bird kinematics and everything computed based on functions of its mass.

Map generation

Tilemap analysis

we attempt to analyze the tilemap to convert it into a better data structure. one important thing we want to be able to do is detection of rooms.

room detection

attempt to simply find all the rooms by using the corners. find the top left corner, go right until a top right corner, and etc. to create a clockwise loop. This can be saved as a Room. to detect doors and such edges, while scanning along the edges, any gaps are listed as exits.

edge detection

after a room detection pass which finds all rooms and their doors, we iterate over all the rooms and each of their doors. from a corner of the door, we scan outward until we intersect another room. if the intersect happens before out-of-range, then we add an edge.

Tilemap generation

Start with a big grid.

Links

  1. https://gamedev.stackexchange.com/questions/121823/how-do-i-implement-level-generation-based-on-graph-grammars

  2. https://ondrejnepozitek.github.io/ProceduralLevelGenerator-Unity/docs/roomTemplates

  3. http://theory.stanford.edu/~amitp/GameProgramming/MapRepresentations.html

  4. https://www.reddit.com/r/roguelikedev/comments/6vuw7i/faq_fridays_revisited_22_map_generation/

  5. https://www.gridsagegames.com/blog/2014/06/procedural-map-generation/

  6. https://ondra.nepozitek.cz/blog/42/dungeon-generator-part-1-node-based-approach/

  7. https://www.reddit.com/r/proceduralgeneration/comments/1sx5ul/graphbased_dungeon_generator_improvement_ideas/

  8. https://gamedev.stackexchange.com/questions/38721/how-can-i-generate-a-navigation-mesh-for-a-tile-grid

  9. https://gamedev.stackexchange.com/questions/22178/is-there-an-algorithm-to-detect-the-mainland-on-a-2d-map

Funny bugs

"Spill the beans"

so i feed this bird, and its social objective tells it it wants to feed me so it goes to a safer distance out of caution, then while trying to feed me it accidentally just releases all of its energy beans and then it starts chasing after them again; meanwhile, Big Bird is like gimme some of those :b:eans

Time log

  • As of today (2020-01-22 00:46:25), have logged 13h45m into the project.

  • As of right now (2020-01-22 19:07:50) have logged about 20h into the project.

  • As of today (2020-03-03 15:31:52), have logged about 61h into the project.

  • As of today (2021-01-20 20:00:16), have logged 136h into the project.

Tasks [0/11]

TODO control [3/6]

DONE add down to slow down

CLOSED: [2020-01-12 Sun 23:21]

  • State "DONE" from "TODO" [2020-01-12 Sun 23:21]

DONE only WALL-tagged colliders should suck velocity

CLOSED: [2020-01-12 Sun 23:34]

  • State "DONE" from "TODO" [2020-01-12 Sun 23:34]

TODO improve collision

  • State "TODO" from "REVIEW" [2020-01-21 Tue 20:27]

  • State "REVIEW" from "TODO" [2020-01-17 Fri 01:26]
    i mean? idk

    DONE wing-wing collision

CLOSED: [2020-01-14 Tue 15:12]

  • State "DONE" from "NEXT" [2020-01-14 Tue 15:12]

    DONE collision momentum

CLOSED: [2020-01-14 Tue 15:12]

  • State "DONE" from "TODO" [2020-01-14 Tue 15:12]

    BLOCKED collision bounce/elasticity

  • State "BLOCKED" from "TODO" [2020-01-21 Tue 20:27]
    do we realy want bounce?

    ARCHIVED raycast checking

CLOSED: [2020-01-21 Tue 21:31] boost tends to clip through things, we want that fixed using a raycast, turning boost off if invalid.

REVIEW tweak handling constants

  • State "REVIEW" from "TODO" [2020-01-14 Tue 15:52]
    tweaked the movement to feel a bit better

REVIEW add interact and utility keys

  • State "REVIEW" from "TODO" [2020-01-17 Fri 01:26]
    there is interact key now

DONE boost

CLOSED: [2020-01-15 Wed 16:01]

  • State "DONE" from "TODO" [2020-01-15 Wed 16:01]

TODO ai [1/3]

TODO systems [12/12]

DONE groundwork

CLOSED: [2020-01-14 Tue 11:43]

  • State "DONE" from "TODO" [2020-01-14 Tue 11:43]

DONE basic ai controller support

CLOSED: [2020-01-14 Tue 11:42]

  • State "DONE" from "NEXT" [2020-01-14 Tue 11:42]

DONE decision debug display

CLOSED: [2020-01-14 Tue 11:43]

  • State "DONE" from "REVIEW" [2020-01-14 Tue 11:43]

  • State "REVIEW" from "DOING" [2020-01-14 Tue 11:42]
    added working display

  • State "DOING" from "TODO" [2020-01-14 Tue 11:42]

DONE logic Sense-Think-Act outline [3/3]

CLOSED: [2020-02-05 Wed 01:03]

  • State "DONE" from "TODO" [2020-02-05 Wed 01:03]

    DONE Sense outline

CLOSED: [2020-01-22 Wed 18:13]

  • State "DONE" from "NEXT" [2020-01-22 Wed 18:13]

    DONE vision

CLOSED: [2020-01-17 Fri 01:27]

  • State "DONE" from "NEXT" [2020-01-17 Fri 01:27]

    DONE object tracking

CLOSED: [2020-01-22 Wed 18:13]

  • State "DONE" from "TODO" [2020-01-22 Wed 18:13]

should be able to track capsules and what happens to them

DONE Think outline

CLOSED: [2020-03-03 Tue 00:38]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:38]

DONE Act outline

CLOSED: [2020-03-03 Tue 00:38]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:38]

DONE personality engine port

CLOSED: [2020-01-22 Wed 18:14]

  • State "DONE" from "NEXT" [2020-01-22 Wed 18:14]

DONE ai representations

CLOSED: [2020-02-05 Wed 01:03]

  • State "DONE" from "TODO" [2020-02-05 Wed 01:03]

DONE personality engine

CLOSED: [2020-02-05 Wed 01:03]

  • State "DONE" from "TODO" [2020-02-05 Wed 01:03]

    DONE opinion

CLOSED: [2020-01-23 Thu 11:11]

  • State "DONE" from "TODO" [2020-01-23 Thu 11:11]

    DONE per-mind opinion storage

CLOSED: [2020-01-23 Thu 11:11]

  • State "DONE" from "TODO" [2020-01-23 Thu 11:11]

    DONE mind for player

CLOSED: [2020-01-22 Wed 18:21]

  • State "DONE" from "TODO" [2020-01-22 Wed 18:21]

    DONE non-control minds

CLOSED: [2020-01-22 Wed 18:15]

  • State "DONE" from "NEXT" [2020-01-22 Wed 18:15]

    DONE fixed personality for player

CLOSED: [2020-01-22 Wed 18:21]

  • State "DONE" from "NEXT" [2020-01-22 Wed 18:21]

    TODO interactions
    TODO signaled capsule feed TODO make feed capsule use trait interaction
    TODO add mood to mind
DONE representation of resources to AI

CLOSED: [2020-03-03 Tue 00:37]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:37]

CANCELED representation of tethers and links to AI

CLOSED: [2020-03-03 Tue 00:40]

  • State "CANCELED" from "TODO" [2020-03-03 Tue 00:40]
    no tethers

DONE signaling review

CLOSED: [2020-02-05 Wed 01:03]

  • State "DONE" from "TODO" [2020-02-05 Wed 01:03]

DONE ai movement

CLOSED: [2020-02-05 Wed 01:03]

  • State "DONE" from "TODO" [2020-02-05 Wed 01:03]

    DONE turn ai

CLOSED: [2020-02-01 Sat 14:40]

  • State "DONE" from "TODO" [2020-02-01 Sat 14:40]

    DONE move ai

CLOSED: [2020-02-01 Sat 14:40]

  • State "DONE" from "TODO" [2020-02-01 Sat 14:40]

    DONE make the code less shit

CLOSED: [2020-02-05 Wed 01:03]

  • State "DONE" from "TODO" [2020-02-05 Wed 01:03]

    PROG show unit plans

  • State "PROG" from "NEXT" [2020-03-03 Tue 00:37]

DONE action planning

CLOSED: [2020-02-06 Thu 09:25]

  • State "DONE" from "TODO" [2020-02-06 Thu 09:25]

    DONE move action planning

CLOSED: [2020-03-03 Tue 00:35]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:35]

    DONE target move planning

CLOSED: [2020-02-06 Thu 09:25]

  • State "DONE" from "TODO" [2020-02-06 Thu 09:25]

    DONE target distance follow planning

CLOSED: [2020-02-06 Thu 09:25]

  • State "DONE" from "TODO" [2020-02-06 Thu 09:25]

    TODO use time validity on task plans too
    TODO add an approach type that is rougher and doesn't attempt to distance itself

sometimes it tries to back off if it gets too close for the feed

TODO refinement [0/3]

TODO make nearby interaction less exploitable

right now being near it while friendly will indefinitely increase opinion

TODO wander movement [3/3]

  • State "DONE" from "TODO" [2020-03-03 Tue 13:32]

    DONE pathfinding over room graph

CLOSED: [2020-03-03 Tue 00:40]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:40]

    DONE translate pathfind results into action queue

CLOSED: [2020-03-03 Tue 13:32]

  • State "DONE" from "NEXT" [2020-03-03 Tue 13:32]

    DONE navigate using structural navigation graph

CLOSED: [2020-03-06 Fri 11:16]

  • State "DONE" from "TODO" [2020-03-06 Fri 11:16]

TODO birds should attempt to bond

DONE ai inspector [6/6]

CLOSED: [2020-03-06 Fri 11:15]

  • State "DONE" from "TODO" [2020-03-06 Fri 11:15]

  • State "DONE" from "TODO" [2020-02-05 Wed 01:03]

    DONE support ai inspector component

CLOSED: [2020-01-14 Tue 13:10]

  • State "DONE" from "TODO" [2020-01-14 Tue 13:10]

    DONE show unit data in inspector

CLOSED: [2020-01-23 Thu 11:06]

  • State "DONE" from "TODO" [2020-01-23 Thu 11:06]

    DONE collect inspect data from both Mind and State

CLOSED: [2020-01-23 Thu 11:06]

  • State "DONE" from "TODO" [2020-01-23 Thu 11:06]

    DONE show unit vision

CLOSED: [2020-01-23 Thu 11:06]

  • State "DONE" from "TODO" [2020-01-23 Thu 11:06]

    DONE show unit thoughts

CLOSED: [2020-03-03 Tue 00:36]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:36]

    DONE value board [2/2]

CLOSED: [2020-03-06 Fri 11:15]

  • State "DONE" from "TODO" [2020-03-06 Fri 11:15]

    DONE separate sections in board display

CLOSED: [2020-03-06 Fri 11:15]

  • State "DONE" from "TODO" [2020-03-06 Fri 11:15]

    DONE board line color support

CLOSED: [2020-03-06 Fri 11:15]

  • State "DONE" from "TODO" [2020-03-06 Fri 11:15]

TODO visual [5/8]

DONE pips [4/4]

CLOSED: [2020-01-22 Wed 20:31]

  • State "DONE" from "NEXT" [2020-01-22 Wed 20:31]

    DONE pip sprites

CLOSED: [2020-01-22 Wed 19:16]

  • State "DONE" from "TODO" [2020-01-22 Wed 19:16]

    DONE only onscreen pips

CLOSED: [2020-01-22 Wed 20:31]

  • State "DONE" from "TODO" [2020-01-22 Wed 20:31]

    DONE update pips based on opinion

CLOSED: [2020-01-22 Wed 20:31]

  • State "DONE" from "NEXT" [2020-01-22 Wed 20:31]

    DONE pips represent opinion

CLOSED: [2020-01-22 Wed 20:31]

  • State "DONE" from "TODO" [2020-01-22 Wed 20:31]

DONE add screen shake

CLOSED: [2020-01-17 Fri 01:26]

  • State "DONE" from "TODO" [2020-01-17 Fri 01:26]

DONE add trail (disabled)

CLOSED: [2020-01-17 Fri 01:27]

  • State "DONE" from "TODO" [2020-01-17 Fri 01:27]

DONE add indicator bar

CLOSED: [2020-02-06 Thu 09:50]

  • State "DONE" from "REVIEW" [2020-02-06 Thu 09:50]

  • State "REVIEW" from "TODO" [2020-01-21 Tue 16:53]
    added indicator bar to show energy

DONE fruits with more energy pulsate faster

CLOSED: [2020-02-06 Thu 09:50]

  • State "DONE" from "TODO" [2020-02-06 Thu 09:50]

TODO overload indicator

show overloaded ship (energy > 2x capacity) by tinting it.

TODO color filters and vfx

TODO menu redesign

TODO mechanics [9/10]

CANCELED phase switching [0/4]

CLOSED: [2020-01-22 Wed 11:53]

  • State "CANCELED" from [2020-01-22 Wed 11:53]
    unnecessary

  • State "BLOCKED" from "TODO" [2020-01-22 Wed 11:51]
    is this needded?

    TODO support different physics based on phase
    TODO implement phase physics
    TODO create phase transitions

CANCELED wall attachment [0/6]

  • State "BLOCKED" from "TODO" [2020-01-22 Wed 11:51]
    is this really needed?

    TODO inspect InfiniteGravity code
    TODO animation support
    TODO input design for controls
    TODO physical snapping to walls
    TODO movement along walls
    TODO boosting off walls

CANCELED tethers

CLOSED: [2020-02-06 Thu 21:41]

  • State "CANCELED" from "TODO" [2020-02-06 Thu 21:41]

    TODO manual tether attachment

DONE disable walls

CLOSED: [2020-03-03 Tue 00:34]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:34]

DONE trees [4/4]

CLOSED: [2020-02-06 Thu 21:41]

  • State "DONE" from "TODO" [2020-02-06 Thu 21:41]

    DONE tree loading

CLOSED: [2020-01-21 Tue 23:37]

  • State "DONE" from "TODO" [2020-01-21 Tue 23:37]

    DONE tree stage animation

CLOSED: [2020-01-21 Tue 23:37]

  • State "DONE" from "TODO" [2020-01-21 Tue 23:37]

    DONE tree growth

CLOSED: [2020-01-22 Wed 00:08]

  • State "DONE" from "TODO" [2020-01-22 Wed 00:08]

    DONE trees dropping fruit

CLOSED: [2020-01-22 Wed 00:08]

  • State "DONE" from "TODO" [2020-01-22 Wed 00:08]

DONE boosting [2/2]

CLOSED: [2020-01-23 Thu 07:59]

  • State "DONE" from "REVIEW" [2020-01-23 Thu 07:59]

  • State "REVIEW" from "TODO" [2020-01-22 Wed 11:50]
    basic boosting mechanics functional

    DONE boosting drains energy

CLOSED: [2020-01-21 Tue 21:48]

  • State "DONE" from "TODO" [2020-01-21 Tue 21:48]

    DONE boost time delay

CLOSED: [2020-01-22 Wed 11:50]

  • State "DONE" from "TODO" [2020-01-22 Wed 11:50]

DONE lanes

CLOSED: [2020-01-23 Thu 07:59]

  • State "DONE" from "REVIEW" [2020-01-23 Thu 07:59]

  • State "REVIEW" from "TODO" [2020-01-22 Wed 11:50]
    basic lanes are fucntioning

    DONE analyze lanes in map and create triggers

CLOSED: [2020-01-22 Wed 11:50]

  • State "DONE" from "TODO" [2020-01-22 Wed 11:50]

DONE capsules [6/6]

CLOSED: [2020-01-22 Wed 11:50]

  • State "DONE" from "REVIEW" [2020-01-22 Wed 11:50]

  • State "REVIEW" from "TODO" [2020-01-22 Wed 11:49]
    add working capsules

    DONE capsule creation

CLOSED: [2020-01-22 Wed 00:20]

  • State "DONE" from "TODO" [2020-01-22 Wed 00:20]

    DONE capsule transfer

CLOSED: [2020-01-22 Wed 00:20]

  • State "DONE" from "TODO" [2020-01-22 Wed 00:20]

    DONE capsules as fruit

CLOSED: [2020-01-22 Wed 00:20]

  • State "DONE" from "TODO" [2020-01-22 Wed 00:20]

    DONE capsule gravity

CLOSED: [2020-01-22 Wed 00:38]

  • State "DONE" from "TODO" [2020-01-22 Wed 00:38]

    DONE capsule acquire animation

CLOSED: [2020-01-22 Wed 00:45]

  • State "DONE" from "TODO" [2020-01-22 Wed 00:45]

    DONE capsule despawn

CLOSED: [2020-01-23 Thu 08:21]

  • State "DONE" from "TODO" [2020-01-23 Thu 08:21]

DONE world state persistence

CLOSED: [2020-02-06 Thu 21:41]

  • State "DONE" from "NEXT" [2020-02-06 Thu 21:41]

NEXT make kinematics computed based on functions of mass

TODO bugs [4/4]

DONE pull Nez

CLOSED: [2020-01-13 Mon 00:09]

  • State "DONE" from "NEXT" [2020-01-13 Mon 00:09]

DONE reintroduce Nez as submodule

CLOSED: [2020-01-13 Mon 00:08]

  • State "DONE" from "NEXT" [2020-01-13 Mon 00:08]

DONE collision glitching

CLOSED: [2020-01-14 Tue 11:42]

  • State "DONE" from "REVIEW" [2020-01-14 Tue 11:42]

  • State "REVIEW" from "TODO" [2020-01-14 Tue 11:42]
    solved by simply increasing the border

    REVIEW look at KinematicBody in Glint and at Mover and see if they can be used to help correct.

  • State "REVIEW" from "TODO" [2020-01-14 Tue 11:41]
    looked at it, might need to port o

    BLOCKED try to replicate pos collision MinTranslation glitching

  • State "BLOCKED" from "TODO" [2020-01-14 Tue 11:42]
    CNR

DONE in MapLoader::adjustColliders center tile

CLOSED: [2020-03-03 Tue 00:35]

  • State "DONE" from "REVIEW" [2020-03-03 Tue 00:35]

  • State "REVIEW" from "TODO" [2020-01-21 Tue 22:15]
    check left tile if center tile is null

check more than just the center tile (see note)

TODO refactor [1/1]

CANCELED refactor collision in WingBody to use a System

CLOSED: [2020-01-23 Thu 11:04]

  • State "CANCELED" from "TODO" [2020-01-23 Thu 11:04]
    unncessary

TODO map [1/2]

DONE representation of map as room graph

CLOSED: [2020-03-03 Tue 00:35]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:35]

  • State "DONE" from "TODO" [2020-01-23 Thu 11:04]

    DONE convert from tile grid to rooms

CLOSED: [2020-01-23 Thu 11:04]

  • State "DONE" from "TODO" [2020-01-23 Thu 1

    TODO interpret rooms and doors as graph

TODO map generation [2/3]

DONE full graph representation mapping

CLOSED: [2020-03-03 Tue 00:35]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:35]

DONE map graph generation

CLOSED: [2020-03-03 Tue 00:35]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:35]

TODO convert from graph to rough tile grid

TODO design [2/3]

DONE weapon/shoot design

CLOSED: [2020-03-03 Tue 00:35]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:35]

TODO tether design

DONE ai design

CLOSED: [2020-01-23 Thu 11:04]

  • State "DONE" from "REVIEW" [2020-01-23 Thu 11:04]

  • State "REVIEW" from "TODO" [2020-01-14 Tue 13:24]
    written section in plan

TODO assets [2/2]

CANCELED phase switch assets

CLOSED: [2020-01-23 Thu 11:03]

  • State "CANCELED" from "TODO" [2020-01-23 Thu 11:03]

DONE game icon

CLOSED: [2020-01-23 Thu 11:04]

  • State "DONE" from "TODO" [2020-01-23 Thu 11:04]

TODO platform [4/4]

DONE resolution

CLOSED: [2020-01-22 Wed 18:37]

  • State "DONE" from "TODO" [2020-01-22 Wed 18:37]

    DONE custom resolution

CLOSED: [2020-01-22 Wed 18:37]

  • State "DONE" from "TODO" [2020-01-22 Wed 18:37]

    DONE custom scaling mode

CLOSED: [2020-01-22 Wed 18:37]

  • State "DONE" from "TODO" [2020-01-22 Wed 18:37]

DONE binary distribution

CLOSED: [2020-01-22 Wed 18:37]

  • State "DONE" from "TODO" [2020-01-22 Wed 18:37]

    DONE native build doesn't include bin

CLOSED: [2020-01-22 Wed 18:37]

  • State "DONE" from "TODO" [2020-01-22 Wed 18:37]

DONE logging

CLOSED: [2020-01-23 Thu 11:03]

  • State "DONE" from "TODO" [2020-01-23 Thu 11:03]

    DONE logging configurable verbosity

CLOSED: [2020-01-22 Wed 20:31]

  • State "DONE" from "TODO" [2020-01-22 Wed 20:31]

    DONE more logging

CLOSED: [2020-01-23 Thu 07:59]

  • State "DONE" from "TODO" [2020-01-23 Thu 07:59]

    DONE logging sinks

CLOSED: [2020-01-23 Thu 11:03]

  • State "DONE" from "NEXT" [2020-01-23 Thu 11:03]

    DONE only trace log AI with inspectors

CLOSED: [2020-01-23 Thu 08:03]

  • State "DONE" from "TODO" [2020-01-23 Thu 08:03]

DONE game exit

CLOSED: [2020-03-03 Tue 00:35]

  • State "DONE" from "TODO" [2020-03-03 Tue 00:35]