## Description:
- `AiAttackBehavior`: Because bots delete stolen structures now, nations
prioritize attacking bots with structures
- `NationMIRVBehavior`: Nations no longer MIRV enemies who already got
MIRVed in the last 30 seconds. Some humans complained about getting
double-MIRVed by nations. And in games with very high starting gold, ALL
nations MIRVed the same player (stop steamroll logic).
- `NationAllianceBehavior`: Fixes a comparison logic bug (Thanks to
Deshack)
- `NationNukeBehavior.ts`: Little atom bomb perceived cost balance
change
- `MIRVExecution`: To make sure the MIRVing nations are attacking the
MIRVed nations (even if they don't share a border), the relation gets
updated in both directions now.
- `SinglePlayerModal` & `HostLobbyModal`: Update the default difficulty
to "Medium" (to synchronize the defaults with the public game default)
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
Resolves#3154
## Description:
#2716 introduced nuke cancellation logic on alliance acceptance via
`AllianceRequestReplyExecution`. The radial menu action, though, calls
`AllianceRequestExecution` instead, which accepts the alliance if a
request has already been made by the other player.
This PR moves the nuke cancellation logic to `GameImpl`, hooking into
the `acceptAllianceRequest` method, therefore accounting for every
alliance acceptance, regardless of the specific action that brought to
that.
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
deshack_82603
## Description:
When a train station is removed, the clusters are recomputed. However
the cluster recomputation code has not been changed from the original
rail network implementation, which was a tree.
The deletion code made assumptions that are not true anymore since we
introduced loops in the network. As a result the cluster recomputation
was very inefficient, although the data was correct.
Changes:
- Fix clusters computation when a structure is deleted
- Structures are frequently deleted in bulk: atom/hydro/MIRV.
Re-computing the clusters when a single structure is deleted would be
inefficient because the recomputed cluster would probably need to be
recomputed again in the same tick.
Instead, when a structure is deleted, flag the cluster as "dirty", and
recompute all the dirty clusters once per tick only.
Previous performances (hydro over a dense area):
<img width="700" height="160" alt="image"
src="https://github.com/user-attachments/assets/cd3ceb42-6d5f-4ad1-b35a-f8e5e0513821"
/>
Now:
<img width="450" height="269" alt="image"
src="https://github.com/user-attachments/assets/55dec3b9-8619-4a6c-9a16-f5368fe40da1"
/>
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
IngloriousTom
## Description:
- Remove alliances on death: after death, alliances would stay active
including countdown timers and (when dead player kept spectating) icons.
Now remove them when player becomes inActive.
- Moved code to private method within PlayerExecution + added comments
in NationExecution and BotExecution for more clarity as to where
removals are performed from at death
- Remove renewal request from Events Display when Alliance doesn't exist
anymore (after death or otherwise).
- Also cleanup this.alliancesCheckedAt when alliance doesn't exist
anymore. Before, old/broken alliance id's would accumulate in it during
a game.
- Removed now-redundant isAlive check in EventsDisplay. Both the
alliances array as the isAlive are updated in the same tick from
PlayerUpdates so now alliance is removed from alliances array on player
death, the other.isAlive() check is no longer needed. Of course we could
keep it in just to be very safe, so just let me know when you're
doubtful about this.
- Attack.test.ts: fix failing test. Player B dies because of the attack,
meaning the alliance now gets removed. Prevent this by gving both a
different, adjecent, starting tile. And to be more clear about what is
needed for the test to pass, add isAlive check for both of them after
the attacks.
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
tryout33
Resolves#2997
## Description:
### New stuff
- Nations can upgrade structures now. They do it if they have too many
structures compared to their territory size
- They prefer to upgrade stuff thats protected by SAMs (based on
difficulty)
- Updated the build order, it also depends a bit on the difficulty now
(easy nations build less SAMs)
- Nations can handle extreme amounts of gold now. 500M starting gold? no
problem. Previously they only built cities
- They stop saving up for MIRV if they can afford it (in some old Enzo
"impossible difficulty experiment" videos you could see nations with
like 300M gold...)
- The save-up-target changes when bombs / hydros / MIRVs are disabled
- Added many checks for disabled units. For example: Don't build SAMs
when missile silos are disabled, focus on factories when ports are
disabled
- Updated the `structureSpawnTileValue` method, SAM-placement depends a
bit on the difficulty now
### Refactor
- Moved all structure related nation code into
`NationStructureBehavior.ts`
- Split up the good old `structureSpawnTileValue` method to make it more
readable
- Cleaned up NationExecution a bit
### A screenshot
<img width="1681" height="755" alt="Screenshot 2026-02-08 001108"
src="https://github.com/user-attachments/assets/c9b3df01-41ca-4c68-b450-b20e7d7d910a"
/>
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
added sam missile immunity (was missing from the immunity list)
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
w.o.n
## Description:
The `RailroadLayer` simply displays tiles as instructed by the core
worker. While it's practical for the layer to only care about the tiles,
it also means it has no understanding of railroads as entities (their
paths, connections, or identities).
It also means that the core worker is responsible for rendering tasks
such as tile orientation and construction animation, which is not
expected.
To support ID-based events and better separation of concerns, the
rendering layer needs to be aware of complete railroads. With this
change, the core worker can send the tiles once and subsequently
reference railroads only by ID for all other events.
#### Changes:
- `RailroadLayer` now stores full railroad data instead of only
individual tiles
- `RailroadLayer` is responsible for animating newly built railroads
- Add a new `RailroadSnapUpdate` sent when a new structure is built over
an existing railroad. This event is used by `RailroadLayer` to keep
railroad ID in sync.
- When hovering over a railroad, the render worker is querying the core
worker about overlapping railroads.
Alternatively, RailroadLayer could compute overlaps itself now that it
has full railroad knowledge, but this logic would need to be duplicated
and kept in sync across workers. Keeping a single source of truth in the
core worker is preferred.
#### Edgecases:
- When a structure snaps over a railroad, the original railroad is split
into two new railroads. If the construction animation is still in
progress, instead of resuming the animation at the correct point on the
new railroads, all remaining tiles are rendered immediately
- Previously, `RailroadUpdate` handled both construction and
destruction. This no longer works with `RailroadSnapUpdate`, as event
ordering is now pretty important and IDs may be lost before they are
consumed.
To address this, RailroadUpdate is split in two:
`RailroadConstructionUpdate` and `RailroadDestructionUpdate`.
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
IngloriousTom
---------
Co-authored-by: jrouillard <jon@rouillard.org>
## Description:
- Removed all code related to generating a client ID on the client. The
server now assigns the client ID and sends it to the client in lobby
messages.
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
w.o.n
## Description:
Train spawning hot-path optimization (trade destination selection)
## Summary
This PR reduces per-tick overhead in train spawning by removing
temporary allocations and reducing work in the
destination-selection path.
The change focuses on `Cluster` trade destination lookup and how
`TrainStationExecution` picks a destination.
## What changed
### 1) Maintain a “trade-capable” station subset per cluster
`src/core/game/TrainStation.ts`
- `Cluster` now maintains:
- `stations`: all stations in the cluster (unchanged)
- `tradeStations`: maintained subset of stations that can act as trade
endpoints (`City` or `Port`)
- `tradeStations` is kept in sync in:
- `addStation()`
- `removeStation()`
- `clear()`
Impact:
- Trade queries no longer scan every station in the cluster; they only
scan `tradeStations`.
### 2) Add cheap eligibility helpers
`src/core/game/TrainStation.ts`
- `hasAnyTradeDestination(player)`:
- Fast early-exit check: returns as soon as it finds any eligible trade
destination.
- `randomTradeDestination(player, random)`:
- Picks a random eligible trade destination directly without
materializing an intermediate `Set`.
### 3) Use reservoir sampling for single-pass random choice
`src/core/game/TrainStation.ts`
`Cluster.randomTradeDestination()` uses reservoir sampling:
- Iterates `tradeStations` once.
- Maintains a running count of eligible stations (`eligibleSeen`).
- Replaces the selected station with probability `1/eligibleSeen`.
Properties:
- Uniform selection among eligible stations.
- One pass instead of “count then pick by index” (two pass).
- Allocation-free.
- Returns `null` when no eligible destination exists.
### 4) Update train spawning to avoid temporary sets
`src/core/execution/TrainStationExecution.ts`
- Previously: `spawnTrain()` called `cluster.availableForTrade()` and
then `random.randFromSet(...)`.
- This built a new `Set` on the hot path.
- Now:
- Early-exit via `cluster.hasAnyTradeDestination(owner)`.
- Destination via `cluster.randomTradeDestination(owner, random)`.
Net effect:
- Less per-tick work and no per-spawn temporary `Set` allocations.
## Why this helps
Train spawning happens frequently and can become a hot path in large
games / large rail clusters.
Avoiding repeated allocations and reducing work inside `tick()` helps
keep frame/update time predictable.
## notes
- Trade rules are unchanged (`tradeAvailable(player)` still gates
eligibility).
- Destination selection remains random-uniform over eligible
`City`/`Port` stations that satisfy `tradeAvailable(player)`.
- `TrainStationExecution` now avoids calling `spawnTrain()` entirely
when `spawnTrains` is falsy (it was already guarded inside).
## Please complete the following:
- [ ] I have added screenshots for all UI updates
- [ ] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [ ] I have added relevant tests to the test directory
- [ ] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
DISCORD_USERNAME
## Description:
This PR reduces server/client tick CPU spent on territory cluster
maintenance by:
- Cutting redundant work in `PlayerExecution.removeClusters()`
(largest-cluster bounding box reuse, fewer allocations in
`isSurrounded()` and `removeCluster()`).
- Making `calculateBoundingBox()` allocation-free per tile by switching
from `gm.cell(tile)` to `gm.x(tile)`/`gm.y(tile)` (it now only allocates
the two result `Cell`s, instead of 1 per tile).
## Commits
- `51de0a1b` core: reduce PlayerExecution cluster overhead
- `6d9d85c5` core: avoid Cell allocations in PlayerExecution
isSurrounded
- `346f6a8c` core(util): speed up calculateBoundingBox by avoiding Cell
allocations
## Notes
- This PR is intended to be behavior-preserving; changes are limited to
hot-path micro-optimizations.
- Follow-up opportunity: `calculateClusters`/flood-fill is now the top
hotspot; further wins likely come from reducing traversal work or
caching.
## Please complete the following:
- [ ] I have added screenshots for all UI updates
- [ ] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [ ] I have added relevant tests to the test directory
- [ ] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
DISCORD_USERNAME
## Description:
Explanation of the bot farming strategy in the discord:
https://discord.com/channels/1359946986937258015/1359949371956789289/1460928540575928478
"the result is that a player can build unlimited factories for 125 000
gold discount, trade with themselves with each train being worth 50 000
gold. First the 25 000 for neutral trade and then another 25 000 when
the bot is harvested."
"If you have a minute and ally people around you it should be trivial to
get 10 of both cities and factories for 1.25mil"
It's debatable if we want to let people do that (close this PR) or see
it as an abusive mechanic.
Here is the fix, bots try to delete all structures now. You can simply
retake them to stop the deletion:
https://github.com/user-attachments/assets/ac1ca846-50bd-42fa-8e25-5ac25a6d627e
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
Co-authored-by: Ryan <7389646+ryanbarlow97@users.noreply.github.com>
## Description:
- Maybe greet nearby players with a 👋 emoji in the earlygame (make
nations feel a little bit more lively)
- Destroying a transport ship / capturing a trade ship of a nation now
updates the relation (mainly on higher difficulties)
- On hard difficulty: Only lift embargos if we have a friendly relation.
On impossible difficulty: Don't lift embargos
- Improve totalPlayers calculation and return value of
`hasTooManyAlliances` in `NationAllianceBehavior`
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
Troop count display for naval invasion message:
<img width="398" height="131" alt="Screenshot 2026-01-28 204213"
src="https://github.com/user-attachments/assets/d7ccf2a8-9974-4f12-9901-e603426a8e56"
/>
On hover, PlayerInfoOverlay shows the troop count now:
<img width="504" height="99" alt="Screenshot 2026-01-28 202916"
src="https://github.com/user-attachments/assets/46f7685f-8c0e-4156-8d02-8a68dbcffde0"
/>
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
For v29, balances the HvN winrate.
In team games, nations now donate troops to their weakest team members
(if they have no attack options available).
How often they donate depends on the difficulty.
This PR also has some other little fixes:
- For HvN games, always return true in `shouldAttack()` (make nations a
bit more aggressive).
- Early exit in `attackWithRandomBoat()` for performance
- Early exit in `findNearestIslandEnemy()` for performance AND to make
sure nations which are encircled by friends don't run into this method
(=> no donation happening!)
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
For v29
HvN winrate is between 10 and 15%, but should be around 50%.
1. Change HvN difficulty to Medium
2. Little balance change in `NationAllianceBehavior`
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
### Short path for multi-source HPA*
Math was not mathing, increased the bounds to 260x260, it is a bit
slower but should work better. The short path was breaking when player
owned a lot of shores. This is because the bounding box of tiles with
less than 120 distance + 10 padding could be as big as 260x260 and the
optimized array was set to 140x140. I made mistake of calculating it as
`2 * (60 + 10)` instead of `2 * (120 + 10)`.
### LoS path refinement
Previously, we ran 2 passes of LoS smoothing on the path. However, since
we are effectively tracing the same path, the line of sight is
essentially the same. This PR makes second line of sight stop on water
tiles with magnitude `n + 1` compared to first path. Practically, this
means it'll attempt LoS exactly 1 tile after previous corner. See
screenshot.
<img width="1299" height="1151" alt="image"
src="https://github.com/user-attachments/assets/726be236-1ff8-406c-896a-02902a762ab0"
/>
### SendBoatAttackIntentEvent
The flow of sending transport ships is currently strange. This PR makes
the flow more sane.
**Old flow**
```
- Player clicks TARGET tile, it can be deep inland
- Client asks Worker for the best START tile to TARGET tile
- Worker answers `false`, since the tile is inland
- Client sends BoatAttackIntent with START=false and TARGET tiles set
- Worker accepts BoatAttackIntent, computes DESTINATION as closest shore to TARGET
- Worker re-computes best START to DESTINATION
- Worker sends boat from START to DESTINATION
```
**New flow**
```
- Player clicks TARGET tile, it can be deep inland
- Client sends BoatAttackIntent with TARGET
- Worker accepts BoatAttackIntent, computes DESTINATION as closest shore to TARGET
- Worker computes START as the best tile to DESTINATION
- Worker sends boat from START to DESTINATION
```
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
moleole
## Description:
"You can pick your spawn in random spawn games in v29. You need to open
the menu and click on the attack button. That's it."
Thats the fix for this problem.
Radial menu no longer allows to attack (pick a spawn) while random spawn
is enabled.
And SpawnExecution got a check so you cannot send malicious intents.
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
Nuke trajectory prediction now will show interception with allied SAMs
if the alliance will break on nuke launch.
Code was also refactored to be shared a bit more.
In addition, if an incoming alliance would break if accepted, the nuke
launch will break the alliance.
<img width="1199" height="1002" alt="nukepr"
src="https://github.com/user-attachments/assets/c31066d9-66cf-4eaa-be3c-e2fbcfe7965a"
/>
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
bibizu
## Description:
When a player is conquered (has less than 100 tiles left) their gold is
transfered to the conqueror. And after that the conqueror gets the last
tiles. But if some of those last tiles are not bordered by the
conqueror, they are given to their neighbour player. However that
neighbour player can be an ally. It is percieved as a bug if an ally
conquers/annexes tiles.
This PR fixes that by adding an isFriendly check to `handleDeadDefender`
in `AttackExecution`.
Now, there are already scenarios possible currently, where a player
survives being conquered. If they have some tiles on a small island for
example. Going from that, there should be no unexpected bugs following
this change. A player can be conquered twice in a game already in the
stats too.
https://discord.com/channels/1359946986937258015/1359946989046989063/1462595261204533248
Example of this happening in an Enzo vid (with his surprised reaction)
and explanation posted here:
https://discord.com/channels/1359946986937258015/1359946989046989063/1460483209308536925
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
tryout33
Resolves#2484
## Description:
- When an alliance is created between two players, any incoming nukes
between them are destroyed mid-air.
This prevents the traitor debuff from being applied on impact, even if
the nukes were launched before the alliance was formed.
- If a player has launched nukes at multiple nations, only the nukes
targeting the newly allied nation are destroyed.
This is what the players will see after the alliance is created (in case
they have launched nukes at each other):
<img width="423" height="125" alt="Screenshot 2026-01-04 092907"
src="https://github.com/user-attachments/assets/6544fb7a-7623-4fc3-b799-89ef8fe897d6"
/>
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
assessin.
## Description:
The recent pathfinding rework broke the naval invasion target.
This change reverts the code to the previous working one.
<img width="232" height="273" alt="image"
src="https://github.com/user-attachments/assets/0cc3ed80-bd77-4e7b-886b-0ce5012840ac"
/>
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
IngloriousTom
## Description:
Bots always attacked Terra Nullius if they shared a border with Terra
Nullius.
But water is Terra Nullius...
So I changed that condition to `this.bot.neighbors().some((n) =>
!n.isPlayer())`.
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
* Add RankedType enum, for now it's just 1v1
* Add new method to MapPlaylist to create 1v1 game config
* Update WinCheck so the last player is declared a winner on 1v1.
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
evan
## Description:
Doesn't need a description :D
https://github.com/user-attachments/assets/8de576fd-050b-4b35-8526-e4c88d1a9f25https://github.com/user-attachments/assets/c99147a1-efdf-426b-96d1-e996e01f89aa
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Playtest
https://pf-pt-2.openfront.dev/
## Pathfinding Refactor pt. 2
<img width="1536" height="1024" alt="image"
src="https://github.com/user-attachments/assets/9477958e-54b7-4c83-b317-ba789e809e9e"
/>
This is a follow-up to a previous PR introducing pathfinding changes.
This time, it introduces a complete refactor of `pathfinding` directory
and breakdown into composable pieces.
### Unified PathFinder interface
`PathFinder<T>` and `SteppingPathFinder<T>` are introduced to unify
**all** pathfinding across the application. First one exposes complete
path, while stepping variant allows the callee to iterate over the path
by calling `.next`. All pathfinders share this one common interface,
which makes them easy to use in any scenario -
`PathFinding.Water(game).search(from, to)`.
`SteppingPathFinder<T>` extends `PathFinder<T>` with an ability to
iterate over the path. It handles caching, storing current index and
invalidation. This allows the units to not care about the inner workings
of the pathfinder and just call `pf.next(current, target)` and receive
instructions on what to do next.
### Common entry point
All pathfinders are now exposed from common `PathFinding` entrypoint:
- `PathFinding.Water`
- `PathFinding.Rail`
- `PathFinding.Stations`
- `PathFinding.Rail`
Additional entry point is introduced for pathfinders which need to work
both in the worker, but also on the frontend, which lacks `Game`
interface. Currently only `UniversalPathFinding.Parabola` is available.
### Spatial Query
New module has been introduced close to `pathfinding` - `SpatialQuery`.
It aims to resolve any questions game may have about finding tiles
meeting criteria. Currently `SpatialQuery.closestShore(player, target)`
and `SpatialQuery.closestShoreByWater(player, target)` are available -
they help answering questions about naval invasion: "What is the best
landing location from user's click?" and "Which our tile should be used
to launch the transport ship?". Under the hood they use very similar
mechanics to pathfinding, so it felt right to put them close by.
### Modular architecture
Pathfinders now support transformers: `MiniMapTransformer`,
`ShoreCoercingTransformer`, `ComponentCheckTransformer`,
`SmoothingTransformer`. Transformers functions like a middleware in the
pathfinding chain. They wrap around the pathfinder and provide
additional functionality. This allows the pathfinder to focus on
actually finding the path instead of doing unrelated things.
Example chain for simple (A*) water pathfinding:
```ts
static WaterSimple(game: Game): SteppingPathFinder<TileRef> {
const miniMap = game.miniMap();
const pf = new AStarWater(miniMap);
return PathFinderBuilder.create(pf)
.wrap((pf) => new ShoreCoercingTransformer(pf, miniMap))
.wrap((pf) => new MiniMapTransformer(pf, game.map(), miniMap))
.buildWithStepper(tileStepperConfig(game));
}
```
The Pathfinder - here `AStarWater` - does not care about the conversion
between minimap and main map tiles. It also does not care if the source
or destination is a land tile. The transformers take care of that. The
pathfinder gets a set of valid coordinates and produces the path -
that's it.
Modular approach makes working on a particular set of utilities much
easier - for example map upscaling is handled consistently across all
pathfinders. Additionally, the pathfinders are not tied to the
particular map resolution used. Pass them a different map and they will
work the same.
### Algorithms
Algorithms used are neatly organized inside
`src/core/pathfinding/algorithms`. They are prefixed with the algorithm
name and suffixed with the use case. File without suffix exposes generic
version ready to traverse any graph with adapters. Specialized versions
either use an adapter or inline logic when performance is critical -
using adapters leads to 20-30% performance loss.
The directory includes `A*` and `BFS` but also other useful utils, such
as `AbstractGraph` used to generate... an abstract graph on top of the
tile map and `ConnectedComponents` helping to identify whether two tiles
are connected by a path without actually computing the path.
### Playground
The playground have been updated with new algorithms, including tweaked
very greedy `A*`.
<img width="2175" height="1424" alt="image"
src="https://github.com/user-attachments/assets/1f833651-0024-4299-bf86-882f5368358c"
/>
### Tests
Yeah, there are some, a little too many if I say so myself. But there
are no useless tests. I had to ensure refactored code works somehow
reliably. This PR comes with trust me bro guarantee, but I would
appreciate someone confirming **naval invasions, nukes (esp. MIRV) and
warships**.
### Discord
`moleole`
GL & HF
If this PR fixes an issue, link it below. If not, delete these two
lines.
Resolves#1225
## Description:
Replace all raw displayMessage strings with events_display.* keys +
params and add the new English translations in resources/
lang/en.json so EventsDisplay can translate them.
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
aotumuri
---------
Co-authored-by: Ryan <7389646+ryanbarlow97@users.noreply.github.com>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
## Description:
### `NationAllianceBehavior`
- `isAlliancePartnerSimilarlyStrong()` now also checks for
`numTilesOwned`, should make it a bit easier to get alliances. The troop
calculation now also uses the players `outgoingAttacks` to make it feel
less random. OF-Discord-Humans complained.
- Rebalanced `checkAlreadyEnoughAlliances()` a bit
### `NationNukeBehavior`
- Don't save up for MIRV if they are disabled
- Don't try to throw atom bombs / hydros if they are disabled
- Hydro-Nations are allowed to throw atom bombs if they are under heavy
attack
- Rebalance `isUnderHeavyAttack()` a bit
- Increased perceived cost
### `NationEmojiBehavior`
- Fix multiple nations congratulated the winner instead of one
- Don't brag with our crown if the game is already over
### `DonateGoldExecution` & `DonateTroopExecution`
- Added `canSendEmoji` checks
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
My first PR about the nation build order.
This one is a bit important for HumansVsNations.
- Nations build more SAMs in team games
- Nations build less factories if they have access to the ocean (instead
of focusing ports and factories with the same priority)
- Nations no longer place defense posts "without reason" - only when
they share a border with someone they haven't allied
I'm planning to make the build order a bit different based on
difficulty.
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
Issue discovered by @DevelopingTom, posted on Discord. After merging
pathfinding PR, transport ships lost the ability to navigate between
land tiles. For now, the quick fix is to select adjacent water tile and
select it for pathing. Conquer logic still applies to original
destination on the shore.
## Please complete the following:
- [ ] I have added screenshots for all UI updates
- [ ] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [ ] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
moleole
## Description:
### Refactor
- Moved `maybeSpawnWarship()` from `NationExecution` to
`NationWarshipBehavior`
- Moved `maybeAttack()` (and sub-methods) from `NationExecution` to
`AiAttackBehavior`
### Betrayal
- Added nice betrayal logic in `maybeBetray()`. Previously that method
was basically just a placeholder for a future implementation.
### Attacking
- Added `veryWeak()` attack strategy for hard and impossible difficulty
nations attack orders to target MIRVed players with higher priority
- Optimized the `weakest()` attack strategy so that nations don't attack
stronger players. This should make nation-attacks feel less random
(humans complained in discord)
- `findNearestIslandEnemy()` and `randomBoatTarget()` also no longer
returns stronger players
- `afk()` and `hated()` attack strategies no longer return MUCH stronger
players
- Several tiny refactorings, fixes and balance optimizations in
`AiAttackBehavior`
### Emojis
- Added some `canSendEmoji()` because I saw some "cannot send emoji"
warnings in the console
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
## Description:
Some FX animations were previously used as UI elements (e.g. nuke area,
naval invasion target, gold text).
This PR moves those animations to a dedicated UI layer.
Those UI elements handle correctly the current zoom level and remain
sharply rendered at all zoom levels.
The new UI layer can be disabled using the same setting that disables
the FX layer.
Performance-wise, this layer is equivalent to the FX layer, as it reuses
the same animations.
### Naval target
Don't scale with the zoom level, but has a minimum zoom level so the
targeted tile can still be easily highlighted by zooming
### Nukes
Has to scale because the size is set, but the border radius is not so
the area is more visible from afar.
### Popup text
Scale with zoom level, and stop showing when zoomed-out:
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
IngloriousTom
## Pathfinding with HPA*
Hi! The primary objective of this PR is to replace per-tile A* with
hierarchical pathfinding - HPA*. In practice, this means we create an
abstract graph on top of the actual map with far fewer points and use it
to decide on general path structure. Only then we go back to tile-level
and build path between selected waypoints. This speeds up long distance
pathfinding by over 1000x in some cases. To make the review easier, it
comes with a benchmark and visual playground.
## PREPROCESSING
H part of HPA* means "hierarchical" and requires preprocessing.
This PR includes pre-processing as part inside `new Game()` constructor.
It takes about 135ms for `giantworldmap` on my machine, which increases
the effective initialization from ~95ms to ~230ms. This time could be
reduced in different ways, which are **out of scope** for this PR.
After confirming the initialization time is bearable on low-end devices,
I argue merging this PR as-is is acceptable tradeoff. It creates small
lag at the beginning of a round but pays for itself in the first minute
of the match.
## Nerdy details
**Architecture**
- HPA*-style hierarchical pathfinding
- 32×32 sectors on minimap with gateway nodes on borders
- Gateway graph built via BFS during preprocessing
- Water component optimization skips unreachable gateway pairs
- A* on gateway graph → local A* within sectors → Bresenham path
smoothing
- Minimap upscaling identical to currently used in MiniAStar
**Key Optimizations**
- Typed arrays instead of high-level primitives
- Stamp-based visited tracking (no need to recreate buffers, O(1)
clearing)
- Optional - enabled by default - caching of tile paths between gateways
- Line of sight smoothing for the final path
## Review Focus
Play with included tools, benchmark and visualization. Pathfinding
should be safe to merge as a black box - you do not need to understand
the details. Outcomes can be tested empirically in-game. Visualize (and
share!) edge cases with included playground. Confirm the 100x speedup is
real with benchmark.
If you plan to dive into the code, I suggest the following order:
- Pathfinding abstraction in `src/core/pathfinding/`
- Pathfinding tests in `tests/core/pathfinding/`
- NavMesh in `src/core/pathfinding/navmesh/` + integration with
`Game.ts`
- Benchmark in `tests/pathfinding/benchmark/`
Do not look at playground's code, it has been created with a clanker.
The design is 100% mine and I spent way too long polishing it, but I
haven't even once edited the code manually. There is probably no
abstraction whatsoever, just do not look at the code, let it play.
## Core Changes
#### Pathfinding (`src/core/pathfinding/navmesh/`)
- HPA* + refinement -> three phased pathfinding: A* over the graph ->
naive path -> refinement
- comes with A* and BFS optimized for for specific needs
#### Pre-Processing (`src/core/pathfinding/navmesh/`)
- identify water bodies to avoid pathfinding between disconnected nodes
- create high-level graph of gateways on top of tile map
#### Abstraction (`src/core/pathfinding/`)
- common `PathFinder` interface that can return full path and also act
as state machine (`.next()`)
- adapters for both new and legacy algorithm with fallback to legacy if
navigation mesh not available
#### Benchmark (`tests/pathfinding/benchmark/`)
- `npx tsx tests/pathfinding/benchmark/run.ts` - no guesswork, numbers
- `npx tsx tests/pathfinding/benchmark/run.ts --synthetic` - 1000s of
synthetic paths
- `npx tsc tests/pathfinding/benchmark/generate.ts` - generate more as
needed, test new maps
- includes ONE synthetic scenario to avoid PR bloat, generate more
locally / later
#### Playground (`tests/pathfinding/playground/`)
- `npx tsx tests/pathfinding/playground/server.ts` - visualize paths
with both new and legacy algorithm
## Benchmarks
### Compared with legacy in default - hand picked - scenario:
```
Initialization: 95.95ms -> 227.29ms
Pathfinding: 3038.43ms -> 6.45ms
Distance: 26972 -> 26810 tiles
```
### 42,000 synthetic routes across all maps
```
Running 42 synthetic scenarios with hpa.cached adapter...
✅ synthetic/achiran | Init: 93.42ms | Path: 139.07ms | Dist: 1481630 tiles | Routes: 1000/1000
✅ synthetic/africa | Init: 87.14ms | Path: 155.08ms | Dist: 1829414 tiles | Routes: 1000/1000
✅ synthetic/asia | Init: 57.60ms | Path: 112.55ms | Dist: 1204082 tiles | Routes: 1000/1000
✅ synthetic/australia | Init: 78.18ms | Path: 77.12ms | Dist: 978375 tiles | Routes: 1000/1000
✅ synthetic/baikal | Init: 78.26ms | Path: 152.14ms | Dist: 1600016 tiles | Routes: 1000/1000
✅ synthetic/baikalnukewars | Init: 81.44ms | Path: 165.90ms | Dist: 1699283 tiles | Routes: 1000/1000
✅ synthetic/betweentwoseas | Init: 29.29ms | Path: 114.99ms | Dist: 1338075 tiles | Routes: 1000/1000
✅ synthetic/blacksea | Init: 30.66ms | Path: 93.14ms | Dist: 949217 tiles | Routes: 1000/1000
✅ synthetic/britannia | Init: 74.12ms | Path: 85.62ms | Dist: 866752 tiles | Routes: 1000/1000
✅ synthetic/deglaciatedantarctica | Init: 105.49ms | Path: 192.93ms | Dist: 1574684 tiles | Routes: 1000/1000
✅ synthetic/didier | Init: 81.51ms | Path: 153.70ms | Dist: 1734876 tiles | Routes: 1000/1000
✅ synthetic/eastasia | Init: 49.29ms | Path: 128.63ms | Dist: 1410270 tiles | Routes: 1000/1000
✅ synthetic/europe | Init: 92.55ms | Path: 178.35ms | Dist: 1525216 tiles | Routes: 1000/1000
✅ synthetic/europeclassic | Init: 33.50ms | Path: 104.40ms | Dist: 1209759 tiles | Routes: 1000/1000
✅ synthetic/falklandislands | Init: 63.00ms | Path: 107.41ms | Dist: 1080251 tiles | Routes: 1000/1000
✅ synthetic/faroeislands | Init: 71.91ms | Path: 49.52ms | Dist: 604613 tiles | Routes: 1000/1000
✅ synthetic/fourislands | Init: 45.75ms | Path: 78.91ms | Dist: 937439 tiles | Routes: 1000/1000
✅ synthetic/gatewaytotheatlantic | Init: 81.00ms | Path: 257.06ms | Dist: 2555551 tiles | Routes: 1000/1000
✅ synthetic/giantworldmap | Init: 214.25ms | Path: 220.42ms | Dist: 1976693 tiles | Routes: 1000/1000
✅ synthetic/gulfofstlawrence | Init: 45.16ms | Path: 96.05ms | Dist: 1014604 tiles | Routes: 1000/1000
✅ synthetic/halkidiki | Init: 74.68ms | Path: 149.39ms | Dist: 1546781 tiles | Routes: 1000/1000
✅ synthetic/iceland | Init: 58.72ms | Path: 78.16ms | Dist: 1001554 tiles | Routes: 1000/1000
✅ synthetic/italia | Init: 29.78ms | Path: 139.93ms | Dist: 1412024 tiles | Routes: 1000/1000
✅ synthetic/japan | Init: 161.07ms | Path: 118.65ms | Dist: 1154393 tiles | Routes: 1000/1000
✅ synthetic/lemnos | Init: 52.59ms | Path: 136.69ms | Dist: 1481101 tiles | Routes: 1000/1000
✅ synthetic/lisbon | Init: 49.27ms | Path: 86.53ms | Dist: 1032011 tiles | Routes: 1000/1000
✅ synthetic/manicouagan | Init: 53.74ms | Path: 110.52ms | Dist: 1307630 tiles | Routes: 1000/1000
✅ synthetic/mars | Init: 29.39ms | Path: 80.55ms | Dist: 1091702 tiles | Routes: 1000/1000
✅ synthetic/mena | Init: 26.37ms | Path: 120.09ms | Dist: 1272751 tiles | Routes: 1000/1000
✅ synthetic/montreal | Init: 26.08ms | Path: 106.77ms | Dist: 1187736 tiles | Routes: 1000/1000
✅ synthetic/newyorkcity | Init: 56.60ms | Path: 181.19ms | Dist: 1753875 tiles | Routes: 1000/1000
✅ synthetic/northamerica | Init: 96.29ms | Path: 123.02ms | Dist: 1217221 tiles | Routes: 1000/1000
✅ synthetic/oceania | Init: 52.81ms | Path: 51.96ms | Dist: 482373 tiles | Routes: 1000/1000
✅ synthetic/pangaea | Init: 21.29ms | Path: 56.58ms | Dist: 716189 tiles | Routes: 1000/1000
✅ synthetic/pluto | Init: 53.89ms | Path: 141.62ms | Dist: 1304362 tiles | Routes: 1000/1000
✅ synthetic/southamerica | Init: 85.19ms | Path: 123.03ms | Dist: 1301403 tiles | Routes: 1000/1000
✅ synthetic/straitofgibraltar | Init: 76.68ms | Path: 108.30ms | Dist: 1304592 tiles | Routes: 1000/1000
✅ synthetic/straitofhormuz | Init: 38.97ms | Path: 67.78ms | Dist: 754920 tiles | Routes: 1000/1000
✅ synthetic/surrounded | Init: 95.35ms | Path: 90.18ms | Dist: 1017142 tiles | Routes: 1000/1000
✅ synthetic/svalmel | Init: 60.58ms | Path: 104.75ms | Dist: 1235501 tiles | Routes: 1000/1000
✅ synthetic/twolakes | Init: 62.05ms | Path: 94.54ms | Dist: 1140807 tiles | Routes: 1000/1000
✅ synthetic/world | Init: 41.43ms | Path: 93.42ms | Dist: 873406 tiles | Routes: 1000/1000
Completed 42 scenarios
Total Initialization Time: 2796.32ms
Total Pathfinding Time: 5026.64ms
Total Distance: 53160274 tiles
```
## Playground
**That's the fun part**. Watch NavMesh running circles around legacy
`PathFinder.Mini` in real time. Debug inner workings, test edge cases,
share URLs for debugging.
https://github.com/user-attachments/assets/34e2e3f5-fbc1-4b1f-917d-820766e98d5d
## Discord Tag
`moleole`
## Description:
The new (awesome) nation emoji updates had a small bug in them when I
was playtesting with a friend where donating troops to them (a human
player) would result in the player automatically sending an emoji reply.
Sometimes these replies were negative-connotations like ❓ and 🥱, which
could impact how other players perceive their donation attempt. This PR
fixes that issue.
### Example of player nation sending emojis automatically
https://github.com/user-attachments/assets/99689966-b784-4c3f-b43b-953a4a102e2d
### Donating to player after fix
https://github.com/user-attachments/assets/ace0c1ee-3eb8-4240-9c78-167dd773cfb2
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
bijx
## Description:
### Refactor
- Moved `findBestNukeTarget()` (and child methods) from
`AiAttackBehavior` to `NationNukeBehavior` because it makes more sense.
- `NationNukeBehavior`: Renamed `mg` to `game` (So its similar to the
other nation .ts files)
- Moved the `removeOldNukeEvents()` method a bit
### New features
- Impossible difficulty nations are now optimized for SAM outranging!
- 1 of 3 nations is now a hydro-nation. They cannot send atom bombs.
They save up for hydros. This is to reduce atom-bomb-spam on the map.
- On impossible difficulty, the crown nukes the second place now (in
FFAs)
- On hard and impossible difficulty, nations now ignore the perceived
cost for nukes if they get heavily attacked. Reasoning: They should stop
saving for MIRV, they should defend with all their gold!
- On medium, hard and impossible difficulty, nations no longer throw
nukes at places where another team member already has a nuke "in the
flying process"
- Optimized `lastNukeSent` a bit (to respect nuke radius)
- Adjusted `maybeSendNuke()` to use 30 instead of 10 randomTiles on
impossible difficulty to improve chances of finding a perfect SAM
outranging spot
- On impossible difficulty, nations now ignore their "most hated enemy"
if the crown has 50%+ of the map (Very big danger). Instead they are
nuking the crown.
- Added `isFriendly` check to `findStrongestTeamTarget()`
### Media
SAM outranging:
https://github.com/user-attachments/assets/d1e88bb6-0060-400b-9c16-24d7399f5949
Team game nukes not hitting the same spot:
<img width="1160" height="708" alt="Screenshot 2026-01-03 042236"
src="https://github.com/user-attachments/assets/c017fb3c-3e3f-45fb-9d45-dd4caba7a59f"
/>
## Please complete the following:
- [X] I have added screenshots for all UI updates
- [X] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [X] I have added relevant tests to the test directory
- [X] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
FloPinguin
---------
Co-authored-by: iamlewis <lewismmmm@gmail.com>
## Description:
Resolve discussions about stalled PR
https://github.com/openfrontio/OpenFrontIO/pull/2460
<img width="724" height="348" alt="image"
src="https://github.com/user-attachments/assets/c2c9fa79-cace-431a-9ca4-b3656612fa9d"
/>
Changes:
- Added a `Player::canAttackPlayer(other)` function to determine whether
a player can be attacked.
- This function is now used in most places where a fight can occur:
- AttackExecution (land attacks)
- Naval invasion
- Warship fight
- Nukes can't be thrown during the truce
- Immunity only affect human players. Nations and bot will fight as
usual, and can be fought against.
- The immunity timer uses minutes in the modal window.
UI:
- The immunity phase is displayed with a timer bar at the top. This is
from the original PR, to be discussed if it's not deemed visible enough:
<img width="632" height="215" alt="image"
src="https://github.com/user-attachments/assets/f5ab9aa0-bd4f-4503-b8d6-b40b121fba65"
/>
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [x] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
IngloriousTom
---------
Co-authored-by: newyearnewphil <git@nynp.dev>
## Description:
Trains are made of a primary unit (`TrainExecution.train`) followed by 6
cars (`TrainExecution.cars`). Currently when any of the units is
destroyed, a message "Your Train was destroyed" is shown. In worst case
scenario, when all cars are blasted by a nuke, 7 messages are displayed
to the user, but the train continues.
Since the actual logic is unaffected as long as the primary unit stays
alive, displaying messages is confusing. This PR fixes it. The message
is only displayed when the primary unit is destroyed. The following cars
are only there for fancy visuals.
## Screencast
##### Current - multiple messages for single train
https://github.com/user-attachments/assets/3df04c71-d899-4f68-af83-36c9d0ffa730
First nuke was a test.
Second nuke destroyed the entire train, prompting 7 messages.
Third nuke destroyed one full train and then some. Prompting many too
many messages.
##### Fixed - one message, only if train was fully destroyed
https://github.com/user-attachments/assets/1f3840a7-6c62-487d-af3a-82de39dad9e8
First train was only partially destroyed, no message was shown, delivery
mission completed A+.
Following 2 trains had the front engine destroyed and therefore were
promptly decommissioned.
## Please complete the following:
- [x] I have added screenshots for all UI updates
- [ ] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [ ] I have added relevant tests to the test directory
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
## Please put your Discord username so you can be contacted if a bug or
regression is found:
moleole
## Description:
`MIRVExecution.separate` is consuming more resources than it needs to.
This PR introduces a series of minor performance changes which do not
alter the behavior of the selection algorithm. Assessing the code
initially, I was convinced there are multiple easy wins - starting with
the proximity check. However, after many hours of mostly math, no
alternative solution came close to the speed of current implementation.
Therefore this PR consists of only a few minor tweaks:
#### Removed `console.log`
This is by far the worst offender, in my test removing the three lines
of console log improved execution time from ~30ms down to ~10ms. The
logs are not very useful. I do not see a clear pattern in the logs
produced by the application therefore they were completely removed for
now. If there is a need for the log in production build, I suggest
adding single line with the number of destinations selected.
```diff
- console.log(`dsts: ${dsts.length}`);
- console.log(`got ${dsts.length} dsts!!`);
- console.log("couldn't find place, giving up");
```
#### Replaced multiple calls to `random.nextInt` with single call to
`random.next`
The flamechart shows calling pseudo random number generator is
expensive. Therefore instead of calling it twice, the code now generates
a single random number and derives further calculations from it. It
remains 100% deterministic and there should not be any noticeable change
to the enthrophy. This saves about ~1.5ms in my tests.
```diff
- const x = this.random.nextInt(
- this.mg.x(ref) - this.mirvRange,
- this.mg.x(ref) + this.mirvRange,
- );
- const y = this.random.nextInt(
- this.mg.y(ref) - this.mirvRange,
- this.mg.y(ref) + this.mirvRange,
- );
+ const r1 = this.random.next();
+ const r2 = (r1 * 15485863) % 1;
+ const x = Math.round(r1 * this.range * 2 - this.range + this.baseX);
+ const y = Math.round(r2 * this.range * 2 - this.range + this.baseY);
```
#### Caching of destination coordinates
Since the target tile coordinates are used a lot, instead of retrieving
them every time with `this.mg.x` and `this.mg.y`, they get cached as
`baseX` and `baseY`. To reduce usage further, I also exposed `x` and `y`
to `isOverlapping` / `proximityCheck` directly instead of passing the
tile. Since available methods operate on `TileRef`, this change requires
the calculations - manhattan and euclidean distance - to be inlined. I
do not think this is a big issue, considering this code is responsible
for very specific task. This saves another ~1.5ms in my tests.
```diff
- if (this.mg.euclideanDistSquared(tile, ref) > mirvRange2) {
+ if ((x - this.baseX) ** 2 + (y - this.baseY) ** 2 > this.rangeSquared) {
```
## Benchmark:
**Before**
```
=== MIRV Performance Benchmark Results ===
MIRV target selection - sparse territory x 53.53 ops/sec ±0.48% (71 runs sampled)
MIRV target selection - dense territory x 53.39 ops/sec ±0.57% (70 runs sampled)
MIRV target selection - giant world map (350 targets) x 1,129 ops/sec ±0.98% (90 runs sampled)
```
**After**
```
=== MIRV Performance Benchmark Results ===
MIRV target selection - sparse territory x 198 ops/sec ±0.39% (85 runs sampled)
MIRV target selection - dense territory x 200 ops/sec ±0.28% (86 runs sampled)
MIRV target selection - giant world map (350 targets) x 1,409 ops/sec ±0.89% (92 runs sampled)
```
## Please complete the following:
- [ ] I have added screenshots for all UI updates
- [ ] I process any text displayed to the user through translateText()
and I've added it to the en.json file
- [x] I have added relevant tests to the test directory*
- [x] I confirm I have thoroughly tested these changes and take full
responsibility for any bugs introduced
\* Tests in separate PR, implemented against master:
https://github.com/openfrontio/OpenFrontIO/pull/2767
## Please put your Discord username so you can be contacted if a bug or
regression is found:
moleole
Aotumuri