-
MIN_CHANCE
andMAX_FRONTIER_SIZE
test cases and proofs - verify
checkHit
commits are all correct (eg. Leech Seed / Substitute etc) -
transitions
called on everyTest
update
-
transitions
called within fuzz tests- empirically prove
MIN_CHANCE
andMAX_FRONTIER_SIZE
- empirically prove
- ensure all docs are updated to reflect
-Dchance
and-Dcalc
and new options patterns - update
DESIGN.md
- explain different engine types and tradeoffs
- explain what is a damage calc
- explain
-Dchance
&-Dcalc
design - explain
transitions
+ runtime analysis and provide tips on optimizing for use case
- update
gen{1,2}/README.md
-
-Dchance
and-Dcalc
covered -
transitions
function - RNG tables
-
- support
-Dchance
and-Dcalc
(de)serialization - efficient + ergonomic pattern for
Options
- efficient way to wrap library and produce an "uber-binary"
- unit test coverage for bindings
- proof of concept for using
-Dchance
/-Dcalc
to power a damage calc (@pkmn/dmg
)
- async API
-
@pkmn/engine
bundled into PocketMon running in browser- WASM bindings
- bundling with Vite working
- Ensure possible to write a wrapper of the JS package for 0 ERROR
- enable
.{ .advance = false, .ebc = false }
for engine and have it preserved across updates - write 0 ERROR worker manager logic and solve bundling/wire up with Web Workers
- benchmark running in browser in all workers
- enable
- optimize
transitions
function - ensure no regressions to benchmark from new features
- profile performance and optimize further
- Refactor
State
param out of Generation II implementation after all tests are complete - Verify all loads and stores of every varialble in
wram.asm
for Generation I & II
https://documentation.divio.com/