Three independently ugly bugs that all fed the same log flood:
* Electron's `WebFrameMain.send` writes every "Render frame was disposed"
to stderr via `console.error`. When the worker fans a second
agent-event past a reloading renderer, we produced thousands of
identical lines and stalled the main loop on stderr I/O. Added a
`mainFrame.detached` guard so the send is short-circuited before
Electron's own catch block fires, and installed a targeted filter in
`suppress-warnings.ts` (now actually imported from the entry point)
so the specific `console.error("Error sending from webFrameMain:"…)`
noise never reaches the terminal. Any other payload still surfaces.
* `dispatchAgentMessage` accepted the human-reader sentinel as a wake
target when an agent replied to `__board__` — `wakeAgentIfPossible`
then called `db.getAgent("__board__")`, threw, and the dispatcher
kept repeating on every send. Strip `HUMAN_SENDER_ID` at
`resolveMessageWakeTargets`, and keep a defense-in-depth guard in the
heartbeat so a stray caller (future automation fan-out) cannot
reintroduce the crash.
* Pi's skills loader rejects any SKILL.md whose frontmatter `name`
diverges from its parent directory. The four bundled `my-writing-style`
copies declared `name: writing-style-skill`, so every install emitted
a warning and dropped the skill. They also lived as orphan gitlinks
(mode 160000 without .gitmodules), and the global `*.md` gitignore
hid every other SKILL.md from version control. Fixed the frontmatter
to match the dir, converted the dangling submodules to plain trees,
and allowlisted `builtin-skills/**/*.md` so a fresh clone carries the
catalogue instead of silently losing it.
2.6 KiB
2.6 KiB
title, impact, impactDescription, tags
| title | impact | impactDescription | tags |
|---|---|---|---|
| Use Case Defines the Transaction Boundary | HIGH | ensures data consistency, prevents partial updates | usecase, transactions, consistency, boundaries |
Use Case Defines the Transaction Boundary
The use case is the natural transaction boundary. All operations within a use case should succeed or fail atomically. Infrastructure manages the transaction; use case defines the scope.
Incorrect (transaction managed inside repository):
class OrderRepository:
def save(self, order):
with self.db.transaction(): # Transaction per operation
self.db.insert('orders', order)
class InventoryRepository:
def reserve(self, items):
with self.db.transaction(): # Separate transaction
for item in items:
self.db.update('inventory', item.sku, decrement=item.qty)
class PlaceOrderUseCase:
def execute(self, command):
order = Order.create(command.items)
self.orders.save(order) # Commits
self.inventory.reserve(command.items) # If this fails, order is orphaned!
self.payments.charge(order.total) # If this fails, inventory is reserved but unpaid
Correct (use case defines transaction boundary):
# application/ports/UnitOfWork.py
class UnitOfWork(Protocol):
def begin(self) -> None: ...
def commit(self) -> None: ...
def rollback(self) -> None: ...
# application/usecases/PlaceOrderUseCase.py
class PlaceOrderUseCase:
def __init__(self, uow: UnitOfWork, orders, inventory, payments):
self.uow = uow
self.orders = orders
self.inventory = inventory
self.payments = payments
def execute(self, command):
self.uow.begin()
try:
order = Order.create(command.items)
self.inventory.reserve(command.items)
self.payments.charge(order.total)
self.orders.save(order)
self.uow.commit() # All or nothing
return order.id
except Exception as e:
self.uow.rollback() # Clean rollback
raise
# Alternative using context manager
class PlaceOrderUseCase:
def execute(self, command):
with self.uow: # Transaction spans entire use case
order = Order.create(command.items)
self.inventory.reserve(command.items)
self.payments.charge(order.total)
self.orders.save(order)
return order.id
Note: For distributed systems, consider saga patterns or eventual consistency rather than distributed transactions.
Reference: Unit of Work Pattern