A serverconfig.env (or just environment variables) can be set to
override the default bind address/port for the backend. This is passed
to the front end as a query parameter.
There are still two more of these that don't show up in openapi.json
because they don't show up in HTTP routes (only in the websocket):
* GameUpdateEvents
* FrozenCombat
I'm not sure if there's a way to forcibly include those in the
openapi.json, if I should add a no-op API to force it to happen, or if I
should just ignore it. For now I'm going with option 3.
operation_ids give us better function names when generating the
typescript API from the openapi.json. BaseModel.Config.title does the
same for type names. Response models (or 204 status codes) need to be
explicit or the API will be declared as returning any.
The React UI running in a browser can't connect to the backend without
punching a hole for CORS, which isn't done by default. We don't need the
API key to protect from browsers, and anything else running on the
user's machine that can access the backend (that's hosted on only
localhost) already has enough control to do damage without using
Liberation as an attack vector.
https://github.com/dcs-liberation/dcs_liberation
This may not be the way to do this long term, but it is how the old map
works so it's at least not a regression. It might be better to generate
events for the between-turn changes in state instead.
https://github.com/dcs-liberation/dcs_liberation/issues/2039
Useful for development if you want to disable API key authentication for
debugging the server without having to pull the generated key out of the
log every time.
There are some TODOs here but th behavior is flagged off by default. The
biggest TODO here is that the time spent frozen is not simulated, so
flights that are engaged by SAMs will unfreeze, move slightly, then re-
freeze.
https://github.com/dcs-liberation/dcs_liberation/issues/1680
If there's a websocket being waited on the shut down won't actually
happen. Add a new event for shut down and send it to break the websocket
out of its loop.
We can't directly use frozen dataclasses from pydcs in our interface
because pydantic can't process them. Pydantic is able to automatically
convert to our modelview type from the pydcs type though.
This is briefly moving us over to my fork of pydcs while we wait for
https://github.com/pydcs/dcs/pull/206 to be merged. The adaptation is
invasive enough that I don't want it lingering for long.
Apparently it's a bad idea to try to make the core data pydantic models,
and those should really be treated more as a view-model. Doing otherwise
causes odd patterns (like the UI info I had leaked into the core type),
and makes it harder to interop with third-party types.
By default logging configuration is defined by
resources/default_logging.yaml. Very noisy loggers (like the
uvicorn.access logger) are kept out of the console and UI logs by
default. Developers (or weird users) can customize their default logging
config by copying the file to resources/logging.yaml and editing as
needed. It would be preferable to load this file form the Liberation
user directory, but because first-time initialization requires the UI,
we want to configure logging before we necessarily know where to find
that.
We don't have any sensitive data, but we do access the file system. On
the off chance that some phishing website decides to try to use
Liberation as an attack vector, prevent access to the API by
unauthorized applications. An API key is generated at each program start
and passed to the front end via the QWebChannel.
I have no idea why the docs think strings are better. pyinstaller wasn't
able to detect the import through the string so the code ended up being
stripped from the package. Use an explicit import to prevent that.
A possible explanation for the infrequent CTDs we've been seeing since
adding fast forward is that QWebChannel doesn't keep a reference to the
python objects that it passes to js, so if the object is GC'd before the
front end is done with it, it crashes.
We don't really like QWebChannel anyway, so this begins replacing that
with FastAPI.
