You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Routing: Revise feature requests from the community. Add affordances to match multiple path segments? Routing to sub-applications? (See also this discussion.) ASGI mount #1950
REST: resource representation creation and consumption; url_for and perhaps also opt-in pydantic support?
Affordances to support API description format generators and frontends (e.g, OpenAPI, API Blueprint) -- check community project(s) already available for Falcon?
Recipes: Things like how to deploy in production, how to spawn background tasks, how to make use of ASGI lifecycle events, how to do AuthN/Z, and how to handle the DAL (SQLAlchemy, asyncpg, but perhaps we can also document piccolo, et al.). What else?
Performance: Cythonize generated router code #1664 -- also for ASGI explore if it is possible to exploit cythonization to get a slight performance bump. See also: perf: undo ASGI cythonization #1860 (comment) -- although cythonization offered a moderate improvement in isolated benchmarks, the actual ASGI servers seem to wrap the app in a way that degrades performance to worse than pure-Python.
4.0.0
CPython 3.11 support
Drop 3.5 & 3.6
Update benchmarks on falconframework.org
Docs: Additional recipes, make recipes easier to find
TBD: Include 1-2 doc improvements, and one "missing feature" such as url_for or basic JWT/OAuth2 affordances that can be expanded upon in future releases. Potentially also include trio support if decided to do it and is ready.
Performance: Basic ASGI perf gate
Features: Baseline JWT/OAuth2 affordances and recipes
Better typing (mypy) support
The text was updated successfully, but these errors were encountered:
Roadmap for 4.x
Themes
ASGI
Frequently requested "missing" features
Docs
Performance
4.0.0
4.1.0
TBD: Include 1-2 doc improvements, and one "missing feature" such as url_for or basic JWT/OAuth2 affordances that can be expanded upon in future releases. Potentially also include trio support if decided to do it and is ready.
The text was updated successfully, but these errors were encountered: