Streamlit Canary is here, a new approach to Streamlit development that eliminates full app reruns, making larger projects faster and more efficient.
If you've ever built a Streamlit app and watched it slow down as your project grew, then Streamlit Canary is the news you've been waiting for! Streamlit's performance can become a real challenge as projects get larger, but this new approach aims to change that significantly.
What does this mean for you? Simply put, your apps won't have to rerun entirely every time a user interacts with them! Instead, Streamlit Canary introduces an event-driven design, similar to a signal-slot pattern. This means your app function runs just once at startup, and after that, interactions only update the specific components that need changing. The result? A much smoother and faster user experience, with applications that feel snappier and handle complexity with ease as they grow.
But the improvements don't stop at performance. Streamlit Canary offers several exciting new features. You can now run your app in a native desktop window, backed by Toga WebView, complete with a window icon, title, and adjustable size.
This version also comes with a richer, built-in component library. Beyond Streamlit's standard components, you'll find practical additions like folder trees, a PDF viewer, a live log panel, and even floating toast messages. For developers, there's the 'dunder_literal' feature, which stops Markdown from accidentally turning names like `__init__` into bold text, keeping your code clear. Even better, Streamlit Canary significantly reduces its dependencies; it doesn't pull in heavy libraries such as NumPy, pandas, SciPy, pyarrow, or Altair, making your deployments lighter and potentially faster. This thoughtful design focuses on making your Streamlit projects perform well from the start, letting you concentrate on building great content rather than battling performance bottlenecks.
What does this mean for you? Simply put, your apps won't have to rerun entirely every time a user interacts with them! Instead, Streamlit Canary introduces an event-driven design, similar to a signal-slot pattern. This means your app function runs just once at startup, and after that, interactions only update the specific components that need changing. The result? A much smoother and faster user experience, with applications that feel snappier and handle complexity with ease as they grow.
But the improvements don't stop at performance. Streamlit Canary offers several exciting new features. You can now run your app in a native desktop window, backed by Toga WebView, complete with a window icon, title, and adjustable size.
This version also comes with a richer, built-in component library. Beyond Streamlit's standard components, you'll find practical additions like folder trees, a PDF viewer, a live log panel, and even floating toast messages. For developers, there's the 'dunder_literal' feature, which stops Markdown from accidentally turning names like `__init__` into bold text, keeping your code clear. Even better, Streamlit Canary significantly reduces its dependencies; it doesn't pull in heavy libraries such as NumPy, pandas, SciPy, pyarrow, or Altair, making your deployments lighter and potentially faster. This thoughtful design focuses on making your Streamlit projects perform well from the start, letting you concentrate on building great content rather than battling performance bottlenecks.
