Interesting AI code gen experiment #70
fwaris
started this conversation in
Show and tell
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
I wanted to add charts to my app. Syncfusion has open sourced some controls now (https://github.com/syncfusion/maui-toolkit/) including charts.
I asked OpenAI Codex preview model (via github copilot) to generate the fabulous-maui bindings. It did! And the bindings seem to work (for the charts I tested). I think this is quite impressive and can save a lot of time/effort when creating new bindings.
In github copilot 'Agent' mode, the generated code is compile-checked and fixed - repeatedy, if needed.
To 'understand' the existing code (syncfusion controls, fabulous maui, etc.) Codex uses the ILSpy tool to decompile assemblies and ingest source.
Thus far other models (e.g. Sonnet 4.5) cannot seem to accomplish such tasks which require deeper understanding of F#/C# code.
See generated code below.
Beta Was this translation helpful? Give feedback.
All reactions