What version of the IDE extension are you using?
0.4.71
What subscription do you have?
API
Which IDE are you using?
VS Code
What platform is your computer?
linux
What issue are you seeing?
When running long-running Python code, the Codex IDE version 0.4.71 starts to require polling to confirm whether the code has finished running, and it will also stop halfway, requiring my intervention.
The old version of Codex could perform parameter scanning tasks upon my command, and it could work continuously until all parameters were scanned, without requiring my manual intervention. (After manually reverting to an earlier version, I found that this was indeed the case.)
What steps can reproduce the bug?
The feeling is that it's too aggressive with asynchronous operations; the older versions were more inclined to proceed step by step
What is the expected behavior?
My expectation is to revert to the old version, which is very helpful for deep learning practitioners. The logic of the new version is somewhat similar to Claude code, but it is not very user-friendly
Additional information
No response
What version of the IDE extension are you using?
0.4.71
What subscription do you have?
API
Which IDE are you using?
VS Code
What platform is your computer?
linux
What issue are you seeing?
When running long-running Python code, the Codex IDE version 0.4.71 starts to require polling to confirm whether the code has finished running, and it will also stop halfway, requiring my intervention.
The old version of Codex could perform parameter scanning tasks upon my command, and it could work continuously until all parameters were scanned, without requiring my manual intervention. (After manually reverting to an earlier version, I found that this was indeed the case.)
What steps can reproduce the bug?
The feeling is that it's too aggressive with asynchronous operations; the older versions were more inclined to proceed step by step
What is the expected behavior?
My expectation is to revert to the old version, which is very helpful for deep learning practitioners. The logic of the new version is somewhat similar to Claude code, but it is not very user-friendly
Additional information
No response