Re: [Python-Dev] py3k, bad sys.path and abort()

2009-01-04 Thread Christian Heimes
Mark Hammond schrieb: > Would it be practical and desirable to handle this situation more > gracefully, possibly just leaving sys.std* set to None and letting > whatever exceptions then occur happen as normal without terminating the > process? Given it is an edge-case, I thought I'd open it here f

[Python-Dev] py3k, bad sys.path and abort()

2009-01-03 Thread Mark Hammond
I've recently noticed that in py3k, the lack of a suitable sys.path will cause Py_FatalError() to be called, which immediately terminates the entire application. On Windows, it is fairly easy for this to happen for developers or anyone who hasn't run the official Python installation; just have