Closed mxlgv closed 3 years ago
I think the old tinypy all numerical values (include integer types) was represented by a float. This may cause some issues casting between system int to and from a TP_NUM object.
The current version uses a double for TP_NUM. And we really should add support to a real, lossless integer type.
Thanks for the answer. Does this mean that you also have this bug?
I remember there was a bug in the old TinyPy. For the x86 platform, for unknown reasons, the arguments were not read correctly. For example, the function I my_func (0xffffffff) in TynyPy returned the value 0x7ffffff. Although if you collect for x64, then everything is fine: the values are the same. 5 days looking for a problem in the source code - so I did not find it. Do you know anything about this?