Exception: 2147483648 is larger than the maximum possible size (2147483647) about bx-python HOT 5 OPEN

--- Long answer, less coherent --- The original test was probably something I wrote. Not very fresh in my memory though (to say the least!). I haven't been able to find my circa 2005 implementation of this. I did find the C implementation. I think MAX is supposed to relate to the maximum interval end that will be storable in the file format. In my 2005 C implementation, it looks like this was called MAX_INTERVAL_END. I presume the file format has a fixed number of bytes in which to represent an interval end. Thus MAX should be independent of the system and instead should depend on the file format/layout/design. Whether MAX should be 2**31, 2**31-1, or 2**32-1 ... I'd have to dig more into the code. (digging) It looks like at some point the file format has been enhanced beyond version 1 (line 105 and comment at line 124). But the comments describing the file format weren't completely updated, since they still claim version 1 (line 8 and lines 25 thru 81). That description matches my 2005 description. At line 451 it writes an interval's start and end to the file without modification. So (assuming line 78 still accurately reflects the file format), we have 32 bits for the end value, and the maximum allowable end would be 2^31. I don't understand why the test at line 133 is max_size<max instead of max_size<=max. In the comment at line 130, does "(0,max_size)" mean the interval is intended to be closed? The only similar reject message I find in my 2005 code rejects when something is > MAX_INTERVAL_END. But the test at line 133 rejects when something is >= MAX_INTERVAL_END. SO ... I think MAX is correct, but that the test at line 133 is probably wrong. It's also possible (probable) that line 120 is wrong. On my machine sys.maxsize is 2**63-1. Bob H