Can't build with Python 2.6 at the moment (C-API changes). Tried this past weekend using Mingw32. Will have to get time to update the code to get the port done.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
NumPy just released version 1.3 with support for Python 2.6. Perhaps the NumPy developers might be able to answer some relevant questions to help move this forward? (They have a mailing list: http://mail.scipy.org/mailman/listinfo/numpy-discussion .)
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Might you post a summary of what creates the compilation problem? It would be really nice to have a Windows installer for 2.6 (and actually, for 3 as well). Thanks.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Anthony Tuininga offers the following.
Is this adeuate info?
The problem is that Python 2.6 uses the new C library that comes with
Visual Studio 2008. Microsoft has seen fit to eliminate a number of
standard C routines. In my case I found equivalent routines in Python
so I could avoid using them. In your case I'm not sure what you are
going to want to do. Its only inside a debugging function so you can
probably quite safely turn it off completely but I'll leave that to
you to decide. If you want further help with this, let me know. I can
also provide you with the equivalent code for Windows that I use in
cx_Logging.
The offending lines are (in the 2.1.1a1 release)
stt/TextTools/mxTextTools/mxstdlib.h (lines 157 and 187-189)
If you comment out these lines the problem goes away completely.
This is the code I use in cx_Logging. The 'd' part is for the date and
the else part is for the time. If you have questions, fire away. :-)
Can't build with Python 2.6 at the moment (C-API changes). Tried this past weekend using Mingw32. Will have to get time to update the code to get the port done.
NumPy just released version 1.3 with support for Python 2.6. Perhaps the NumPy developers might be able to answer some relevant questions to help move this forward? (They have a mailing list: http://mail.scipy.org/mailman/listinfo/numpy-discussion .)
Might you post a summary of what creates the compilation problem? It would be really nice to have a Windows installer for 2.6 (and actually, for 3 as well). Thanks.
Anthony Tuininga offers the following.
Is this adeuate info?
The problem is that Python 2.6 uses the new C library that comes with
Visual Studio 2008. Microsoft has seen fit to eliminate a number of
standard C routines. In my case I found equivalent routines in Python
so I could avoid using them. In your case I'm not sure what you are
going to want to do. Its only inside a debugging function so you can
probably quite safely turn it off completely but I'll leave that to
you to decide. If you want further help with this, let me know. I can
also provide you with the equivalent code for Windows that I use in
cx_Logging.
The offending lines are (in the 2.1.1a1 release)
stt/TextTools/mxTextTools/mxstdlib.h (lines 157 and 187-189)
If you comment out these lines the problem goes away completely.
This is the code I use in cx_Logging. The 'd' part is for the date and
the else part is for the time. If you have questions, fire away. :-)
#define DATE_FORMAT "%.4d/%.2d/%.2d"
#define TIME_FORMAT "%.2d:%.2d:%.2d.%.3d"
#if defined MS_WINDOWS
GetLocalTime(&time);
#else
gettimeofday(&timeOfDay, NULL);
localtime_r(&timeOfDay.tv_sec, &time);
#endif
}
#ifdef MS_WINDOWS
if (*ptr == 'd')
sprintf(temp, DATE_FORMAT, time.wYear, time.wMonth,
time.wDay);
else
sprintf(temp, TIME_FORMAT, time.wHour, time.wMinute,
time.wSecond, time.wMilliseconds);
#else
if (*ptr == 'd')
sprintf(temp, DATE_FORMAT, time.tm_year + 1900,
time.tm_mon + 1, time.tm_mday);
else
sprintf(temp, TIME_FORMAT, time.tm_hour, time.tm_min,
time.tm_sec, (int) (timeOfDay.tv_usec / 1000));
#endif