[tl;dr version: please take the survey at
https://www.surveymonkey.com/r/KVYHFKK]
According to our documentation (not that we've tested in awhile) FiPy currently
supports all Python versions from 2.4 on. If you use the 2to3 utility, FiPy
also works with Python 3, but this isn't a configuration we test very much.
Python has advanced a lot over the years. There are already some features like
context managers ('with', introduced in Python 2.5) that we're starting to use
in a limited way in non-critical areas of FiPy.
Until our primary solver libraries (PySparse and PyTrilinos) support Py3k, we
don't see trying to do real work in anything but Python 2, but we certainly
don't want to be an obstacle to adopting what is clearly the future of Python.
Furthermore, if you are doing real work in Py3k, we'd love to know about it.
To that end, a FiPy user (enormous thanks to Mike Müller!) has proposed a pull
request <https://github.com/usnistgov/fipy/pull/473> that enables FiPy to run
in either Python 2 or Python 3 without conversion. This pull request uses the
'python-future' package <http://python-future.org> that looks pretty spiff, but
imposes a requirement of Python 2.6+ or Python 3.3+
<http://python-future.org/faq.html#supported-versions>. There is an alternative
compatibility library called 'six' <https://pythonhosted.org/six/> that
supports older versions of Python, but which also looks like it would be much
more cumbersome to use in the long run.
It would be very helpful if you told us a bit about how increasing the minimum
required version of Python (whether via this pull request or for any other
reason) might impact your work:
https://www.surveymonkey.com/r/KVYHFKK
_______________________________________________
fipy mailing list
[email protected]
http://www.ctcms.nist.gov/fipy
[ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]
