On Oct 2, 2006, at 11:10 AM, hzhatlboro wrote:
Traceback (most recent call last):
File "C:\FiPy\FiPy-1.1.win32\FiPy-1.1\examples\diffusion
\mesh1D.py", line 779, in ?
exec(fipy.tests.doctestPlus._getScript())
File "<string>", line 638, in ?
TypeError: iteration over non-sequence
That's a bug that has been fixed in the Subversion repository, but
which hasn't made it into a full release yet.
See <http://matforge.org/fipy/changeset/1865#file0>
For the 64-bit WinXP, I can not get the installation right even if
I uninstall everything before a new installation. I tried Python
2.3-0.9.9 and 2.4-1.0.0 but both failed
What happens?
Later on, I tried Python-2.3.5-1076-wx26-n238, it worked except for
the same error as mentioned above.
So, with Python-2.3.5-1076-wx26-n238, FiPy 1.1 works on your 64-bit
WinXP system, with the exception of this error? That sounds good.
Does this error matter at all? It seems to me that line 779 is the
last line of that program?
The line numbers are deceptive. As you say, line 779 is practically
at the end of the program:
exec(fipy.tests.doctestPlus._getScript())
All this does is load in the docscript (from line 43 to line 773) and
run it as a program. The Traceback line:
File "<string>", line 638, in ?
refers to where the error occurs within the docstring, and gets you
close.
638 + 43 = 681 and the changed line is 678; I don't know why it isn't
exact.
I wish these error messages were more informative, but don't know how
to do that.
