Hello,
So I assume you first upgraded from 1.0.x (some x) to 1.0.15, or installed
1.0.15 from scratch, and then did an upgrade from that.
For some reason it looks like the C code that has changed since 1.0.15 (so
should have been installed in that last upgrade) has not been picked up.
This could be:
(1) That C code was not downloaded for some reason
(2) That C code was not compiled for some reason
(3) That compiled code was not picked up for some reason
(1) is unlikely unless you explicitly chose not to install all the
upgrades. (2) could happen, but recently other people had have problems
with (3).
To check (1):
cd ccpnmr1.0/c/memops/global
grep 'runFit' py_fit.c
You should get
static PyObject *runFit(PyObject *self, PyObject *args)
{ "runFit", (PyCFunction) runFit, METH_VARARGS },
To check (2):
cd ccpnmr1.0/c/memops/global (so same as above)
ls -l py_fit.c FitMethod.so
and the time for the latter should be >= the time for the former.
To check (3):
cd ccpnmr1.0/python/memops/c
ls -l FitMethod.so
This should be a symbolic link to the ccpnmr1.0/c/memops/global version of
FitMethod.so, but for some people seems to end up as a copy. If so, the
copy is out of date (so the ls -l here would give an older date than the
one in (2)). If that is the problem, then in this directory do:
source linkSharedObjs
And similary in ccpnmr1.0/python/ccp/c and ccpnmr1.0/python/ccpnmr/c.
Wayne
On Tue, 11 Dec 2007, Michelle Rowe wrote:
> As Mark has mentioned we have been having trouble with Analysis upgrades.
> The updates that cause errors have been narrowed down to three scripts:
> CalcRatesPopup, FollowShiftChangesPopup and DataAnalysisBasic. When these
> are added Analysis no longer restarts. The tracebacks of each are attached.
>
>
>
> Has anyone else had similar trouble?
>
>
>
> Michelle
>
>
|