including libttsh11 fixed the problem. Thank you! Now I can get on with fixing everything that Python 3 broke... err changed. :)
-- Cliff On Sat, Aug 28, 2010 at 11:20 AM, Alexander Gattin <xr...@yandex.ru> wrote: > Hello, > > On Sat, Aug 28, 2010 at 09:27:05AM -0400, Cliff > Martin wrote: > > Yes, our entire toolchain is 64 bit - a mix of > > stuff I have downloaded and built and some > > packages from HP (in the form of depot files) > > GCC was downloaded from HP, for example. > > I see. I bootstrapped from bundled cc, hence all > the problems. > > > Python -d did not generate any additional > > information, and so was not helpful (should this > > work?). > > Oops I was wrong about the python -d -- > correct option is -v of course... > > > Python -v did, however, and it came up with a > > number of unresolved symbols all seeming to be > > from libnnz11.so. I tried linking against all of > > the *.so files in ORACLE_HOME/lib, but I don't > > remember trying libttsh11 specifically. I will > > try it again on Monday. > > You're using Oracle 11 vs our v10 (we also have > v8, v9 and v11 in production, but not on this > HP-UX server), but I think the problem with the > libnnz is the same: Oracle doesn't put correct > shared library dependencies into the libnnzXX.so > dynamic section header (it should list > libttshXX.so as NEEDED but apperently doesn't). > > Probably their distribution for Solaris is better, > I didn't check (I'll ask our DBAs on Monday). > > -- > With best regards, > xrgtn >
-- http://mail.python.org/mailman/listinfo/python-list
