Hi, > I had a bug report that the Lanczos algorithm wasn't working[1], and it > turned out I hadn't enabled the relevant code path in the wrapper. The > user was using a dense matrix in the given example, so I enabled the > algorithm for dense types, and compressed and coordinate types. This was > all fine until I attempted a build on Windows, with VC9. > > [1] https://github.com/viennacl/pyviennacl-dev/issues/11 > > Aside from a number of double->float conversion warnings -- I can't see > where they're supposed to originate, even having looked at the relevant > lines of code
All (or at least most...) conversion warnings (both floating point and integers) are supposed to be fixed with a commit I pushed today. The issue with dense types looks like one of these boundary cases where Visual Studio just needs some additional love (maybe a C++03/C++11 issue). I'll look at it. Best regards, Karli ------------------------------------------------------------------------------ Is your legacy SCM system holding you back? Join Perforce May 7 to find out: • 3 signs your SCM is hindering your productivity • Requirements for releasing software faster • Expert tips and advice for migrating your SCM now http://p.sf.net/sfu/perforce _______________________________________________ ViennaCL-devel mailing list ViennaCL-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/viennacl-devel
