On May 18, 5:30 pm, William Stein <wst...@gmail.com> wrote: > On Tue, May 18, 2010 at 8:21 AM, Pablo Winant <pablo.win...@gmail.com> wrote: > > I guess the problem does not arise in Python's namespace. Unless I'm missing > > something, I don't see how 'import pytave' statement could conflict with > > singular python modules or objects. > > > There seem to be an incompatibility between octave and singular shared > > libraries (.so file) related to C++ templating. According to pytave's dev, > > That makes sense. We've had problems like this before. E.g., at one > point suddenly MatPlotLib had a new C++ type with the same name as > something in another C++ library that Sage links in (maybe licecm or > NTL). The fix that time was to change libecm or NTL. >
It's been a while since I coded in C++, but isn't there always a simple way out, just by putting these classes in different namespaces? (IMHO it's a silly of a library developer to call something "Array" and put in into the global namespace, hoping that noone else ever is going to call "Array" something entirely different...) Dima > William > > > > > > > "The problem is conflicting implementations of Array<int>. Both liboctave > > and Singular have their own template <class T> class Array with an int > > instantiation". I thought it was impossible to simultenously load the two > > .so files with the same python interpreter. Is it correct ? > > > Pablo > > > Le 18/05/2010 16:41, Dima Pasechnik a écrit : > > >> Certainly, such name clashes are not uncommon (e.g. cvxopt defines its > >> own > >> matrix that clashes with Sage's matrix) > >> IMHO if you import the needed things from the package you need, doing > >> re-namings when needed, as usual (from foo import blah as fooblah) > >> you can avoid the problem you describe. > > >> HTH, > >> Dmitrii > > >> On May 18, 10:44 am, Pablo Winant<pablo.win...@gmail.com> wrote: > > >>> Hello, > > >>> I am still trying to use Pytave under Sage. After some work, I > >>> eventually managed to compile every needed library (boost, octave), but > >>> I went through an annoying bug. > >>> It is possible to use the library with 'sage -python' or 'sage > >>> -ipython' but when imported at the sage command prompt it fails. > >>> According to pytave developper, it is due to an incompatibility between > >>> the templates of octave and singular, the class Array ending up being > >>> declared twice. > > >>> For this reason, I am wondering whether it is possible to restrict the > >>> initial import into sage interpreter so as to avoid importing singular. > >>> Would something usable remain if I remove all modules depending on it (I > >>> don't really need pure mathematics functions anymore) ? What would be > >>> the right way to do it ? > > >>> Best, > > >>> Pablo > > >>> P.S. the attempts I made are tracked > >>> herehttps://bugs.launchpad.net/pytave/+bug/578767. Once it is done I would > >>> be more than happy to contribute doc/script on the matter. > >>> <https://bugs.launchpad.net/pytave/+bug/578767> > > >>> -- > >>> To post to this group, send an email to sage-devel@googlegroups.com > >>> To unsubscribe from this group, send an email to > >>> sage-devel+unsubscr...@googlegroups.com > >>> For more options, visit this group > >>> athttp://groups.google.com/group/sage-devel > >>> URL:http://www.sagemath.org > > > -- > > To post to this group, send an email to sage-devel@googlegroups.com > > To unsubscribe from this group, send an email to > > sage-devel+unsubscr...@googlegroups.com > > For more options, visit this group at > >http://groups.google.com/group/sage-devel > > URL:http://www.sagemath.org > > -- > William Stein > Professor of Mathematics > University of Washingtonhttp://wstein.org > > -- > To post to this group, send an email to sage-devel@googlegroups.com > To unsubscribe from this group, send an email to > sage-devel+unsubscr...@googlegroups.com > For more options, visit this group athttp://groups.google.com/group/sage-devel > URL:http://www.sagemath.org -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
