Re: location of arch-dependent python module

2020-11-18 Thread Kasper Peeters
> CMake is not the usual choice of upstream build system for Python > modules This is a module written in C++, for which CMake makes a lot of sense (the tool originally did not have Python bindings at all). > It looks like this is the solution that you were referring to: > > https://github.com/k

Re: location of arch-dependent python module

2020-11-17 Thread Kasper Peeters
> I can of course use a Debian-specific toolset to do the installation, > but that doesn't really solve my problem To clarify that: no, I am not attempting to replace the Debian toolset for a package meant for the Debian repos; someone is taking care of the Debian package of my software for the of

Re: location of arch-dependent python module

2020-11-17 Thread Kasper Peeters
> The default tools should do that automatically, which ones are you > using and how did you solve it with them? The problem is finding the right default solution in the right default tool ;-) On a serious note, my software builds on various platforms/distros and I need packages for as many of th

Re: location of arch-dependent python module

2020-11-17 Thread Kasper Peeters
> Most Python packages appear to put .so files in a different location: > > $ apt-file search x86_64-linux-gnu | grep python | grep .so | grep > packages | grep ^python3- | head -n1 > python3-acora: /usr/lib/python3/dist-packages/acora/_acora. Yeah, I should have done my homework. What I missed w

location of arch-dependent python module

2020-11-16 Thread Kasper Peeters
I am trying to package a program which contains a Python module. This module is not written in Python, but in C++/pybind11. It seems to me that the '.so' file this produces should go into an arch-dependent directory, e.g. /usr/lib/x86_64-linux-gnu/python3.6/site-packages/ However, these are no

location of arch-dependent python package

2020-11-16 Thread Kasper Peeters
I am trying to package a program which contains a Python module. This module is not written in Python, but in C++/pybind11. It seems to me that the '.so' file this produces should go into an arch-dependent directory, e.g. /usr/lib/x86_64-linux-gnu/python3.6/site-packages/ However, these are no

Re: RFS: LiE -- computer algebra package for Lie group computations

2007-02-09 Thread Kasper Peeters
> * tipp: remove the useless empty line at the end of changelog > > * tipp: it's common to list debhelper as the first depends in > Build-Depends > > * tipp: remove the useless empty line at the end of the manpage. > > * tipp: do not use more than one empty line at the time as a >

Packages (using C++) for two different distributions

2006-12-20 Thread Kasper Peeters
Hi, (Please redirect me to the right mailing list if this is off-topic here, apologies for the waste of bandwidth). I have written a program which makes use of the C++ interface of libgmp. On Ubuntu Dapper, this leads to a dependence on libgmpxx3, while the upgraded gcc on Edgy makes this depend

RFS: LiE -- computer algebra package for Lie group computations

2006-12-20 Thread Kasper Peeters
Having cleaned up my package with the help of Rodrigo Gallardo, I would now like to get sponsorship for LiE -- computer algebra package for Lie group computations LiE is a computer algebra system that is specialised in computations involving (reductive) Lie groups and their representatio

Re: RFS: LiE computer algebra for Lie groups

2006-12-12 Thread Kasper Peeters
> Your work in packaging has a copyright and thus needs a license. Ok, all done. DD's, can I get sponsorship for http://www.aei.mpg.de/~peekas/debian/lie_2.2.2-1.dsc (all other files in http://www.aei.mpg.de/~peekas/debian/) Thanks! Best, Kasper -- To UNSUBSCRIBE, email to [EMAIL PROTEC

Re: RFS: LiE computer algebra for Lie groups

2006-12-12 Thread Kasper Peeters
> Please post the complete URL to your .dsc file, it's easier for > potential sponsors to grab it. http://www.aei.mpg.de/~peekas/debian/lie_2.2.2-1.dsc > FWIW: Dear DD's, pending minor comments below, I believe this package > is close to ready for sponsorhip. I have fixed all the remaining probl

Re: RFS: LiE computer algebra for Lie groups

2006-12-11 Thread Kasper Peeters
> Once you make these changes, repost your sponsor request. I have made all the changes you mentioned in your two previous emails. New files are at http://www.aei.mpg.de/~peekas/debian/ There's only one warning left when running lintian on the package: W: lie: description-synopsis-might-not

Re: RFS: LiE computer algebra for Lie groups

2006-12-07 Thread Kasper Peeters
> If that is all the license statement then yes, you need > clarification. Please get in touch with the author. Ok, I'll do that. > We need to see your source package to comment. Please post an URL to > the .dsc .dsc:http://www.aei.mpg.de/~peekas/lie_2.2.2-1.dsc .tar.gz: http://www.aei.mpg.d

RFS: LiE computer algebra for Lie groups

2006-12-07 Thread Kasper Peeters
Name:LiE Short: Computer algebra system for Lie groups. License: "publically available for free" (probably needs some coordination with the author). Description: LiE is a computer algebra system that is specialised in computations involving (reductive