Re: [sage-devel] Building Sage 7.4 final release on OS X El Capitan

2016-11-07 Thread Paul Masson
As a check of the build process, I did download a pristine copy and it built with multiple jobs. The copy that did not build cleanly is the one I've been updating all summer from the GitHub master, which is certainly not pristine. I did run a make dist-clean on it but that did not solve the

Re: [sage-devel] Building Sage 7.4 final release on OS X El Capitan

2016-11-07 Thread Jeroen Demeyer
On 2016-11-03 20:40, Paul Masson wrote: OSError: [Errno 2] No such file or directory: '/Users/Masson/Downloads/GitHub/sage/local/lib/python2.7/site-packages/backports_abc-0.4-py2.7.egg' Error installing Cython This looks very much like https://trac.sagemath.org/ticket/21672 which should be

Re: [sage-devel] Building Sage 7.4 final release on OS X El Capitan

2016-11-07 Thread Erik Bray
On Thu, Nov 3, 2016 at 8:40 PM, Paul Masson wrote: > The one package that stuck in my mind was Cython, and for some reason the > error log for that one is not yet overwritten. Here's that error message: > > OSError: [Errno 2] No such file or directory: >

Re: [sage-devel] Building Sage 7.4 final release on OS X El Capitan

2016-11-03 Thread Paul Masson
The one package that stuck in my mind was Cython, and for some reason the error log for that one is not yet overwritten. Here's that error message: OSError: [Errno 2] No such file or directory: '/Users/Masson/Downloads/GitHub/sage/local/lib/python2.7/site-packages/backports_abc-0.4-py2.7.egg'

Re: [sage-devel] Building Sage 7.4 final release on OS X El Capitan

2016-11-02 Thread Francois Bissey
It is probably not limited to El Capitan it would have been useful to get some more details. Francois On 3/11/2016, at 2:07 PM, Paul Masson > wrote: I had difficulty over the last couple days getting Sage 7.4 to build. Something like ten

[sage-devel] Building Sage 7.4 final release on OS X El Capitan

2016-11-02 Thread Paul Masson
I had difficulty over the last couple days getting Sage 7.4 to build. Something like ten packages would fail to build, but not always the same ones. It finally occurred to me that there might be some race condition going on. Sure enough, running make without multiple jobs solved the problem.