Hmph...it doesn't seem to be possible to "install" the _openbabel.so
file to site-packages with distutils. Sure, I can hardcode the
copying, but that was what I originally wanted to avoid.
If no-one has any suggestions, I'll ask about this on a Python list...
- Noel
On 9 December 2011 20:00, Mar
On Fri, Dec 9, 2011 at 12:56 PM, Geoffrey Hutchison
wrote:
>
> On Dec 9, 2011, at 8:08 AM, Noel O'Boyle wrote:
>
>> I think about it, we could just use distutils to install the package
>> after it has already been compiled (this was where the problem was).
>
> That would be my suggestion. CMake's
On Dec 9, 2011, at 8:08 AM, Noel O'Boyle wrote:
> I think about it, we could just use distutils to install the package
> after it has already been compiled (this was where the problem was).
That would be my suggestion. CMake's logic does not work for installing Python
modules on Mac, because it
On 7 December 2011 20:20, My Th wrote:
> 2011/12/7 Noel O'Boyle :
>> I'm afraid, it's not possible to ask disutils the location of the
>> site-packages folder (believe me, I have tried, googled, everything).
>> As a result, we have no choice. In any case, this simplifies
>> everything and I wouldn
2011/12/7 Noel O'Boyle :
> I'm afraid, it's not possible to ask disutils the location of the
> site-packages folder (believe me, I have tried, googled, everything).
> As a result, we have no choice. In any case, this simplifies
> everything and I wouldn't change it back now. The user just needs to
I'm afraid, it's not possible to ask disutils the location of the
site-packages folder (believe me, I have tried, googled, everything).
As a result, we have no choice. In any case, this simplifies
everything and I wouldn't change it back now. The user just needs to
add the lib directory to PYTHONPA
Hi!
Please teach CMake to install bindings in standard locations as
distutils is doing. That should help distributions also.
Reinis
--
Cloud Services Checklist: Pricing and Packaging Optimization
This white paper is int
Hi Noel,
On Tue, Dec 6, 2011 at 3:25 PM, Noel O'Boyle wrote:
> Hi there,
>
> I eat my words, Marcus - you were right. We should have been using the
> CMake build system for the bindings, rather than using distutils for
> Python, something else for Ruby, and so forth. There were various
> reasons
Hi there,
I eat my words, Marcus - you were right. We should have been using the
CMake build system for the bindings, rather than using distutils for
Python, something else for Ruby, and so forth. There were various
reasons we (or I rather) took that decision, but they are overwhelmed
now by the m