[PyCUDA] Fwd: installing/upgrading issue with siteconf.py, setup.py and 0.94rc
Vince Fulco
vfulco1 at gmail.com
Sun Sep 5 14:27:07 PDT 2010
Dear Andreas- Thanks for your patience.
"...Further, which of those packages actually requires setuptools? A
quick look at their repositories suggests none of them do."
I can only state that using yum install for the libraries mentioned
previously after installing distribute states the need for setuptools
installation. Why, if as you mention each independently does not, I
have no idea.
"> Lastly, the build log you posted a while ago didn't show any
issues--what do you need help resolving"
After an install of Pycuda a la the build log you mentioned, any
attempt at running tests indicates missing python packages and a fail.
This didn't happen in 0.93 and trying to re-install 0.93 at this
juncture also fails.
Will attempt another clean re-installation of everything and report back asap.
Best, V.
On Sun, Sep 5, 2010 at 4:18 PM, Andreas Kloeckner
<lists at informa.tiker.net> wrote:
> Hi Vince,
>
> On Sat, 21 Aug 2010 09:51:29 -0400, Vince Fulco <vfulco1 at gmail.com> wrote:
>> To follow up on install failure detailed below, went ahead and removed
>> python-setuptools while installing distribute-0.6.4. The ongoing
>> unresolved issue is that to re-install numpy, scipy, rpy prior to
>> PyCuda requires the setuptools dependency which messes up the install
>> as previously mentioned. Would appreciate any further insights, V.
>
> Since distribute aims to replace setuptools, it should be installed at
> the very beginning of the process. (With distribute installed, "import
> setuptools" gives you distribute code.)
>
> Further, which of those packages actually requires setuptools? A quick
> look at their repositories suggests none of them do.
>
> Lastly, the build log you posted a while ago didn't show any
> issues--what do you need help resolving?
>
> HTH,
> Andreas
>
--
Vince Fulco, CFA, CAIA
612.424.5477 (universal)
vfulco1 at gmail.com
A posse ad esse non valet consequentia
“the possibility does not necessarily lead to materialization”
More information about the PyCUDA
mailing list