Research paper "Energy Efficiency across Programming Languages: How does energy, time, and memory relate?"
alister
alister.ware at ntlworld.com
Wed Sep 20 12:11:18 EDT 2017
On Wed, 20 Sep 2017 14:14:24 +0100, Paul Moore wrote:
> On 20 September 2017 at 13:58, alister via Python-list
> <python-list at python.org> wrote:
>> On Tue, 19 Sep 2017 14:40:17 -0400, leam hall wrote:
>>
>>> On Tue, Sep 19, 2017 at 2:37 PM, Stephan Houben <
>>> stephanh42 at gmail.com.invalid> wrote:
>>>
>>>> Op 2017-09-19, Steven D'Aprano schreef <steve+comp.lang.python@
>>>> pearwood.info>:
>>>>
>>>> > There is a significant chunk of the Python community for whom "just
>>>> > pip install it" is not easy, legal or even possible. For them, if
>>>> > its not in the standard library, it might as well not even exist.
>>>>
>>>> But numpy *is* in the standard library, provided you download the
>>>> correct version of Python, namely the one from:
>>>>
>>>> https://python-xy.github.io/
>>>>
>>>> Stephan
>>>>
>>>>
>>> Many of us can't pip install; it's in the OS supplied vendor repo or
>>> it doesn't go on the machines.
>>>
>>> Leam
>>
>> dnf install <package>
>> or apt_get install <package>
>>
>> most of the mainstream modules seem to be there (certainly numpy)
>
> You're missing the point. A significant number of Python users work on
> systems where:
>
> 1. They have no admin rights 2. Their corporate or other policies
> prohibit installing 3rd party software without approval that is
> typically difficult or impossible to get 3. Quite possibly the system
> has no network access outside of the local intranet 4. The system admins
> may not be able or willing to upgrade or otherwise modify the system
> Python
>
> Writing code that works only with stdlib modules is basically the only
> option in such environments.
>
> Having said that, I don't advocate that everything be in the stdlib
> because of this. A lot of things (such as numpy) belong as 3rd party
> packages. But that doesn't mean that "get XYZ off PyPI" (or "install XYZ
> alternative Python distribution/version") is a viable solution to every
> problem.
>
> Paul
not missing the point you said previously "it's in the OS supplied vendor
repo or it doesn't go on the machines."
dnf/yum or apt_get install form the "vendor supplied repo"
I fully understand that even this may require various hoops to be jumped
through before it can happen
--
Minnie Mouse is a slow maze learner.
More information about the Python-list
mailing list