[Python-ideas] stdlib crowdsourcing

anatoly techtonik techtonik at gmail.com
Fri Jun 1 17:08:21 CEST 2012


On Tue, May 29, 2012 at 9:02 AM, Nick Coghlan <ncoghlan at gmail.com> wrote:
> Once again, you're completely ignoring all existing knowledge and
> expertise on open collaboration and trying to reinvent the world. It's
> *not going to happen*.

It's too boring to live in a world of existing knowledge and
expertise, and yes, I am not aware of any open collaboration stuff
expertise. Any reading recommendations with concentrated knowledge
that can fit my brain?

> The standard library is just the curated core, and *yes*, it's damn
> hard to get anything added to it (deliberately so). There's a place
> where anyone can post anything they want, and see if others find it
> useful: PyPI.

The major drawbacks of remote packages in general is that it bring
back project compilation from the old days. The biggest Python
advantage at all times was "copy and run" ability.

The drawbacks of PyPI for this proposal are:
1. every function you need will require a separate upload to PyPI
2. you can't upload function with the same stdlib name, but slightly
different implementation as it is used in different projects
3. you can't find functions that people recommend to be included into stdlib
4. it is hard (impossible) to gather feedback on the quality of these proposals

> The standard library provides tools to upload to PyPI, and, as of 3.3,
> will even include tools to download and install from it.

I am glad 3.3 is giving virtualenv and bootstrap stuff. It would
really rock, if the new feature won't be settled in stone right after
release and will gain a few UX iterations with allowed break-ability.

As for PyPI, the major drawback of it is security - DNS attack for a
couple of minutes, and one of your automatically deployed nodes is
trojan ready. I remember PyPI password are stored in clear-text on
developer's machine, but I don't remember if anyone turned off HTTP
basic authorization on PyPI to protect passwords travelling to PyPI
with every upload from intercepting. It would be an interesting
exercise to sniff PyPI passwords over WiFi during next conference
(i.e. https://ep2012.europython.eu/) and match those to the
developer's accounts on *.python.org ;)

> If you don't like our ecosystem (it's hard to tell whether or not you
> do: everything you post is about how utterly awful and unusable
> everything is, yet you're still here years later).

You're absolutely right - I like the Python ecosystem, otherwise I
wouldn't stick there. It is like a vintage car - awesome, nice
looking, and there is even this new twisted pyusion engine inside,
but.. well - it's not for youngsters.

> If you think the PyPI UI is awful or inadequate, follow the example of
> crate.io or pythonpackage.com and *create your own*. There's far more
> to the Python universe than just core development, stop trying to
> shoehorn everything into a place where it doesn't belong.

I have absolutely no idea how aforementioned post touches PyPI UI.
Speaking about PyPI enhancements and ecosystem, instead of reinventing
bicycles I'd rather patch existing one. The only problem is that
patches are not accepted.
https://bitbucket.org/loewis/pypi/pull-request/1/fix-imports-add-logging-to-console-in



More information about the Python-ideas mailing list