Template:Did you know nominations/Python Package Index

The following is an archived discussion of the DYK nomination of the article below. Please do not modify this page. Subsequent comments should be made on the appropriate discussion page (such as this nomination's talk page, the article's talk page or Wikipedia talk:Did you know), unless there is consensus to re-open the discussion at this page. No further edits should be made to this page.

The result was: promoted by Cwmhiraeth (talk) 05:26, 17 August 2016 (UTC)

Python Package Index

edit

5x expanded by Σ (talk). Self-nominated at 03:28, 3 August 2016 (UTC).

  • @Σ: New enough and long enough (5x expanded), neutral, no copyvio or close paraphrasing, AGF that the exact source cited verifies the hook (the book isn't in google books, it seems) but it's trivially verifiable elsewhere. Nominator is exempt from QPQ. Two small things:
  • The current ref 10 (PEP 301) says the article contains PD text from that source, but I don't see any. Also, while this used to be common to just incorporate it with a note at the end of the article, it's good practice to quote and cite it inline, both to avoid plagiarism issues and to keep track of what's attributed.
  • On the question you linked about current ref #13 (the code cite), that seems fine to me, it's plainly obvious in the code and involves no interpretation. I have sort of the opposite reaction to the sources, in fact; the ones that are traditionally "reliable" (books) are oooooold. Nothing to stop this from being a DYK, of course, but it's an irony of the way sourcing policies have evolved that it's hard to write about open-source software using the same kinds of sources that would be considered best for the subject by just about everyone else on the internet. Also, to be fussy, the Hussain source seems very esoteric considering it supports a fairly simple statement - I think this is a case where you do want a primary source. Opabinia regalis (talk) 06:36, 6 August 2016 (UTC)
  • @Opabinia regalis: Thanks for taking the time to look at this. I've added another source that links "cheese shop (PyPI)" with the Monty Python cheese shop.

    Actually, I've had six DYKs in the past! I somehow remember that back in the day, you were required to review another DYK only if you had made a certain number of DYK nominations in some duration. It appears that I either remembered wrong or the rules have changed. But I've just reviewed Template:Did you know nominations/Thomas L. Martin for this.

    My justification for the public domain note is that I took this excerpt straight from the PEP without modification: included mechanisms for the capture of package metadata, but did little with the metadata save ship it with the package. Though the sentence is quite literally copy-and-pasted, I don't think it should count as plagiarism with the citation and PD notice template.

    Thanks for double-checking ref 13. But I'm not sure what you mean about the Hussain source being "esoteric", could you explain?

    And yes, I agree that it's frustrating how sourcing for software works. Part of my reason to use hard sources is to establish notability, that PyPI isn't a minor thing that shouldn't be merged to the Python article itself. Another part of it is that sometimes there's simply no "regular" sources that outright document it. Some things, such as Egg files being zip files in disguise, are easily verifiable by using file (command) or straight up trying to unzip it as you would a zip file. Hence when "regular" sources don't say it, I end up citing mailing lists, source code, the ?:action=browse on the PyPI url itself, and could possibly even cite (to an extent) where http://cheeseshop.python.org takes you...

    Σσς(Sigma) 22:05, 6 August 2016 (UTC)

  • @Σ: Sorry for miscounting QPQs, either the search didn't pick up all of the old ones or I'm just blind :) On the PD thing, it's easier to either put it in quote marks or just rephrase (e.g., "was able to collect and distribute metadata but did not use it for other purposes"). On Hussain, I meant that it's kind of obscure - a paper from a German conference in 2011 - and seems to be used only to support the statement that you can host your package yourself rather than having PyPI do it, which should be easy to source from the current documentation. Opabinia regalis (talk) 02:58, 7 August 2016 (UTC)
  • Thanks for the proposed rewording, I've incorporated it into the article.

    As for the Hussain ref, this is an example of what I'm talking about when I say that sometimes there's simply no "regular" sources that outright document it. It's easy to put a link to the project and nothing more on PyPI - just fill out nothing but the "Project URL" field when you're creating the PyPI project entry - but I found nothing, not even official documentation, that said that. What do you think should be done? Σσς(Sigma) 07:01, 12 August 2016 (UTC)

  • @Σ: Oops, I forgot to get back to this; sorry! You're right, that's rather hard to find a "source" for, other than mailing-list chatter. Anyway it's hardly a DYK-specific problem, more of a problem for our definition of "reliable". Good to go! Opabinia regalis (talk) 21:30, 15 August 2016 (UTC)