The page https://scikit-hep.org/documentation has an issue with the displaying of the table because the formatting insists on adding a line between the definition of the header and the programmatic filling of the table entries:

| Package | README | Documentation |
| ------- | ------ | ------------- |

{% for project in items %}

Present render:

A page on conda forge adding & maintenance would be nice, and has been requested a few times.

@kratsg recently broached the idea of having a way to demo some interesting public applications of the Scikit-HEP stack to potential users. For example, he recently made a Binderized GitHub Gist of Reproducible ATLAS SUSY Summary Plots which is a great example.

I had suggested that one approach might be an application gallery similar to matplotlib's. This could be a way to both demonstrate actual public uses of the Scikit-HEP stack out "in the wild" and be a handy thing to have on hand to point to in talks and demos.

I'm not sure if it would live directly on the website or somewhere adjacent.

Thoughts from @henryiii and @eduardo-rodrigues?

pyhf is used in multiple analyses in ATLAS, and uproot is used all over the place and many projects are used in work for IRIS-HEP. It would be good to add both of these collaborations to the Who uses Scikit-HEP? page.

The style seems to be somewhat broken (mentioned in #64 ). Investigate.

We should add caps checks to rule out PyTest, CMake, PyBind, and any other incorrect capitalizations.

Might be able to add a spelling check, too.

We could remove the readme links and just append /#readme to the project.url if readme: is missing. It's really only used on one page that likely is not the one most people would use; the main page just has normal links. A link to the package itself should not auto-scroll the page, because users might want a variety of information from visiting a project, like the most recent download, the stars button (very important!), the clone button, the issues tab, etc. I hope everyone knows how to scroll down on GitHub (and would really like it if GitHub made the files collapsible), but I don't want to hide the important links at the top of the page, at least for the "normal" link. On the packages page, linking to the README's is fine, since it's a separate link that says README.

Originally posted by @henryiii in #99 (comment)

The current intro text is quite long; if you have a smaller screen, you may not even know that there is a list of packages below, which is one of the most useful parts of the main page. I recommend shortening it quite a bit, and maybe even moving most of it to a separate "about" page. Thoughts, especially @eduardo-rodrigues ?

As seen in the main page https://scikit-hep.org/, the lists of packages within each group are not ordered alphabetically. Nicer to ensure the ordering.

Indeed the tutorials available as a Jupyter notebook at https://scikit-hep.org/scikit-hep-tutorials are not mentioned anywhere, and they should. (See discussion at #108.)

A possible place to mention the tutorials is the https://scikit-hep.org/resources page.

Stefan gave us a data-domain for scikit-hep.org; this snippet should be added to every HTML page:

<script defer data-domain="scikit-hep.org" src="https://views.scientific-python.org/js/plausible.js"></script>

It doesn't matter where on the page it appears (see a long conversation I had with myself here: scikit-hep/awkward#1819), and I think the defer could be replaced by async. You just need to find a template that gets rendered on every page you care about. (In my long conversation, I decided that I didn't care about redirects, but that's up to you.) There are templates in _includes and _layouts.

By the way—context for anyone seeing this without background—this is a follow-up of a discussion we had on the Scikit-HEP administrators list. The scientific-python.org organization is providing website statistics through Plausible, which we can use on our documentation sites. The reason we want it is to get a sense of what our users are interested in. In particular, the Awkward Array library is organized as a collection of ak.* functions, each with its own page; if we find 100× more website traffic to ak.this than ak.that, then we'll know that ak.this is more important (should get more developer attention) than ak.that. The usefulness to the scikit-hep.org pages is different: "Are users looking at 'Project News'? 'Packages'? 'User Information'? ...)

In the presentations at SciPy, they said that Plausible is less intrusive than Google Analytics (which I've used in the past). In particular, Google Analytics violates some regulations in Europe, but Plausible does not. So the choices seem to be well thought-through.

I've navigated to the web-page from pypa/gh-action-pypi-publish#15 (comment) and then saw a typo there. I wanted to send a patch (#70) but It took me about 10-15 min just to find this repo.

Please explore how it could be made better discoverable. Maybe add a corner-banner saying "edit this page on GitHub" or something like that...

At the moment the "User information" menu contains "Installing conda" and "Resources". I would argue that "Installing conda" is for the "Developers information" (not spefic to Scikit-HEP in particular) and would move the "Getting in touch" presently under menu "About" to menu "User information", which seems more natural.

Thoughts?

This page claims

For Python packages without binary extensions and fairly simple builds can use a modern build system instead of the classic but verbose setuptools

The phrasing implies that PEP 621 is somehow incompatible with building wheels including compiled artifacts, which is wrong:

E.g. hatch supports build hooks, which allow you to run custom code just like setup.py.

We should list a few things that you should not do, like using setup.py directly, especially not setup.py test, which is discouraged and deprecated. Maybe include common mistakes, like accidentally included the "tests" folder in the install (or maybe even the SDist).

Add pre-commit.ci as a use recommendation to the Developer Information pages given that Anthony Sottile has approved of advertisement of the project.

A proper page on writing typed Python and using mypy would be useful.

Are you going to add info on these files somewhere on the site? The Scikit-HEP shell figure might actually be really nice somewhere at https://scikit-hep.org/about with some bit of text.

Originally posted by @eduardo-rodrigues in #282 (comment)

This is a great idea that @eduardo-rodrigues had that I was reminded of by a comment from @pelmer. 👍

IMO a guide should be added to scikit-hep.org, probably under developer. Maybe could be combined in an "extras" page that covers license, badges, etc?

Originally posted by @henryiii in scikit-hep/hist#416 (comment)

Can scikit-hep.org have HTTPS support added? It would be nice in general to have this, as this also affects all the project pages as well.

Since this is a custom domain, and not just using scikit-hep.github.io I assume it is more involved then clicking the "Enforce HTTPS" in Settings. Though, I assume that @henryiii knows how to do this, as he has HTTPS working for https://iris-hep.org/.

Should we maybe have a page on news as for example the scikit-learn website https://scikit-learn.org/ has? This would be a place to add major news such as the first release of a new package, a package or project joining the group, major releases that can potentially make a difference (backward incompatibilities), etc. One could also add bits of info such as what I recently mentioned on the Gitter channel, namely "A cute piece of info: Scikit-HEP made it to the PDG online particle physics information".

What do you think, @henryiii, @HDembinski, @matthewfeickert?

In PR #193 Scikit-HEP got a news page! 🎉 At the moment all the releases are noted as "major" even if they are "minor" releases in SemVer.

Example: boost-histogram v1.0.0 and pyhf v0.6.0 are both described as "major"

Should we adopt SemVer conventions and change the SemVer minor releases to be described as "minor"?

As first brought up in the Scikit-HEP Gitter it would be nice if we designed a Scikit-HEP badge that would link to the project webpage that all projects would have in their READMEs. As the READMEs end up getting put up on PyPI during publishing it could be a useful way to get a curious user from PyPI to the project quickly. So maybe something along the lines of

but something more interesting.

As @henryiii did a great job with the color pallet choices for the PyPI badges

I'm hoping that he has some thoughts on the badge design as well. 🙂

A page like this might be useful:

https://www.scipy.org/citing.html

Given scikit-hep/pyhf#1449 and @kratsg's scikit-build Issue which motivated the PR, it seems that current use of prune in the MANIFEST.in section of the Packaging docs

prune *
graft src
graft tests

include LICENSE README.md pyproject.toml setup.py setup.cfg
global-exclude __pycache__ *.py[cod] .*

might need to get revised to something like

prune **
graft src
graft tests

include LICENSE README.md pyproject.toml setup.py setup.cfg
global-exclude __pycache__ *.py[cod]

as prune * is not actually removing all files from the manifest to start with.

If we're not missing anything we're happy to make a PR for this, but I thought we'd bring it up for discussion first.

@henryiii @eduardo-rodrigues let us know if you have any concerns here.

The main User information web page contains an invalid link: [Conda installation instructions](Conda installation instructions)

As we advocate for the use of pytest, the https://github.com/pre-commit/pre-commit-hooks hook name-tests-test might be a useful hook to add to the suggested hooks. Though as

is about a minimal configuration if the idea here is to add hooks that are agnostic to most other choices then I could make a PR to https://github.com/scikit-hep/cookie instead (if people think this is a good idea in either place I'll happily make a PR to wherever).

The current grouping leaves quite a few packages as single items in a group. Here's my suggested changes:

Put Simulation into Interface to HEP libraries, which would have pyhepmc, numpythia, and pyjet.

Rename Units and Constants to Math or Basics (or something like that), so Vector can be added there later.

Event Processing is a little odd with only one item in it, but I think that's okay. Will eventually have Awkward 1 and 0.

It would be nice to have a pytest page, discussing how to do (proper) pytest, with parametrized fixtures and such.

Please fill in the GitHub webpage link (top right) with scikit-hep.github.io, so that someone finding the repository can click to switch to the webpage.

With Github now supporting Dark mode, and iscinumpy & iris-hep supporting it since launch, now Scikit-HEP is the main offender for shining a white screen in my face at night. Needs a dark mode. ;)

Or even better define HEP on the main page.