I think spam bots have existed ever since online forums have existed. On GitHub, you can report them using a link in their profile (at the bottom of the left-hand panel).

First time I see some on Github, but thanks for the info.

Is this a paid course? If so has anyone on the Matplotlib team taken it so we know it is reasonable?

While we have linked to paid resources in the past, I think that is mostly for people we know from the project.

Is this a paid course? If so has anyone on the Matplotlib team taken it so we know it is reasonable?

While we have linked to paid resources in the past, I think that is mostly for people we know from the project.

Yes it is. @story645 has access to it, and I can give access to anyone in the matplotlib team (I only need an email), as I already said to @tacaswell by email.

I think we should be cautious about linking commercial resources, including books, as it gives the impression we think these resources are "good". I'm not at all saying that this particular contribution is not "good" or not "worth the money". However, I think we should reconsider linking to commercial resources, or if we do, the "Resources" page should have a disclaimer that we are listing these for informational purposes, and don't endorse any commercial links on the page.

While we have linked to paid resources in the past, I think that is mostly for people we know from the project.

@JosephBARBIERDARNAL has a bunch of third party packages and his collaborator (Yan Holtz) is the creator of the python graph gallery which is listed on the external resources page.

However, I think we should reconsider linking to commercial resources, or if we do, the "Resources" page should have a disclaimer that we are listing these for informational purposes, and don't endorse any commercial links on the page.

What about a paid resources section with a disclaimer there? Though we've never explicitly endorsed the free resources either.

I'm also gonna loop in the @matplotlib/steering-council since I sent 'em an email about this as a follow up to last weeks call discussion.

I think we should be cautious about linking commercial resources, including books, as it gives the impression we think these resources are "good". I'm not at all saying that this particular contribution is not "good" or not "worth the money". However, I think we should reconsider linking to commercial resources, or if we do, the "Resources" page should have a disclaimer that we are listing these for informational purposes, and don't endorse any commercial links on the page.

Fundamentally, listing something in "External resources" is a kind of endorsement. We would not add random stuff without knowing it is any good. There's no difference in whether it's free or paid content.

What we should do is

• add a statement that this is all 3rd party content and not affiliated to matplotlib. - I'm not sure that "informational purposes, and don't endorse" is really the right wording, but we can work on that
• mark all paid content as such, e.g. by adding (paid content) or a similar tag to the relevant entries. - I would not make separate sections for free/paid.
• Have a documented minimal bar for adding content here, which is somewhat higher than regular docs content. - It could be something like "two approvals and no veto by a core developer"; exact conditions t.b.d. and "content must be good quality. Approvers should have the impression that it can be beneficial to users"

add a statement that this is all 3rd party content and not affiliated to matplotlib. - I'm not sure that "informational purposes, and don't endorse" is really the right wording, but we can work on that

The third party packages page has:

A list of packages that extend Matplotlib. These are maintained and distributed independently from Matplotlib so go to the website listed for instructions.

Maybe a mash up:

A list of education resources for learning Matplotlib. These resources are maintained and distributed independently from Matplotlib and therefore not affiliated with Matplotlib.

Have a documented minimal bar for adding content here, which is somewhat higher than regular docs content

We should also have that for the third party page - I think the current standard is more or less that it's a documented and tested installable package.
ETA: I think the educational equivalent is something like the scope of the resource - it should provide a fairly comprehensive overview of matplotlib in service to its aims.

We would not add random stuff without knowing it is any good. There's no difference in whether it's free or paid content.

"two approvals and no veto by a core developer"

I'd be fine with that, if the members who approve actually view the material to vet it. However, I won't be watching 30 h of commercially available material unless I am also paid, and I wouldn't expect anyone else to do that - even textbook reviewers typically get honorariums for their time.

I'd also probably be OK if we wanted to have "sponsored content" where someone pays the project a nominal fee to include an advertisement that has clearly has no endorsement from us

"content must be good quality. Approvers should have the impression that it can be beneficial to users"

So skimming through mpl-journey, I think they've got a clearer definition of figure and axes than we do in our docs 😅 and they've got a clean scaffolded and chunked way of unrolling the information about mpl. I think for their audience, it's nicely motivated - axes has methods, want to make chart x? here's the methods on axes for that (with lots of interactivity that we can't have on our docs) - while still being concept first and wrapping up each section with quizzes and small modules. They also weave in visualization concepts/advice/best practices in a way I'd find out of scope for our docs.

Which they're covering:

• overview of mpl - figure, axes, basic customization via parameters
• customization/styling - mostly focusing on color/colormaps and theme but also mentions things folks ask for like ticks/spines/legend
• annotations - text/annotate/legend/fonts/arrows (mostly via custom library but we have multiple open discussions on arrows needing to be sorted out)+ styling (some of their custom font stuff they do via a third party library might get easier once all the font work is complete)
• custom layouts - subplots/subplot_mosaic/inset_axes/
• maps - geopandas based, covers maps, choropleths, scatters(bubbles)

tldr: I think it's a solid well crafted resource that I could see being especially good for folks who aren't scientists.

ETA: I guess I should add the disclaimer that I have no financial stake/conflict of interest w/ this resource, I've just interacted with both Yan and Joseph on socials b/c they've been sharing a lot of "how to make pretty mpl" for I think years now.

I'd also probably be OK if we wanted to have "sponsored content" where someone pays the project a nominal fee to include an advertisement that has clearly has no endorsement from us

The original offer was for an affliate link/coupon code but that was shut down b/c of concerns about tracking.

Thanks @story645 !

Some other things that might be relevant:

• Here is one of the lesson that I think can give you a good idea of the course: https://www.matplotlib-journey.com/module1/build-any-chart
• our course is really opiniated --> we don't try to be exhaustive about matplotlib, only what we consider the most useful (for example, we never explain how to use the pyplot interface)
• it's in continuous improvement

I'd be fine with that, if the members who approve actually view the material to vet it. However, I won't be watching 30 h of commercially available material unless I am also paid, and I wouldn't expect anyone else to do that - even textbook reviewers typically get honorariums for their time.

You don't have to do a tutorial / read a book completely to get a sense of quality. In fact, I believe that the assessment can often be done quite quickly by looking at a few passages. Therefore, I'd claim that paying for review in this specific context is barking up the wrong tree. It would come with it's own set of challenges. I would start with the base assumption that the topic can be handled like any other review in the sense that everybody is free to decide what they want to review. Only if such topics (of which we'll still have very few), then we'd need to discuss other approaches.

thanks!

I'll merge based on two positive reviews.