The site currently has data and content intertwined with each other in the HTML making the site a bit difficult to maintain/grow.

I propose to introduce flask with Frozen-Flask to provide tooling to make the site easier to oversee, maintain, and extend.

Main changes:

• Convert index.html and utrechtteam.html to templates
  • abstract out the bulk of the data on papers, and team members into yaml files. During build, this data is populated into the templates
• Other pages remain untouched
• Markdown support (flask-markdown) which compiles md to HTML during build
  • Markdown files can be included directly in the website
  • Markdown can be written directly within the HTML files

Everything according to plan, building the site will create the exact same website files that currently exist. I'm already familiar with this tooling, so this PR shouldn't take me too long to implement.

Blocked by #73

It might be interesting to invite Parcels user to contribute to the code with such a page.

In the instructions, we should link to this nice explanation on how to fork and contribute to a repo:
https://www.asmeurer.com/git-workflow/

This might be interesting as well
https://opensource.guide/how-to-contribute/

While everything else scales well, the animation at the top is cropped on mobile chrome:

Recreation

• Go to tutorials section
• Ctrl + click a tutorial
• Notebook opens in new tab, and current tab relocates to a 404 (eg. https://oceanparcels.org/documentation_stuck)

Initial thoughts
Clicking a link such as:

uses the captureTutorialLink callback:

I think the fix is just removing 'hitCallback': function(){document.location = url;}.

Happy to work on this in the coming week or two when I have time. Feel free to claim if someone else wants to patch this beforehand.

Summary

Errors per input

Errors in index.html

• [TIMEOUT] https://planktondrift.science.uu.nl/ | Timeout

Errors in waddendrifters.html

• [ERR] https://www.noordhollandsdagblad.nl/cnt/dmf20231122_38615721 | Failed: Too many redirects: error following redirect for url (https://login.mediahuis.com/authorize?response_type=code&client_id=8VmYoMJIvB1mZRDamQPQLYj8fvfnlaZ8&redirect_uri=https%3A%2F%2Fwww.noordhollandsdagblad.nl%2Fauth%2Fcallback&scope=openid+profile+email+offline_access&ui_locales=nl&prompt=none&state=eyJyZXR1cm5UbyI6Imh0dHBzOi8vd3d3Lm5vb3JkaG9sbGFuZHNkYWdibGFkLm5sL2NudC9kbWYyMDIzMTEyMl8zODYxNTcyMSIsImlzU2lsZW50Ijp0cnVlfQ): too many redirects

Errors in vegu21.html

• [TIMEOUT] https://planktondrift.science.uu.nl/ | Timeout
  Full Github Actions output

From Parcels v.2.4.2., there is no more parcels_get_examples script, so step 4 of the installation instructions needs to be updated. I think it should now be a matter of downloading the zip at the Tutorial and Documentation page, right? Can you confirm, @VeckoTheGecko?

Recently we have added many links between tutorial notebooks and the documentation for a more natural guide on specific parcels topics. We also added Google Analytics event tags to the tutorial tiles on the homepage to see how often people visit the notebooks. These event tags do not cover all the other links however. This is not a major problem but something to be aware of.

Prettier support is deprecated for pre-commit. Migrate to Biome.

The old website had a 'parcels documentation' tab, sorted by topic and with links to fragments of the source code. It was not properly updating with the releases and was quite ugly, however it offered a nice quick access to the code and supported quick learning/familiarising for users that wanted to dig deeper then the tutorials (or prefer reading small parts of the code, without having to open the full source code in a project and search in there). Is this coming back in the new website?

Is there any documentation about the datasets in examples-data (which are in turn used throughout the parcels examples)? Perhaps it would be worth adding a file examples-data/README.md to add quick descriptions to these. This can be propagated through to users in parcels if we want as well just by linking to the README here on GitHub.

Motivation

parcels_tutorial.ipynb mentions

since we know that the files are in what’s called Parcels FieldSet format, we can call these files using the function FieldSet.from_parcels()

I feel like the following would be better

The "MovingEddies_data" dataset contains data that is in Parcels FieldSet format. We can ingest this data into parcels using FieldSet.from_parcels(). For more information on available datasets see link