These are ASCII-sortable:

0001_Introduction.ipynb

0010_Chapter-1.ipynb

ISO8601 w/ UTC is also ASCII sortable.

# Jupyter notebooks as lab notebooks

## Disadvantages

### Mutability

With a lab notebook, you can cross things out but they're still there.

- [ ] ENH: Copy cell and mark as don't execute (or wrap with ```language\n``` and change the cell type to markdown)

- [ ] ENH: add a 'Save and {git,} Commit' shortcut

CoCalc (was: SageMathCloud) has (somewhat?) complete notebook replay with a time slider; and multi-user collaborative editing. ("Time-travel is a detailed history of all your edits and everything is backed up in consistent snapshots.")

### Timestamps

You must add timestamps by hand; i.e. as #comments or markdown cells.

- [ ] ENH: add a markdown cell with a timestamp (from a configurable template) (with a keyboard shortcut)

### Project files

You must manage the non-.ipynb sources separately. (You can create a new file or folder. You can just drag and drop to upload. You can open a shell tab to `git status diff commit` and `git push`, if the Jupyter/JupyterHub/CoCalc instance has network access to e.g. GitLab or GitHub)

## Advantages

### Reproducibility Executable I/O cells

The version_information and/or watermark extensions will inline the software versions that were installed when the notebook was last run

Dockerfile for OS config

Conda environment.yml (and/or pip requirements.txt and/or pipenv Pipfile) for further software dependencies

BinderHub can rebuild a docker image on receipt of a webhook from a got repo, push the built image to a docker image repository, and then host prepared Jupyter instances (with Kubernetes) which contain (and reproducibly archive) all of the preinstalled prerequisites.

Diff: `git diff`, `nbdime`

### Publishing

You can generate static HTML, HTML slides with RevealJS, interactive HTML slides with RISE, executable source with comments (e.g. a .py file), LaTeX, and PDF with 'Save as' or `jupyter-convert --to`. You can also create slides with nbpresent.

MyBinder.org and Azure Notebooks have badges for e.g. a README.md or README.rst which launch a project executably in a docker instance hosted in a cloud. CoCalc and Anaconda Cloud also provide hosted Jupyter Notebook projects.

You can template a gradable notebook with nbgrader.

GitHub renders .ipynb notebooks as HTML. Nbviewer renders .ipynb notebooks as HTML.

There are more than 90 Jupyter Kernels for languages other than Python.

https://github.com/quobit/awesome-python-in-education#jupyte...