search

projectnessie / nessie

Nessie: Transactional Catalog for Data Lakes with Git-like semantics
https://projectnessie.org
Apache License 2.0
984 stars 125 forks source link

Demo notebooks - comparable static versions instead of interactive ones #2811

Open datocrats-org opened 2 years ago

datocrats-org commented 2 years ago

The Try Now button on projectnessie.org redirects to here. That site links to 4 different python notebooks, each as freestanding examples that can be run interactively (spun up) using an external service like binder.

One way that is faster and easier to consume is hosting a static version of the notebook having already been run and showing the code inputs and outputs.

Also, instead of showing each of the examples in isolation in their own environments and flavors, it can help to have the ability to switch between programming languages or dialects such as this example from spark: https://spark.apache.org/docs/latest/sql-data-sources-parquet.html That lets folks read in the language or dialect of their choice and to compare and contrast feature maturity between versions.

snazy commented 2 years ago

@XN137 isn't this what you've merged recently?

XN137 commented 2 years ago

In https://github.com/projectnessie/nessie-demos/pull/293 we made it so the notebook files in the repo keep the output from the last run.

so potentially in our docs we could include links to github like so: https://github.com/projectnessie/nessie-demos/blob/main/notebooks/nessie-iceberg-demo-nba.ipynb

there users could follow along and see the output etc, without having to run things themselves in mybinder (which right now had a 503 service problem for me). wdyt?

not sure if this fully aligns with the idea of "hosting static versions" from the issue description or the last paragraph about switching between flavors easily.