32:T46b,slide`

Different Groups of People!

Typical Dataset Owners

Typically project managers whose main responsibilty is to get the data out, cataloged and secured.
Any code released is done through different system & by different people.
If government or big business, they are likely prohibited from linking to outside code that uses the data by rules that prohibit any appearance of "endorsement".

Typical Code Owners

Interested in their own applications, not all applications of the dataset.
Only very recently faced any demand from academic publishers to show where exactly the data came from or include access to it.
Each will pick a different way and place to reference the data if they reference it at all.

SWUNG deals with both datasets & code. It also has less difficulties in terms of being accused of endoresement. This puts it in a unique position.

`33:T6c0,slide`

Let's Fork Someone Else's Open Source Catalog

Thanks Lawerence Livermore National Laboratory

SWUNG VERSION!

Forked repo for SWUNG

live site with exploration of Subsurface Geoscience code project metadata connections

Parts of the Project

Python code to get metadata from GitHub API (used as is)
YAML to say which repositories & organizations to look at (you can submit pull request to get the repos you care about in here! Thanks to those that have!)
D3.js code for all the visualizations (used as is)
Jekyll & Angular site (painful to fix all the links so it can be reused. That's why a few parts still say LLNL and not SWUNG)

`34:T59d,full = html``35:T706,// Code adopted from here: https://observablehq.com/@mircopp/import-of-notebook-as-a-runtime-environment srcdoc = (notebook, cells, opts = {}) => { const { css = '' } = opts; return ` ${cells.map((cell, index) => `

`).join('')} ${notebook} ` }36:[{"id":"08694994414f77a48b2d343e0e7cfd34e128af6cde1702f99033d8e060ec86ecc2c9d0c8daec4c44167d369fe9919f38736d418a5f0151a628ccd63cc72d970b","href":"https://static.observableusercontent.com/files/08694994414f77a48b2d343e0e7cfd34e128af6cde1702f99033d8e060ec86ecc2c9d0c8daec4c44167d369fe9919f38736d418a5f0151a628ccd63cc72d970b","name":"Screen Shot 2021-04-22 at 11.46.06 PM.png","type":"image/png","lastModified":1619153209978.451,"size":152424},{"id":"f2feecb6a8c0c5895b895ecbe439d0675196f8f03301e9ef07091faef2ccfe083b4532b9b12828d50ca2859fad2d337a30dee47fb2dc51c3e77b25cd18762513","href":"https://static.observableusercontent.com/files/f2feecb6a8c0c5895b895ecbe439d0675196f8f03301e9ef07091faef2ccfe083b4532b9b12828d50ca2859fad2d337a30dee47fb2dc51c3e77b25cd18762513","name":"Screen Shot 2021-04-20 at 9.02.06 PM.png","type":"image/png","lastModified":1618970801346.903,"size":944975},{"id":"f865fc83e368b687a6a31ddd316a502e7272a57368c053af103b6260c94a5b8692461c53e533186165b4b823aa63e4b450078c2aceb94506e09ed98b80cfc4a1","href":"https://static.observableusercontent.com/files/f865fc83e368b687a6a31ddd316a502e7272a57368c053af103b6260c94a5b8692461c53e533186165b4b823aa63e4b450078c2aceb94506e09ed98b80cfc4a1","name":"Screen Shot 2021-04-22 at 11.53.38 PM.png","type":"image/png","lastModified":1619153640357.255,"size":231552}] 30:{"title":"Visualizing Connection Between Code Projects Could Nudge Developers to Contribute in More Impactful Ways","cells":[{"id":1542,"value":"md `# Visualizing Connection Between Code Projects Could Nudge Developers to Contribute in More Impactful Ways`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1028,"value":"slide`\n

More Visible Connections Between Projects Can Nudge Developers in Ways that Stregthen Open Soure Communities

\n\n

Real World Thing <=> Code

Code <=> Dependency

Code <=> Dependency <=> Code

Data <=> Code

\n
\n

A Lightning Talk for Transform 2021

A geoscience + coding online conference of https://softwareunderground.org/

by Justin Gosses

\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":988,"value":"slide`

You've heard of developer tools

IDE (Integrated Development Environment)

Source Control (GitHub, GitLab, etc.)

\n\n

And there's developer relations tool for individual projects/tools

Discord

Twitter

What about tools that build up communities across projects?

Transform! This conference

More simple, visible Connections Between Code Projects, Contributors, and Data <== THIS TALK

\n\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":942,"value":"slide `\n

Skeptical?

\n
\n

That's okay. I have an example of cross-project visibility that accelerates the community

\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1011,"value":"slide`\n

Awesome Open Geoscience

Condenses a lot of hard earned knowledge into a transferrable form, a list with descriptions & links contributed by the wider community of tools they actually use.

\n\n

\n 50 contributers \n

\n 708 stars\n

\n Shortens time to start\n

\n Helped people discover other SWUNG resources\n

\n\n

\n\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1515,"value":"html `

`","pinned":true,"hidden":false,"mode":"ojs"},{"id":1293,"value":"\n\nslide`\n

Now let's talk about making easier connections between open data and the code to use it

\n
\n
\n

Real World Thing <=> Code

Code <=> Dependency

Code <=> Dependency <=> Code

Data <=> Code

\n\n\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1154,"value":"slide`\n

\n\n

Hypothysis 1:

\n
\n
\n\n

It would save people a ton of time if they could easily find code already written for an open dataset they might be interested in.

\n
\n

The technology already exists to build direct links between open data & open code

\n
\n

...but the culture to do so doesn't exist

\n\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1176,"value":"$32","pinned":false,"hidden":false,"mode":"ojs"},{"id":1193,"value":"slide`

A PROPOSAL

1. Add \"dataunderground\" topic tags to code repositories that use data from dataunderground.

\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1217,"value":"FileAttachment(\"Screen Shot 2021-04-22 at 11.53.38 PM.png\").image()","pinned":true,"hidden":false,"mode":"ojs"},{"id":1199,"value":"FileAttachment(\"Screen Shot 2021-04-22 at 11.46.06 PM.png\").image()","pinned":true,"hidden":false,"mode":"ojs"},{"id":1286,"value":"slide`\n

Ta Da!

\n
\n

Now there is a way to find all the code that use datasets cataloged on dataunderground

\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1197,"value":"slide`\n

A SLIGHTY MORE COMPLICATED PROPOSAL

(1) Add \"dataunderground\" topic tag to code repositories that use data from dataunderground.

\n
\n

(2) Add an additional topic tag with the name of the dataset on dataunderground, so that way its easy to find all the code that leverages that dataset.

\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":1208,"value":"slide`\n

This is dataunderground!

If you own the dataset, you can easily add any key:value pair, including one for \"github_topic\"

\n\n