Skip to content

Latest commit

 

History

History
106 lines (74 loc) · 3.67 KB

README.md

File metadata and controls

106 lines (74 loc) · 3.67 KB
Raw

silverorange Retreat Charts

This contains a bunch of custom charts rendered with D3 that can be exported and added to slides.

The charts are rendered using D3 and there is a lot of control over the visual style of charts.

Charts are rendered as SVG on a HTML canvas so you can right-click and save the charts as PNG. This makes it easier to import charts into slide software. Charts are rendered at 3x normal size so the exported PNG assets look good on high-dpi screens.

Running

yarn install
yarn start

Editing Data

Chart data comes from src/data.ts. Each exported dataset causes a separate chart to be rendered.

Site Release Data

Site release history is taken from Slack channels and cleaned up. The Slack history is manually copied from relevant Slack Channels and then cleaned up using grep to remove non-release messages and fix missing user-associations.

In future, a cleaner approach may be using a Slack data export for a specific year. There does not seem to be an option to export a single channel.

Newer projects use GitHub actions to deploy on merge:

  • air-institutes
  • es-translation-unit
  • emrap-go-client
  • emrap-go-server

Pull Request Data

This data comes from a series of GitHub GraphQL queries. Because the data comes from GitHub, if repositories are transferred outside our organization visibility they will disappear from the stats.

For organizations where there are multiple contributors outside of silverorange (i.e. HippoEducation) individual user queries are used to only track contributions from silverorange staff.

Since we are only interested in merged PR counts in a date range, we do not need to paginate data.

This data is currently collected semi-manually by updating the search date ranges for each year for each orginization and/or user. This could be scripted in the future.

Repository Data

This data comes from a series of GitHub GraphQL queries. Because the data comes from GitHub, if repositories are transferred outside our organization visibility they will disappear from the stats.

Since we are interested in repository creation date, we need to paginate data for the silverorange org where the number of source repos is more than 100.

A CSV is generated from the GraphQL data. Some repositories are incorrectly considered source forks by GitHub. These are manually updated in the CSV with a count, yes/no field. This is done so we can consistently ignore them year-over-year.

Android Releases

This data comes from the release history and current release information in the Google Play Console. Data is collected manually from the console for apps in the EM:RAP organization.

We stopped releasing the Hippo apps as of March 2022 with the inclusion of Peds RAP.

iOS Releases

This data comes from the release history in the Apple App Connect console. Data is collected manually from the console for apps in the EM:RAP organization.

We stopped releasing the Hippo apps as of March 2022 with the inclusion of Peds RAP.

Platform and Language Use

This data is assembled manually by looking at active sites and apps and cross-referencing with GitHub repositories. Mono-repos are expanded into separate projects during counting. If sites have been retired, the retirement date is noted in the manually assembled CSV file.

Only active sites and apps are considered. Libraries, packages, and inactive sites are not counted.

Platforms are counted separately if there is more than one project using the platform, otherwise they get categorized as other. Node and React platforms are somewhat arbitrarily grouped together:

  • Express, Apollo, jagQL, Node, Gulp -> Node
  • Create-React-App, Next.js -> React
  • React-Native -> React Native