Library is open source!
October 27, 2019
I spent last summer interning at the New York Times on the Interactive News Team. It’s a really fantastic group of people who build tools for the newsroom: data pipelines for election night live coverage, NYT’s anonymous tip collection system, self-serve tools for reporters to make their own interactive stories. My first project involved working with the obituary desk— I built a tool that powers their yearly Notable Deaths interactive. It was a great lesson in building human-centered software.
My other project was Library, a tool for collaborative documentation with Google Docs. It was open-sourced earlier this year! The New York Times uses it to share information across the newsroom. To use it, you simply configure Library with a shared folder or team drive. It will then ingest everything to create a searchable, organized interface, where you can view the content without having to go to Google.
As an open source project, I’m excited to see where Library can be helpful. I’ve been a part of many organizations (particularly in college) where we lacked a place for relevant institutional knowledge, and something like this would be lightweight and helpful. You don’t need much coding knowledge to get up and running: to deploy to Heroku, you just need to click a button and configure some Google environment variables. NYT is partnering with Northwestern University’s Knight Lab to test and pilot the tool.
I love this project because it met our users where they were. While there wasn’t a way to easily share up-to-date information across the newsroom, Google Docs was already widely adopted. Library solved the former problem quite nicely, without prescribing a radically new workflow. Teams could easily add to and maintain their existing docs on Library. It was a thoughtful approach that reflected deep knowledge of and empathy for its users. This is always how I want to build software (and I’m lucky to get to do that now at Stripe!).
To read more about Library, check out our post.