@SachaGreif@GatsbyJS I guess it depends on where the build happens, locally, you could probably store this on the FS or even SQLite... If this should happen only on CI, you often have the ability to cache artifacts (for example GH actions), if this is meant to run on various envs, remote caching.
Maintaining a lib, I found it difficult to have an overview of its lifecycle & various metrics, especially when it contains several packages, so I decided to create my own thing. A first widget showing when packages were initially released. #react#dataviz
@techgirl1908 https://t.co/OWctgTA6QR, I'm not super objective about it obviously 😅, but even as a maintainer of the lib, I use it quite often myself to play with some charts. For the why? I think interactive docs/playgrounds are the future 🚀
@onionhammer Adding this to the interactive playground is a bit difficult, then you might have invalid JSON, incorrect data types... which involves a bunch of extra UX, I'm more thinking about providing a different tool for that, some kind of Chart Builder App. Could be something like this.
@bbeckshack Yes, you have 2 behaviors, you can see the count for a node, for example for "Would Use in 2018", and you can also see the counts for a transition when hovering a link, for example "Would Use in 2018" -> "Would Not Use 2019" with the counts.
Working on a potential new chart for the #StateOfJS to visualize a cohort analysis for a technology across the years (using fake data atm 😁). What do you think?
@bbeckshack Thank you for the feedback, I agree, it's a little busy, that's a good idea, I could highlight a specific flow on hover for example, and 'mute' other nodes/links.
@chris_07734 Thank you for the feedback! The number of respondents varies from year to year, so the height depends on this. I was worried the gradients would be confusing for new responses, I'm glad it's not 😅. For now it's a prototype using fake data, might look quite different in the end.