This is a
very bizarre take. If you want to just blindly copy and paste d3 code, you may have issues with the docs being hosted on observable. But absolutely zero part of the d3 api (or core design patterns of use) have anything to do with observable.
It’s like saying, “I refuse to use PyTorch because their docs are built with mkdocs.”
Moreover, even if the coupling were limiting (which, again, it is not), it’s odd to attack observable since everything they put out is fully open-source. It’s not like they’re hiding docs behind a paywall, they’re actively contributing to the viz ecosystem through basically everything they release (eg Observable Plot, Observable framework).
As far as echarts goes, it’s a great tool. The declarative syntax for charting always feels a bit odd, but it’s easy enough to wrap into component libraries. AFAIK it still powers many of the big BI viz tools used today (eg AWS Quicksight).
It’s like saying, “I refuse to use PyTorch because their docs are built with mkdocs.”
Moreover, even if the coupling were limiting (which, again, it is not), it’s odd to attack observable since everything they put out is fully open-source. It’s not like they’re hiding docs behind a paywall, they’re actively contributing to the viz ecosystem through basically everything they release (eg Observable Plot, Observable framework).
As far as echarts goes, it’s a great tool. The declarative syntax for charting always feels a bit odd, but it’s easy enough to wrap into component libraries. AFAIK it still powers many of the big BI viz tools used today (eg AWS Quicksight).