It would be nice if data visualization would become more independent from the technology that produces the data.

I tried to create some abstraction layer in [1] (see playground at [2]), it powers my open source debug visualizer extension for vscode. It basically bundles a few popular libraries and establishes a common data format that abstracts from both the visualization and the data source.

I feel like data visualization reinvents itself too often. The table in the authors article is nothing compared to what perspective js offers with advanced filters/group options etc. [3]. There is also SandDance [4] which offers a lot of advanced visualizations.

I would love to see something like the language service protocol for visualizations!

[1] https://github.com/hediet/visualization [2] https://hediet.github.io/visualization/?darkTheme=1 [3] https://perspective.finos.org/ [4] https://marketplace.visualstudio.com/items?itemName=msrvida....

> I would love to see something like the language service protocol for visualizations!

Then have a look at vega and its abstraction layer vega-lite [1]. It is based upon a grammar which allows you to constain the kind of visualization you would like to have. The rest is filled in by the vega-lite compiler. I've been toying around with it lately, and I am really impressed by it. [1]: https://vega.github.io/vega-lite/examples/

This looks great - I hadn't seen [4] before either, going to try it out

This whole article looks like an ad for authors' book and Chord PRO library.

The crate doesn't perform plotting, it only does a ureq HTTP call to the author's web service for data visualization.

I found it a little overwhelming too. I was hit with an add 3 times by the time I hit the bottom of the article. Kind of too aggressive in my book

it's a free article taken from the book using a free rust crate which has a pro version

it's a long running series of numerous high quality articles

Nothing wrong with that as long as the article is still useful

this is the hn way

I've bought a book from this person ("Data is beautiful"), and it's crap. If you go to the website home page, you'll see that the only thing that it does is use one (The author's?) library.

Hello - I am the person :)

You're right of course - 6 of the 11 parts in the book are dedicated to chord diagrams. Shortly after starting the book I started exploring chord diagrams again and was enjoying creating visualisations with them a little too much.

Please bear in mind that it's an early access book that I'm still working on. I can see you grabbed a copy when it was listed at 7.49 (it increases as I add more), and I have to defend it a little and say I don't believe the book is crap, just incomplete! I have been working on custom visualisations which will be working their way into the D3.js and Data is Beautiful book soon.

Anyway, thank you for the feedback, I have sent an email to offer a refund, because selling the books is supposed to support further work, and I don't feel comfortable if someone feels unhappy after doing that.