having used claude extensively for the past couple of months, i think there is still a lot of value to understanding your code and steering it intentionally. the amount of incidental complexity claude accumulates otherwise (which mostly results in bugs that it can't ever fix reliably) is staggering.
Great article from @scientificdiscovery.dev on why data viz matters and how to do it well
www.scientificdiscovery.dev/p/salonis-gu...
ht @visualisingdata.com
Illustration of an individual wearing glasses comfortably leaning back on a cushion. Text reads: DELPHI TOOL. simple handmade conveniences. A collection of small, low stakes and low effort tools. No logins, no registration, no data collection. I can't believe I have to say that. Long live the handmade web.
This collection of free and useful tools is so wonderful
tools.rmv.fyi
These are so elegantly done
New in Svelte: Comments inside component/element tags!
Congratulations, Frank. Not sure if this is a physical move for you but either way, I'm envious of the weather you'll be enjoying
I identified with all 20 of @parabole.bsky.social's salient learnings
www.theplot.media/p/give-me-20
Impactful
Glad you like it. I've noted the request, but can't say when or whether I'll get to it.
I have enjoyed this post by Nikita Prokopov so much, as a Mac user, as a designer, and most of all as a stickler.
Lots of takeaways for UI and information designers, starting with "adding an icon to everything is exactly the wrong thing to do".
Before and after redesign of a credit monitoring email showing how intentional information design can make it easier for the reader to understand, with specific annotations describing the changes.
Each month I receive this email from TransUnion and each month I have to struggle to determine if everything is okay.
Intentional information design can make it better. #unsolicitedRedesign
The same two line charts showing City of Edmonton water usage during the 2010 Winter Olympic men’s gold medal hockey game. Each compares water usage with the previous day’s. The second however, annotates the chart showing how the peaks in water usage coincide with the ends of periods and thw valleys coincide with period starts and Canada’s winning goal along with the medal ceremony.
Annotating points of interest can provide crucial context and meaning, bringing the story of your chart alive to viewers.
#VizTips
The same two line charts showing a facetious data set of butternut squash data which hovers around 210K from 2018 to 2022 then sees a drop from 2023 to 2024 down to around 130K then returning to previous levels. The second chart gives viewers hints that the axis doesn’t start at zero by placing additional space below the data, removing the baseline, and adding a broken axis symbol to the y-axis.
It is fine to start your line chart axes at something other than zero, especially when small variations are important.
You can give hints that your axis doesn’t start at zero by
placing additional space below the data, removing the baseline, and/or adding a broken axis symbol.
#VizTips
Before and after visuals of data using a bar chart to compare 70+ values vs using a dot plot to compare those same visuals.
Using dots instead of bars can reduce the heaviness of a chart with many items.
#VizTips
A stacked area chart of viking revenue streams vs six separate area charts and a stacked bar chart of the same data. Patterns hidden in the stacked area chart are more easilly seen in the smaller area charts
Stacked charts can obscure patterns in data. Only the first series and total afford easy interpretation. Unless those two are more important than the rest, consider breaking things up into multiple smaller charts to understand both proportions and patterns for each of your series.
#VizTips
A vertical bar chart next to a horizontal bar chart. Both show the same set of data (McDonalds Burger Calorie Counts), but the horizontal chart’s bar labels are easier to read because the text reads left to right like normal, rather than each one tilted a ta 45 degree angle like in the the veritical bar chart.
Horizontal bar charts provide room for longer titles and keep your audience from having to tilt their heads.
#VizTips
Two maps of US unemployment rates by county. They are identical except that the legend in the second has been converted into a histogram.
Your map’s legend can be turned into a histogram, helping your audience understand the distribution in addition to decoding the colours.
#VizTips
Two charts showing population over time for 3 age groups (0.24, 25-64, 65+). The first uses a legend to show what the lines mean while the second puts the labels directly on the lines
Direct labeling makes it easier to quickly interpret the visuals, allowing the viewer to find the information they need right where they need it.
#VizTips
two maps, side by side, showing the same 3 locations near the Edmonton river valley. The first map has full colour satelite imagery in the background which makes the locations harder to locate and distinguish. The second has a desaturated simplified base map in the background making the locations much easier to distinguish from the background.
Maps are for context. Simplified, desaturated maps pushed to the background allow your data points to pop.
#VizTips
An image showing two bar charts of Automobile Units Sold in 2019 by automaker. The left image sorts automakers alphabetically while the right sorts the automakers by the number of units sold.
Sorting by the data values instead of alphabetically creates a cleaner chart and makes rankings immediately apparent. #VizTips
two images of a man’s back. the first, labelled ‘before’, is very hairy. The second, labelled ‘After’, is not hairy.
A thread with all my data visualization tips and redesigns.
Mainly for my own convenience.
It's the standard naming convention for tailwind colours, and while I don't use tailwind, I do like the naming convention. I think it started with 100 to 900 like the font weights and then in old school coding with BASIC fashion they added a 50 and 950 when moving to 11 colours.
My personal projects are my escape from reality—each one its own little world that rewards slow, diligent work. I enjoy watching them evolve from nothing or a forgotten something into a brand new thing for others to enjoy. I can’t remember a time when I wasn’t making something and I love it.
Thanks, Frank. It was quite fun to work on.
You'd have to look at how chroma.js does it's interpolation for how it handles out of gamut, but the short answer is no the tool does not provide any warnings for this.
A screenshot of the color tool showing oklch inputs of colour that look blue and pink and the resulting interpolation between them
Thanks to chroma.js it looks like the answer to the first question is yes
No, that sounds like a lot of work.
No I hadn't seen David's work before. Looks pretty cool.
If ever I can be of service to you and the phenomenal work you create, count me in.
Nicholas's process is intricate, exacting, and inspiring. That something I made might make its way into it brings me considerable joy.
