For months I kept doing something embarrassing when I used my own star map tool. I’d lean toward the screen.
The star map lives at ricoordonio.com/tools/star-map/. You give it a date, a location, and a few display options and it draws the night sky for that moment. Stars as white points on navy, sized by magnitude, labeled where they have names. I built it partly because I wanted to know what I was looking at whenever I stepped outside at night, and partly because canvas-based astronomical projection seemed like a fun problem.
The leaning happened when I was trying to identify a specific star. I’d see a bright one near what I thought was Orion, suspect it was Betelgeuse, and then realize I couldn’t read the label without having my face eight inches from the monitor. There were four other stars crowded into the same visual quadrant, and the label text was small enough that I needed to squint past Rigel to confirm what I was already 80% sure about.
The obvious fix was a magnifier. Circular overlay, follows the cursor, 3x zoom. Standard practice for dense data visualization. I’ve seen it on medical imaging software and map applications. Star maps are data. Made sense.
Here’s what I didn’t figure out until I was using it: a magnifier on a star map is backwards.
When you use a real telescope, you’re magnifying the object itself. More glass, more detail, the star gets larger. With a bright enough target and enough aperture you can resolve a disk instead of a point.
That’s not what happens with a software star map. The stars are already points. There’s no sub-point detail to reveal. All I got when I zoomed in 3x was the same points, bigger, with fewer of them visible at once. I wasn’t seeing more of Betelgeuse. I was just removing its neighbors from frame so I could finally read the label without Rigel in the way.
What I built was disambiguation. The circular lens clears visual noise from a small region so you can isolate one object. It looks exactly like a magnifier and operates nothing like one.
I realized this the first time I used it, when I noticed I was hovering to exclude instead of hovering to zoom. I wasn’t trying to see the star better. I was trying to see the star alone.
The tool works. I use it. I’m not going to rename it a “noise reducer” or add explanatory text. But I think about it now whenever I see someone lean toward a screen. That posture is diagnostic. Somebody built the thing right and left out one step.
If users are moving their bodies toward your interface, that’s not a screen size problem. That’s a density problem. Find what their body is doing and use it to make your tool better.