Comment by seletskiy
12 hours ago
Dataviz suggestions from top of the page to the bottom:
- Location search & display: no English version, only localized names, which is hard to reason if you don't know native language.
- Summary: a) what does precip. in % mean and how it correlates with cm? b) pressure is only displayed in hPa, while some countries prefer mmHg; c) what does ozone index mean and why is it important?
- Next 24 hours: visual indication for temp variation through the day would be easier to reason than just by looking at numbers (as discussed in TFA).
- Next 7 days: a) unnecessary precision for y-axis (e.g. 10.9°C vs. 11°C); b) band overshoot actual values (e.g. if I see 10..−10°C, I assume that would be max/min temp, but in fact it is 8..−9°C, which is impossible to tell without hovering mouse over); c) no horiz. line through 0°C; d) no horiz. lines through y-axis ticks, which makes it harder to reason about values closer to the end of the graph; e) precip. in cm tells little, especially when band is alike (0.00..0.80 cm) - peaks on graphs look like a lot, in fact they are not? g) seeing blue precip. graph subconsciously means 'rain' to me, while in fact it would be snow; f) labels for y-axis are at the same time very small, rotated 90° and also take too much horizontal space from the graph.
- Map: moving mouse over next 7 days graph causes time shown on map to change that would make sense if map's timeline would cover all 7 days, but it only covers small part of today.
- Week: a) fog icons look like they have solid white square background, which seems to be off compared to other icons; b) low/high values are hard to reason about, especially when it says 'Low … at 11am' and there is no tick labeled '11am' (10am .. 12pm) - displaying a line through coldest/warmest hours with °C value next to it would be much easier to understand.
Also: displaying air quality prediction based on last year's AQ would be helpful.
No comments yet
Contribute on Hacker News ↗