Search

[Shex]

This is literally nothing, but I noticed the scale of all the nosecount graphs are really weird, being binary and scaling it in weird steps (1-2-4-8-16-32-...).
That creates really unusable graphs because you can't compare them at all. Half the size of the graph doesn't mean half the amount of people.

Equal intervals would be way more useful there

[Fenrikur]

Thank you for the feedback, it is really appreciated! (:

Last year, complaints were that – due to the huge difference between certain data points – the resolution of the plots was too low especially in the segments with lower quantities. This lead to bars representing, for example, the number of attendees per country, to become close to invisible for countries with only a few attendees in comparison to those making up a majority of our attendees.
To further illustrate the issue, check out the attached screenshots from EF23 and EF24.

For this reason, we went for a logarithmic scale this time in order to provide "more detail where it matters", to quote a user on github.

[Zefiro]

Nice idea. And using 2 as base? The nerdy half of our attendees will love that

*purrr*