Join GitHub today
Content-Aware label collision #1150
When rendering map with a lot of cities, administrative boundaries, subway stations and dualway-roads it's typical to have repeating labels.
The idea is to remember the text for each label (in addition to bbox), and if one of possible label positions intersects with same-text label (configurable: just text, text+font, text+font+color, text+font+color+size), just omit this text.
It will make dealing with double-labelled subway stations a lot easier. :3
Also, the feature should work transparently with shields, letting user tell "in some cases it's OK to render just shield image without text".