From 1d46e83e3087f0d392c74c40d6b562359ed91576 Mon Sep 17 00:00:00 2001 From: Bruno de Oliveira Abinader Date: Tue, 28 Jun 2016 10:23:48 +0300 Subject: [core] Use the right types for zoom scale logic 32 bit integers should be enough for zoom scale logic. In shape annotation logic, 'maxAmountOfTileFeatures' requires 64 bits because we are multiplying the zoom scale with the extent, which might give a number higher than std::numeric_limits::max(). --- include/mbgl/util/constants.hpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/mbgl/util') diff --git a/include/mbgl/util/constants.hpp b/include/mbgl/util/constants.hpp index 607c388e95..3c0b3eb93e 100644 --- a/include/mbgl/util/constants.hpp +++ b/include/mbgl/util/constants.hpp @@ -18,7 +18,7 @@ constexpr float tileSize = 512; * In practice, all features are converted to this extent before being added. * * Positions are stored as signed 16bit integers. - * One bit is lost for signedness to support featuers extending past the left edge of the tile. + * One bit is lost for signedness to support features extending past the left edge of the tile. * One bit is lost because the line vertex buffer packs 1 bit of other data into the int. * One bit is lost to support features extending past the extent on the right edge of the tile. * This leaves us with 2^13 = 8192 -- cgit v1.2.1