More readable code + factors #1
Conversation
src/location-distance-utils.c
Outdated
|
|
||
| static double r2d(double rad) | ||
| // compensate for shrinking longitude towards poles | ||
| float lng_scale(double lat) |
There was a problem hiding this comment.
static? also, why float and not double?
There was a problem hiding this comment.
Ah! it's my bad, it should be double,dist type was float in my head for some reason.
I'm not sure what you mean by, the original r2d was static.
src/location-distance-utils.c
Outdated
| double longitude_s, | ||
| double latitude_f, | ||
| double longitude_f) | ||
| double longitude_s, |
There was a problem hiding this comment.
extra change, please squash with the next commit where it is fixed.
There was a problem hiding this comment.
That's my formatter doing :p
src/location-distance-utils.h
Outdated
| @@ -24,6 +24,8 @@ | |||
|
|
|||
| G_BEGIN_DECLS | |||
|
|
|||
| float lng_scale(double lat); | |||
There was a problem hiding this comment.
why is that exported? Is there any user of it outside location-distance-utils.c
There was a problem hiding this comment.
Ultimately yes, for terrain and bearing calculations, it shouldn't be long to add them.
src/location-distance-utils.c
Outdated
| dist = dist * 1.609344; /* kilometers */ | ||
|
|
||
| return dist; | ||
| float dlat = (float)(latitude_f - latitude_s); |
There was a problem hiding this comment.
I haven't verified the maths here (it was too long ago I was playing with trigonometric functions when I was student), so could you provide some evidence that either original code does it wrongly or the replacement calculates the same values given same input values.
There was a problem hiding this comment.
The original code uses Spherical law of cosines which gives good results down to distances as small as few meters,
what is used here is Equirectangular approximation which is accurate for small distances only but performs well.
of course there's Haversine which is accurate for all king of distances but considered relatively slow.
I could add Haversine calculations, what do you think?
here's a link talks about all three laws
|
I think this is fine. @Dastan-glitch If you could keep the existing formatting (it's done with And yeah, a better commit message would be nice. |
No description provided.