[mapserver-dev] msTransformShapeAGG vs. msTransformShape

[Steve Lime]

I didn't try to figure out why some features disappeared between 5.2 and 5.4. The msTransformShape
function has been stable so it's elsewhere, perhaps the default symbol size change from 0 -> 1 had an
effect and features are being rejected elsewhere? Would have to dig further.

The change between  4.10 and 5.0 was done based on some benchmarking DMSolutions did if I recall. It 
removed a test for removing colinear points I believe. Should test all 3 methods with GD I think. A 5.4.1 or
6.0 effort.

The effects are subtle and I only noticed because we have a simple map that draws lakes for the whole
state but only 20 of them so you notice if one goes missing.

Steve

>>> Thomas Bonfort <thomas.bonfort at camptocamp.com> 04/21/09 12:41 AM >>>
Do you have an idea why the change in behavior between versions since 5.0 ?

> The msTransformShapeAGG() seems to function fine and I'm wondering why we don't just use
> it everywhere. I don't see anything in the function that seems to be AGG specific.

It isn't agg specific, but it produces a series of points suitable for
renderers that can treat fractional pixel values, whereas the gd one
trims points that fall on the same pixel. We should benchmark the
perfromance penalty that will/might affect the gd renderer if we make
it use the agg function before deciding if we can switch.

thomas

-- 
www.camptocamp.com
+33 4 79 26 57 97
_______________________________________________
mapserver-dev mailing list
mapserver-dev at lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/mapserver-dev