[mapserver-users] trying new label positioning - no change
I'm trying the new label positioning improvements for polygons in trunk from http://trac.osgeo.org/mapserver/ticket/606 but AUTO seems to be ignoring the initial CC in the new ordering - with no collisions with other labels, all polygons are labelled at the UL corner of the polygon center. Are there special conditions for AUTO/CC to work? A typical label I'm trying is: LABEL COLOR 0 100 222 SHADOWCOLOR 218 218 218 SHADOWSIZE 1 1 OUTLINECOLOR 255 255 255 TYPE TRUETYPE FONT helvetica-italic SIZE 9 POSITION AUTO WRAP '|' PRIORITY 5 MINFEATURESIZE 16 END AUTO: inline: Picture-1.png POSITION CC does work to label polys at CC: inline: Picture-2.png PS. the new gravity positioning is nice! - William Kyngesburye kyngchaos*at*kyngchaos*dot*com http://www.kyngchaos.com/ Earth: Mostly harmless - revised entry in the HitchHiker's Guide to the Galaxy ___ mapserver-users mailing list mapserver-users@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/mapserver-users
Re: [mapserver-users] trying new label positioning - no change
William: In my testing a polygon layer is definitely starting at position CC. Any chance you're using an annotation layer here? I have found an issue if that's the case. The decision to using one set of positions vs another is made based on the layer type NOT the geometry type. So annotation layers are treated as point layers regardless of the source geometry. As the code sits now that information is lost. Wasn't an issue previously since we didn't vary things. I'll file a bug on this issue. Let me know if this isn't the case for you. One work around is to ditch annotation layers for polygons. They really aren't necessary with the addition of the label priority support in 5.0. Steve William Kyngesburye wokl...@kyngchaos.com 02/17/09 2:55 PM (Sorry, OSX mail habit, the images would be inline next to appropriate text descriptions) The first is POSITION AUTO. Not what I was expecting. The second is POSITION CC. What I was expecting for AUTO, now. On Feb 17, 2009, at 2:29 PM, Steve Lime wrote: So what are pictures 1 2 representing? There's certainly a difference and I would have expected picture 2 to represent the new labeling with CC as the first choice. Steve On 2/17/2009 at 12:42 PM, in message 7da25c07-8345-491d-b082-b268ee480...@kyngchaos.com, William Kyngesburye wokl...@kyngchaos.com wrote: I'm trying the new label positioning improvements for polygons in trunk from http://trac.osgeo.org/mapserver/ticket/606 but AUTO seems to be ignoring the initial CC in the new ordering - with no collisions with other labels, all polygons are labelled at the UL corner of the polygon center. - William Kyngesburye kyngchaos*at*kyngchaos*dot*com http://www.kyngchaos.com/ The equator is so long, it could encircle the earth completely once. ___ mapserver-users mailing list mapserver-users@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/mapserver-users ___ mapserver-users mailing list mapserver-users@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/mapserver-users
Re: [mapserver-users] trying new label positioning - no change
On Feb 17, 2009, at 11:02 PM, Steve Lime wrote: William: In my testing a polygon layer is definitely starting at position CC. Any chance you're using an annotation layer here? I have found an issue if that's the case. The decision to using one set of positions vs another is made based on the layer type NOT the geometry type. So annotation layers are treated as point layers regardless of the source geometry. As the code sits now that information is lost. Wasn't an issue previously since we didn't vary things. I'll file a bug on this issue. Let me know if this isn't the case for you. O. I've always used annotation layers for feature annotation. But, if it always treats annotation layers as point layers, why does ANGLE FOLLOW work on annotation layers of line features, or gravity positioning on annotation of poly features? Or does that trigger stepping out of the assumed-point default somehow? I did notice when poking around in the source that the labelling routines didn't know about the geometry of individual features. In my own early (v4.4) attempts at reordering the positions for polys (before the recent changes) I added a parameter to the addlabel() functions to pass the shapetype. One work around is to ditch annotation layers for polygons. They really aren't necessary with the addition of the label priority support in 5.0. I'm not sure what you mean here - how does label priority invalidate annotation layers? - William Kyngesburye kyngchaos*at*kyngchaos*dot*com http://www.kyngchaos.com/ Oh, look, I seem to have fallen down a deep, dark hole. Now what does that remind me of? Ah, yes - life. - Marvin ___ mapserver-users mailing list mapserver-users@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/mapserver-users
Re: [mapserver-users] trying new label positioning - no change
Comments inline... William Kyngesburye wokl...@kyngchaos.com 02/17/09 11:34 PM On Feb 17, 2009, at 11:02 PM, Steve Lime wrote: William: In my testing a polygon layer is definitely starting at position CC. Any chance you're using an annotation layer here? I have found an issue if that's the case. The decision to using one set of positions vs another is made based on the layer type NOT the geometry type. So annotation layers are treated as point layers regardless of the source geometry. As the code sits now that information is lost. Wasn't an issue previously since we didn't vary things. I'll file a bug on this issue. Let me know if this isn't the case for you. O. I've always used annotation layers for feature annotation. But, if it always treats annotation layers as point layers, why does ANGLE FOLLOW work on annotation layers of line features, or gravity positioning on annotation of poly features? Or does that trigger stepping out of the assumed-point default somehow? A shapes geometry is used when deciding where to compute a label point. In this case the we're deciding on what the acceptable position array is and that's done when processing the cache, long after the label points original geometry has been destroyed. I did notice when poking around in the source that the labelling routines didn't know about the geometry of individual features. In my own early (v4.4) attempts at reordering the positions for polys (before the recent changes) I added a parameter to the addlabel() functions to pass the shapetype. I think I'd just pass a shapeObj pointer to the msAddLabel() and harvest what is necessary: type and a few indexes. One work around is to ditch annotation layers for polygons. They really aren't necessary with the addition of the label priority support in 5.0. I'm not sure what you mean here - how does label priority invalidate annotation layers? The principle reason to use annotation layers was to control the order in which labels are stuffed into the cache. They were especially useful with polygons since you typically draw those first but if you want to emphasize labels then you'd have to use annotation to make the labels show before other stuff. Label priorities fix that. You can give a polygon layer high priority labels without having to resort to annotation layers later in the mapfile. The benefits (besides working around this bug) are two-fold: 1) performance, you only have to process the polygon features once... 2) brevity, shorter, more concise layer definitions are good... Certainly annotation layers are still useful but principally for things like copyright notices or road shields. Steve ___ mapserver-users mailing list mapserver-users@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/mapserver-users
Re: [mapserver-users] trying new label positioning - no change
On Feb 18, 2009, at 12:15 AM, Steve Lime wrote: I did notice when poking around in the source that the labelling routines didn't know about the geometry of individual features. In my own early (v4.4) attempts at reordering the positions for polys (before the recent changes) I added a parameter to the addlabel() functions to pass the shapetype. I think I'd just pass a shapeObj pointer to the msAddLabel() and harvest what is necessary: type and a few indexes. If you think this is something that can be done, it would be nice have geometry and annotation layers both use the same positioning rules based on the geometry type (see comments below). I'm not sure what you mean here - how does label priority invalidate annotation layers? The principle reason to use annotation layers was to control the order in which labels are stuffed into the cache. They were especially useful with polygons since you typically draw those first but if you want to emphasize labels then you'd have to use annotation to make the labels show before other stuff. Label priorities fix that. You can give a polygon layer high priority labels without having to resort to annotation layers later in the mapfile. The benefits (besides working around this bug) are two-fold: 1) performance, you only have to process the polygon features once... 2) brevity, shorter, more concise layer definitions are good... Certainly annotation layers are still useful but principally for things like copyright notices or road shields. Steve Hmm, but mixing the annotations with their geometry in a single layer means you can't turn geometry and annotations on and off separately - I often want to show just the feature geometry, yet be able to turn on their annotations if I need it. Also, what about the drawing order? Won't other feature layers then draw on top of earlier feature layers and annotations of those layers? Or does the label cache force ALL annotations to draw last, no matter what layer they're on? I guess I need a visual diagram of the [now many] various possible combinations of layer types, annotations, priorities and layer ordering to make sense of this. - William Kyngesburye kyngchaos*at*kyngchaos*dot*com http://www.kyngchaos.com/ Time is an illusion - lunchtime doubly so. - Ford Prefect ___ mapserver-users mailing list mapserver-users@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/mapserver-users
