[mapserver-commits] r8455 - trunk/docs/development/rfc

svn at osgeo.org svn at osgeo.org
Sun Jan 25 12:29:53 EST 2009


Author: tbonfort
Date: 2009-01-25 12:29:53 -0500 (Sun, 25 Jan 2009)
New Revision: 8455

Modified:
   trunk/docs/development/rfc/ms-rfc-49.txt
Log:
update rfc49 with status, comments from review period, and voting history


Modified: trunk/docs/development/rfc/ms-rfc-49.txt
===================================================================
--- trunk/docs/development/rfc/ms-rfc-49.txt	2009-01-25 17:29:16 UTC (rev 8454)
+++ trunk/docs/development/rfc/ms-rfc-49.txt	2009-01-25 17:29:53 UTC (rev 8455)
@@ -4,10 +4,10 @@
 MS RFC 49: Symbology, Labeling, and Cartography Improvements
 ======================================================================
 
-:Date: 2008/12/03
+:Date: 2009/01/26
 :Authors: Thomas Bonfort, Stephen Lime
 :Contact: Thomas.Bonfort at camptocamp.com , Steve.Lime at DNR.State.MN.US
-:Status: Draft
+:Status: Passed and Implemented
 :Version: MapServer 5.4
 :Id: $Id$
 
@@ -36,7 +36,7 @@
 These keywords can reduce mapfile size and maintainance, by avoiding
 creating multiple layers or classes.
 
-http://trac.osgeo.org/mapserver/ticket/XXXX
+http://trac.osgeo.org/mapserver/ticket/2865
 
 
 OUTLINEWIDTH on line layers
@@ -73,30 +73,16 @@
 whatever the map extent or scale, this line layer would always be rendered by a
 red line with a one pixel black outline.
 
-http://trac.osgeo.org/mapserver/ticket/XXXX
+http://trac.osgeo.org/mapserver/ticket/2865
 
-add LABELMETHOD to layerObj
-------------------------------------------------------------------------------
-The current algorithm ensures that a label point is *in* a polygon, but that
-point is not always optimal. A better approach would be to compute a true
-centroid or center of gravity instead of using the bbox center. 
 
+add LABEL to layersObj
+--------------------------------------------------------------------------------
 
-(or perhaps just method to labelObj?)
- 
-- this would allow users to choose label placement algorithms that meet their
-  needs. For example, in D2 below improving label placement will come with a
-  reduction in performance. In some situations simply using a shapes bounding
-  box center is perfectly adequate.
+.. note::
+  
+  This feature has not yet been implemented
 
-- The default would be the improved placement.
-
-http://trac.osgeo.org/mapserver/ticket/2793
-
-
-
-add LABEL to layersObj
---------------------------------------------------------------------------------
 so that a default can be defined for all classes (sort of like templates). A
 class label would take precedent.
 
@@ -125,3 +111,39 @@
 
 None Expected
 
+Comments from Review period
+--------------------------------------------------------------------------------
+
+- the LABELMETHOD proposal is withdrawn
+- fractional values also for pattern,gap, offset:  a good candidate
+  for 6.0 (when pattern and gap would be moved to the styleObj?)
+- there was concensus for adding parametered scaling, usefull for
+  thematic mapping (eg size proportional to a feature attribute)
+  
+  - by allowing simple expressions, eg SIZE  ( 18 + [poptotal] * ((
+    43-18)/(5000-30000))  )
+  - or by adding pseudoitems at the layer level:
+    ::
+      
+      PSEUDOITEM
+       'mySize' (mrange([itemname], mininput, maxinput, minoutput,
+      maxoutput))
+       'myText (commify(round([itemname], 2)) + ' ac')
+      END
+      
+      LABELITEM 'myText'
+      CLASS
+       STYLE
+        COLOR 255 0 0
+        SIZE [mySize]
+       END
+       LABEL
+        ...
+       END
+      END
+      
+   this enhancement will be addressed in a specific rfc.
+
+Voting History
+--------------------------------------------------------------------------------
++1 : SteveW, UmbertoN, TamasS, SteveL, DanielM (,AlanB)



More information about the mapserver-commits mailing list