[mapserver-dev] strange behavior when size != gif size (UMN 5.4.2)

[Steve Lime]

What output driver? GD or AGG?

>>> On 8/20/2009 at 10:05 AM, in message
<1250780754967-3479658.post at n2.nabble.com>, teeschke <daniel.teske at web.de>
wrote:

> Dear List..
> 
> actually I'm comparing my current used Mapserver (5.0.2) with the actual
> (5.4.2) and I detected a strange side effect when a size assigned to a gif
> symbol:
> 
> size < gif size: the symbol gets brightly
> size > gif size: the symbol gets darkly
> 
> here is my sample code for your better reproduction:
> 
> Part of my Symbolset:
> 
> SYMBOL
>     NAME 'some_gif_symbol'
>     TYPE PIXMAP
>     IMAGE "some_gif_symbol.gif"
> END
> 
> Part of my Mapfile:
> 
> LAYER
> 	NAME 'dw_mg_geom_view'
> 	STATUS ON
> 	TYPE POINT
> 	DATA "some_shape"
> 	TEMPLATE 'foo.html'
> 	LABELITEM 'KARTO_NR'
> 	CLASS
> 		NAME '1-3 (size < gif size)'
> 		EXPRESSION (1 <= [KARTO_NR] AND [KARTO_NR] <= 3)
> 		STYLE
> 			SYMBOL 'some_gif_symbol'
> 			SIZE 10 #equals gif size -3
> 		END #STYLE
> 		LABEL
> 			FONT arial
> 			TYPE TRUETYPE
> 			SIZE 9
> 			POSITION AUTO
> 			COLOR 0 0 0
> 		END #LABEL
> 	END #CLASS
> 	CLASS
> 		NAME '3-6 (size = gif size)'
> 		EXPRESSION (3 < [KARTO_NR] AND [KARTO_NR] <= 6)
> 		STYLE
> 			SYMBOL 'some_gif_symbol'
> 			SIZE 13 #equals gif size
> 		END #STYLE
> 		LABEL
> 			FONT arial
> 			TYPE TRUETYPE
> 			SIZE 9
> 			POSITION AUTO
> 			COLOR 0 0 0
> 		END #LABEL
> 	END #CLASS
> 	CLASS
> 		NAME '6-9 (size > gif size)'
> 		EXPRESSION (6 < [KARTO_NR] AND [KARTO_NR] <= 9)
> 		STYLE
> 			SYMBOL 'some_gif_symbol'
> 			SIZE 16 #equals gif size +3
> 		END #STYLE
> 		LABEL
> 			FONT arial
> 			TYPE TRUETYPE
> 			SIZE 9
> 			POSITION AUTO
> 			COLOR 0 0 0
> 		END #LABEL
> 	END #CLASS
> END #LAYER
> 
> the 13x13 gif symbol:
> http://n2.nabble.com/file/n3479658/some_gif_symbol.gif 
> 
> the getmap result:
> http://n2.nabble.com/file/n3479658/getmap.png 
> 
> This phenomenon happens only in the 5.4.2! the 5.0.2 works correct. 
> In the Changelog I found something: "Made sure computation of default symbol
> size is consistent with older versions (#2963)"
> 
> 
> Thanks and best regards..
> ..Daniel