single classed layers and GetLegendGraphic

[Bart van den Eijnden]

Tom,

how does your client know how many rules a layer has, i.e. how does it
distinguish between single rule layers and multiple rule layers, if it
does not perform a GetStyles request to get the rules?

Best regards,
Bart

On Wed, 10 Nov 2004 08:43:07 -0500, Kralidis,Tom [Burlington]
<Tom.Kralidis at EC.GC.CA> wrote:

>> -----Original Message-----
>> From: bartvde at xs4all.nl [mailto:bartvde at xs4all.nl]
>> Sent: Wednesday, 10 November, 2004 02:33
>> To: Kralidis,Tom [Burlington]
>> Cc: mapserver-dev at lists.umn.edu
>> Subject: Re: [UMN_MAPSERVER-DEV] [Fwd: Re:
>> [UMN_MAPSERVER-DEV] single classed layers and GetLegendGraphic]
>>
>>
>> For reference, this was my original post on the wms-dev list.
>>
>> http://lists.eogeo.org/pipermail/wms-dev/2004-May/000518.html
>>
>> The answer Keith gave only applied to the case of multiple
>> rules, but I did not suspect this behaviour to be different
>> from the case where there is only one rule ... but apparently
>> it is in Cubewerx's implementation ...
>>
>> I do see Tom has a point now, but still think the spec is not
>> fully well-thought at the moment ...
>>
>> Tom, btw what is your purpose for only getting the legend
>> graphic without a label in the case of a single class layer?
>> Are you putting the wms_title next to the legend graphic in
>> the WMS client (or using no label at all?), because otherwise
>> you need to know the name of the Rule anyway to put it next
>> to the legend graphic (and use a GetStyles request to know this name).
>>
>
> Yes, I'm basically writing out a table like:
>
> Icon|Title
> Icon|Title
>
> ...so it's a bit redundant if the title is shown twice.
>
>> Btw2: does Cubewerx use the GetLegend request in their LegendURL?
>>
>
> No, it's a vendor specific thing they use in CubeXPLOR.
>
>> Best regards,
>> Bart
>>
>> >> As I wrote already, I agree that this may be a valid
>> interpretation,
>> >> and that was our initial interpretation of the spec.
>> However the spec
>> >> is vague on this and in bug 653 it was requested that we align
>> >> MapServer's
>> >> behavior with other (prior) implementations (cubeserv) so
>> >> that clients
>> >> can better interoperate.
>> >>
>> >
>> > Here are my CubeSERV observations:
>> >
>> > CubeSERV OGC:WMS 1.1.1 GetLegendGraphic requests:
>> >
>> > This is a layer with multiple classes:
>> >
>> >
>> http://ceoware2.ccrs.nrcan.gc.ca/cubewerx/cubeserv/cubeserv.cgi?servic
>> > e=
>> >
>> WMS&version=1.1.1&request=GetLegendGraphic&layer=CANADNAODA_L%
>> 3ACEOWARE2
>> > &format=image/png
>> >
>> > This is a layer with a single class (i.e. all point symbols are the
>> > same)
>> >
>> >
>> http://ceoware2.ccrs.nrcan.gc.ca/cubewerx/cubeserv/cubeserv.cgi?versio
>> > n=
>> >
>> 1.1.1&service=WMS&request=GetLegendGraphic&layer=GCDB_INDEX:CE
>> OWARE2&for
>> > mat=image/png
>> >
>> > The layer title does not show up as part of the image.  This is
>> > default behaviour.
>> >
>> > Here is the same layer (single class) with the layer name
>> embedded in
>> > image via 'GetLegend'
>> >
>> >
>> http://ceoware2.ccrs.nrcan.gc.ca/cubewerx/cubeserv/cubeserv.cgi?versio
>> > n=
>> >
>> 1.1.1&service=WMS&request=GetLegend&layers=GCDB_INDEX:CEOWARE2
>> &format=im
>> > age/png
>> >
>> > Here are both layers in a GetLegend:
>> >
>> >
>> http://ceoware2.ccrs.nrcan.gc.ca/cubewerx/cubeserv/cubeserv.cgi?versio
>> > n=
>> >
>> 1.1.1&service=WMS&request=GetLegend&layers=GCDB_INDEX:CEOWARE2
>> ,CANADNAOD
>> > A_L:CEOWARE2&format=image/png
>> >
>> > ..Tom
>> >
>> >
>>
>>
>
>



--
Using Opera's revolutionary e-mail client: http://www.opera.com/m2/