You need to log in before you can comment on or make changes to this bug.
The WMC document generated using the DownloadContext widget is invalid according to the schema. These are the respective wms_ metadata tags from my MAP file which are relevant, if I understood correctly Mapserver can only deal with empty styles which is why my wms_style is an empty string: "wms_stylelist" "" "wms_style_default_title" "" "wms_style" "" There is an empty StyleList in the WMC document, StyleList should always contain a Style element. If the StyleList is left out, the document validates. <Layer queryable="0" hidden="0"> <Server service="WMS" version="1.1.0" title="Geostreets"> <OnlineResource xlink:type="simple" xlink:href="http://blabla&"/> </Server> <Name>layer_1</Name> <Title>Geostreets</Title> <SRS>EPSG:28992</SRS> <FormatList> <Format>image/jpeg</Format> <Format current="1">image/png</Format> </FormatList> <StyleList> </StyleList> </Layer>
Why not just leave out the wms_style_list parameter in the mapfile?
That's a good point Daniel. This does solve the problem. Although I always think it is unpretty for an application to generate invalid XML. I have changed the resolution to WONTFIX.
Agreed that's not pretty, you may want to file a (low priority) MapServer bug about this specific issue: an empty list is an obvious case that should be handled better. Note that this whole parameter validation issue is a grey area, and there are many other ways to break the XML other than this one. The problem is that if we add too much validation in MapServer to catch all the possible invalid values then we make the code more complex than it needs to be, and eventually end up with something bloated, slow, and harder to maintain. Not an easy choice...
I have created a bug with a low priority in the Mapserver bugzilla for this. Reference is: http://mapserver.gis.umn.edu/bugs/show_bug.cgi?id=595
Since this is a mapserver issue, I will close this bug and future comment I assume will be tracked under the mapserver bug. http://mapserver.gis.umn.edu/bugs/show_bug.cgi?id=595
closing bug based on last comment
