Turning an ArcGISDynamicMapServiceLayer into a FeatureLayer

[SpencerWilliams]

We have pleasantly surprised by many area County Appraisal Districts and the data that they have available.  I modified the dynamic legend code from the code gallery to suit our needs and be able to display things like landfills, recycling centers, active water wells, etc.  The problem is, there is a ton of data (attributes) that come with the overlays.  I was displaying them via ArcGISDynamicMapServiceLayer and that worked great.  However, I want my ArcGISDynamicMapServiceLayer and my MapTips too.  I have been successful in creating a transparent feature layer that sits on top of the ArcGISDynamicMapServiceLayer that displays the overlay with MapTips along with the dynamic legend.  It works great, setting the *NEW" OnDemand mode is also quite handy as there can be lots of data returned.  The new problem occurs when I have multiple layers - ie layer 0 = active wells, layer 1 = inactive wells in the same overlay.  The ArcGISDynamicMapServiceLayer handles this by design, but to do it via feature layers cleanly, I am stuck.  I was about to start coding a function that would tell me how many layers there are via layerDefinitions in the ArcGISDynamicMapServiceLayer and then create a feature layer (ie some awful code like someOtherLayer.Url = url + "/0" , + "/1" , + "/2" , + "/3" and so on and keep track of them somehow when I figured I should ask if there was a better way to do this.  Of course I could always edit the MXD to have them be sublayers/ or fused?  But I???m more of a programmer than a GIS guy, not to mention we would have to re-do it every quarter for every overlay for every county/datasource???.not my idea of fun.

Thanks,

Spencer


Forestar Group USA
NYSE:FOR

[dotMorten_esri]

From the sound of it, I have a feeling what you are trying to do is going to cause a lot of headaches, so before we go any further into a solution, I want to raise a concerns:
It sounds like you are going to be downloading an ENOURMOUS amount of data under the covers. You are downloading both geometry and a bunch of attributes for all the layers as the user zooms and pans, even though the same data is already present in the dynamic layer, and the user might not even be interested in the specific data. This could very quickly become a HUGE bottleneck, both on the network and performance-wise on the client.

As an alternative, I suggest you use the Identify task on the map. When the mouse stops over a certain point on the map for for instance .5 seconds, you automatically perform an identify using the IdentifyTask and pop up a custom maptip window next to the mouse. This has the benefit of not overloading the network, your server, the client's framerate, and only requesting the data that the user was interested in. Would that not cover your use-case?

[SpencerWilliams]

Morten,

As always, thanks for the quick response.  Your solution would cover my use case; however, I see 2 potential problems with this.  Let's say I have 3 overlays that are present at the same time: 1 shows wells, 1 shows landfills and the other shows tax parcels.  Because the Identify task needs to be tied to the map control (correct me if I'm wrong here), there is no way for me to know which service to query.  I would end up having to query against 3 services every time the map is clicked - be it for panning, zooming, etc.  I could limit the click event to only be active when the user chose a tool like "Identify Feature Tool" and while that was selected it would += the event and on tool "drop" then -= the eventhandler.  Back to my scenario - a well on top of a parcel would always return both (unless I made the user choose the overlay layer that they wanted the data from - haha, writing this post is giving me more ideas :))  I would need to do something like the Identify example shows with a drop down for each layer's info that gets returned.  All in all, it would save me the bottleneck that you describe.  Because I am talking about maptips here, it could just display a stackpanel for each layer that returned results.  I think I will go the Identify route, although if you have any comments on this post feel free to let me know.

Thanks,

Spencer

Forestar Group USA
NYSE: FOR

[SpencerWilliams]

Ok the identify task is definitely the way to go.  However, because the layer that I am clicking on (I know it's the map that handles the click, but anyway) doesn't support Graphics - how do I create a map tip for the data that is returned?  The data is a graphic itself - the only way I can think to do it is:

clear graphics in a tempLayer
create a graphic where their mouse is in the tempLayer
then identifyTask_ExecuteCompleted fires and adds the maptip to the only graphic that is in that tempLayer (the one that I just added)
and every mouse pause (.5 seconds) does this again and again while an overlay layer is present

is that the approach?

[dotMorten_esri]

You don't create a MapTip as such. You just instantiate a usercontrol that contains the information you need, add it to the page, and position at the click position (you can use e.GetPosition(parentOfMyCustomMapTopControl) to get that location). Use the margins/alignments on the control to position it.