Representations and thematic maps

By Charlie Frye, Esri Chief Cartographer

Representations - thumbnail

This blog entry addresses one of the new features in the ArcGIS mapmakers’ suite of tools. At version 9.2, we added the ability to store representations in your geodatabase; this has wide ranging possibilities for map makers and data publishers. You will definitely be seeing more blog entries about other uses of representations in the coming months. Continue reading

Posted in Mapping | Tagged , | Leave a comment

How to track pending tiles and display a busy indicator in a Web mapping application

Rex Hansen contributed this post about how to use some of the enhanced JavaScript in Service Pack 2 to track pending tiles and display a busy indicator (such as an animated “Loading” graphic) over the Web ADF’s Map control: 

As a Web ADF developer working in an asynchronous communication environment, it is often beneficial to provide an end user with some indication that a user action is being processed. Since most Web ADF applications are centered on working with a map, the ability of an end user to effectively interact with map contents is essential. The Web ADF has the ability to asynchronously retrieve map data from multiple sources and consolidate it in a single map control. In general, data sources often differ in the time it takes to respond to a request. Since the Web ADF Map control is capable of rendering map data as it is returned to the browser, it’s possible that some portion of data in the map is visible and accessible before another portion. In this case, it will likely be important to let the end user know when the map control has finished loading map data from any and all sources.

To support this capability, 9.2 service pack 2 includes an enhanced Web ADF JavaScript Map object. The JavaScript Map object has a set of “event handlers” on the pendingTiles property. The pendingTiles property references an array of map image tiles to be rendered. The array is updated when the map needs new image tiles based on the current extent. Events on the pendingTiles property are listed below:

Event Description
add_onRequestsPending Triggered when the number of items in the pendingTiles array changes from 0 to a higher value
add_onRequestsRemove Triggered when an item is removed from the pendingTiles array
add_onRequestsCompleted Triggered when the number of item in the pendingTiles array changes to 0

Use these handlers on the Map object’s pendingTiles property to register a JavaScript function with the event. For example:


where map is the Map object and showBusyIndicator is a JavaScript function to call when the number of items in the pendingTiles array changes from 0 to a higher value.

The JavaScript function showBusyIndicator may appear as follows.

function showBusyIndicator(sender) {


            if (sender!=null) {

                window.status = “Pending Tiles: “ + sender.pendingTiles.length;


The argument to the function is a reference to the JavaScript Map object. This argument can be used to gain access to map properties, such as the number of map image tiles left in the pendingTiles array. In this example, the number of pending tiles is output to the browser window’s status bar. If the argument is null, the pendingTiles array contains 0 items. The Web ADF includes two convenient JavaScript functions to show or hide a layer (div) based on its id, named showLayer and hideLayer, respectively. The functions are contained in the display_common.js file which is by default embedded with the Web ADF controls. In this example, the showLayer function is used to make the contents in the div tag with an id of “BusyIndicator” visible.

You can show the number of pending tiles and a "busy indicator" in a Web mapping application

Included below is a simple Web page with a MapResourceManager, Map, and a div tag containing an image. The JavaScript Map object events are handled after the form to let the content of the form load before interacting with it.

<%@ Page Language=”C#” AutoEventWireup=”true”  CodeFile=”Default.aspx.cs” Inherits=”_Default” %>

<%@ Register Assembly=”ESRI.ArcGIS.ADF.Web.UI.WebControls, Version=, Culture=neutral, PublicKeyToken=8fc3cc631e44ad86″

    Namespace=”ESRI.ArcGIS.ADF.Web.UI.WebControls” TagPrefix=”esri” %>

<!DOCTYPE html PUBLIC “-//W3C//DTD XHTML 1.0 Transitional//EN” “”>

<html xmlns=”” >

<head runat=”server”>

    <title>Untitled Page</title>



    <form id=”form1″ runat=”server”>


        <esri:MapResourceManager ID=”MapResourceManager1″ runat=”server”>




        <esri:Map ID=”Map1″ runat=”server” Height=”453px” Width=”556px” MapResourceManager=”MapResourceManager1″>




     <div id=”BusyIndicator” style=”z-index: 1000; left: 25px; width: 100px; position: absolute; top: 422px;height: 100px”>

        <img src=”images/CircleThickbox.gif” />



   <script language=”javascript” type=”text/javascript”>


        function showBusyIndicator(sender) {


            if (sender!=null) {

                window.status = “Pending Tiles: “ + sender.pendingTiles.length;




        function showPendingTiles(sender) {

            if (sender!=null) {

                window.status = “Pending Tiles: “ + sender.pendingTiles.length;




        function hideBusyIndicator(sender) {


            if (sender!=null) {

                window.status = “”;




        // add busy indicator functions to the map

        map = Maps["Map1"];








Posted in Services | Tagged , , | 6 Comments

How to produce tint bands for boundaries

By Charlie Frye, Esri Chief Cartographer

Tint Bands - Thumbnail

Not long ago we were asked how to create tint bands for boundaries like the ones on National Geographic’s and other national atlas political maps. We didn’t have an immediate answer, but promised to look into it. Continue reading

Posted in Mapping | Tagged | 3 Comments

Welcome to the Mapping Center blog

By Aileen Buckley, Mapping Center Lead

MC Banner - thumbnail

Welcome to the Mapping Center Blog, created to bring map makers and ArcGIS software experts together for the purpose of sharing techniques for making maps. The Mapping Center Blog is part of a larger endeavor called Mapping Center which is a repository of mapping and cartographic know-how. Continue reading

Posted in Mapping | Tagged , | 2 Comments

Configuring your server to display a "Data not available" tile for empty map cache areas

Sterling Quinn of the Server team contributed this post on configuring your web server to display custom tiles in areas where you have not yet completed your map cache.


At the ESRI Developer Summit, several of you asked how we displayed a “Data not available” tile in empty areas of the ArcGIS Online services. This kind of tile can be useful if someone pans to the edge of the map or navigates to an area that you have not completed caching. Configuring your server to return a “Data not available” tile can in some cases yield a better user experience than returning nothing.



To display the tile, you need to create a custom error response on your virtual cache directory for HTTP Error 404: “Not Found”. Instead of an error message, the Web server returns the tile.

Following are steps for this process in IIS. Before you perform these steps, you should put the blank or “Data not available” tile in your cache directory. The tile you use must have the same dimensions and image format as the other tiles in the cache.


1. In IIS Manager, right-click the Virtual Directory for the specific cache and select Properties.


2. Click the Custom Errors tab, scroll down, and select the 404 error code.


3. Click the Edit button. In the URL box, specify the tile that IIS should return whenever a tile is missing. It is important to use a URL and not just a path to a file.


4. Click OK. Your dialog should look similar to the one below. Click OK again to return to IIS Manager.



You can download a sample, "Map data not yet available" tile (512 X 512 JPG) that you can use here


Posted in Services | Tagged | 4 Comments

Using ASP.NET AJAX with the Web ADF

Bryan Baker of the .NET SDK team wrote this great article on integration ASP.NET AJAX with the Web ADF.   


Microsoft’s new ASP.NET AJAX enables developers to refresh portions of a Web page in a relatively simple way. To enable a control to use AJAX, the developer can put the control inside an UpdatePanel. Then any events on the control are automatically handled via AJAX, and updates are passed back without refreshing the entire page. For example, a page might have two drop-down lists. The first lists types of features, such as Cities and Countries. When the user picks a value in the first drop-down, the second list is populated with new values corresponding to the first list’s value, such as a list of cities.

Web ADF developers might want to use this approach to update the Map or other ADF controls. For example, when the user selects a city in the second drop-down list, the Map would zoom to the selected city. The problem is that ASP.NET AJAX was released after the 9.2 version of the Web ADF, and these new features were not available to be incorporated. Putting Web ADF controls inside an UpdatePanel will not work correctly, and is not supported. At 9.3 ESRI does plan to support Web ADF controls with ASP.NET AJAX.

You can have both Web ADF and ASP.NET AJAX controls on the page. The challenge is doing something like the example above, where controls in an UpdatePanel communicate with Web ADF controls. This post shows one approach for this. It turns out that it takes less than a dozen lines of code beyond what would have been needed had the Map been inside the UpdatePanel! You can follow along if you have installed Visual Studio 2005, the Web ADF, and the ASP.NET AJAX 1.0 extensions.

Note that the use of ASP.NET AJAX is not supported by ESRI for version 9.2 of the Web ADF. If you use ASP.NET AJAX, you must be prepared to resolve any issues you encounter with using the Web ADF. You may find help at the ESRI Forums, but ESRI Support will not be able to assist with development issues.

Add the controls

I first created a new website using the “ASP.NET AJAX-Enabled Web Site” template in Visual Studio. This creates a website and adds some extra items to support AJAX into both the Default.aspx page and the web.config file. You’ll notice the ScriptManager control already added to the Default.aspx page, which is a non-visual control that handles the AJAX functionality. If you wanted to add ASP.NET AJAX into an existing page or website, you’d need to add these same items that the template adds in. See the ASP.NET AJAX documentation for details.

ASP.NET AJAX and Web ADF controls on the page On the Default.aspx page, I added an UpdatePanel from the AJAX Extensions toolbox tab. I dragged two standard DropDownList controls into the UpdatePanel. The first DropDownList will display a list of layers. We could obtain these from the Web ADF controls on page startup, but I just added them manually for now to the Items property. I added two items: Cities and Countries. I also set the AutoPostBack property for both DropDownList controls to True. We need to do this to trigger a postback (actually a “partial postback” as ASP.NET AJAX calls it) when the user changes the selection.

Next I added a Map control and a MapResourceManager control from the ArcGIS Web Controls. Tip: having a drop-down list just above the map inteferes with the drop-down’s functioning, so I put the map above the drop-down lists for this demo. I set the MapResourceManager property of the Map, and I added a resource item (map service) to the MapResourceManager as required to enable the map to display the service. In my case I used a world map with cities and countries. My simple page looks like the example here.

Update DropDownList items with ASP.NET AJAX

When the user changes the first drop-down list, the second list should display a list of cities or countries. Using ASP.NET AJAX allows us to treat this like a standard postback event. Behind the scenes, ASP.NET AJAX handles the request using AJAX methods rather than a full postback. Fortunately we don’t have to deal with those details here.

I double-clicked on the DropDownList1 on the design page, which creates the method to handle the user selection in the code-behind page. We need to also fill the second DropDownList at startup, so we’ll create a separate method and call it from both the drop-down’s change method and the Page load method (which we can create by double-clicking on the design page). The UpdatePlaceList method fills the second DropDownList with locations and coordinates that we’ll use later for zooming the map. The code below is in VB; both C# and VB versions are available in the download link at the end of this post.

Protected Sub Page_Load(ByVal sender As Object, ByVal e As System.EventArgs) Handles Me.Load

If Not Page.IsPostBack Then


End If

End Sub

Protected Sub DropDownList1_SelectedIndexChanged(ByVal sender As Object, ByVal e As System.EventArgs)


End Sub

Sub UpdatePlaceList()


If DropDownList1.SelectedValue = "Cities" Then

DropDownList2.Items.Add(New ListItem("Cape Town", "17|-36|21|-32"))

DropDownList2.Items.Add(New ListItem("Oslo", "9|58|13|62"))

DropDownList2.Items.Add(New ListItem("Washington", "-79|37|-75|41"))

ElseIf DropDownList1.SelectedValue = "Countries" Then

DropDownList2.Items.Add(New ListItem("Australia", "112|-43|157|-10"))

DropDownList2.Items.Add(New ListItem("Brazil", "-73|-33|-31|6"))

DropDownList2.Items.Add(New ListItem("China", "75|16|136|55"))

End If

End Sub

Zoom the Map when user selects a place

When the user selects a place in the second DropDownList, we need to zoom the map to the coordinates of that place. The challenge is to have the map’s new extent be communicated to the client. Since the Map control cannot be in the UpdatePanel, we have to use another approach to communicate the results of the extent change to the Map.

The user selection in DropDownList2 triggers another partial postback. To set up the code for this, I double-clicked on DropDownList2 in design mode to create the method to handle the selected-index change. Inside this new method, I obtained the coordinates to zoom to from the DropDownList2.SelectedValue property. I parsed this value into an array of coordinate values, created a new envelope, and set the envelope’s extent to these coordinates. I then set the Map’s extent to this new envelope. This changes the map extent on the server. However, the client won’t be aware of the change and won’t obtain a new map unless we tell it to.

To get the client to update the map, we use a feature in ASP.NET AJAX to pass information to the client. The ScriptManager.RegisterDataItem method adds information that will be passed back and evaluated on the client.

The Map’s CallbackResults will have the information required to update the map on the client. We pass this information to the client using the RegisterDataItem method. We’ll see shortly how the client uses this information to update the map.

The code below obtains the callback results and registers them with the ASP.NET AJAX RegisterDataItem method. The ScriptManager1.IsInAsyncPostBack property ensures that we’re doing an AJAX partial postback rather than a full page postback.

Protected Sub DropDownList2_SelectedIndexChanged(ByVal sender As Object, ByVal e As System.EventArgs)

Dim zoomString As String = DropDownList2.SelectedValue

Dim coordArr() As String = zoomString.Split("|")

Dim new_extent As New ESRI.ArcGIS.ADF.Web.Geometry.Envelope

new_extent.XMin = Double.Parse(coordArr(0))

new_extent.YMin = Double.Parse(coordArr(1))

new_extent.XMax = Double.Parse(coordArr(2))

new_extent.YMax = Double.Parse(coordArr(3))

Map1.Extent = new_extent

If ScriptManager1.IsInAsyncPostBack Then

Dim callbackString As String = Map1.CallbackResults.ToString()

ScriptManager1.RegisterDataItem(Map1, callbackString)

End If

End Sub

Handle the results on the client

The results that we just registered with ScriptManager need to be handled on the client. ASP.NET AJAX has a handler approach to process items passed back to the client. I inserted the code below into the .aspx page, at a point below the asp:ScriptManager tag. This code registers a client-side handler, the PageLoadingHandler function, that will run when the page loads. This handler obtains the items registered on the server with RegisterDataItem, and checks whether any items for the map control are present.  Since our server-side code added an item for the map, the code runs the next line.

This next line calls processCallbackResult and passes it the callback results we registered on the server. The processCallbackResult function is part of the Web ADF JavaScript library, which is automatically downloaded when the page uses Web ADF controls. The processCallbackResult function applies the callback results on the client by retrieving a new map. One caveat: the objects and methods in the client-side library can change, so this method could change with future versions of the Web ADF.

<asp:ScriptManager ID="ScriptManager1" runat="server" />

<script type="text/javascript" language="javascript">


function PageLoadingHandler(sender, args) {

var dataItems = args.get_dataItems();

if (dataItems['Map1'] != null)

processCallbackResult(dataItems['Map1'], 'Map1');



With this code in place, when the user selects an item in the places drop-down list, the server sets the map extent to that place, and the map on the client gets notified to retrieve a new map.

Tip: be careful about dragging controls around the page in Design mode when script blocks are within the page body. Doing so caused Visual Studio to remove my script block, which disabled the map update since no handler was available on the client!


We’ve seen how you can use controls in an ASP.NET AJAX UpdatePanel to control a Map in the Web ADF. Although it takes more work than if the Web ADF controls were embedded in the UpdatePanel, it is possible to do these tasks now. In our case it required less than a dozen lines of additional code compared to when the controls are all inside the UpdatePanel.

Download the code for this demo (includes C# and VB)

Try it out here

Posted in Services | Tagged , | 3 Comments

Using custom logos in ArcGIS Explorer

Mark Bockenhauer of the ArcGIS Explorer team contributed the following post on how to display a custom logo in ArcGIS Explorer.



It is common with paper maps to display a logo perhaps a company logo or government seal.  The same can be done on digital ArcGIS Explorer maps by leveraging the Home Server.

The Home Server is a user defined server that ArcGIS Explorer connects to when it starts up.  Typically ArcGIS Explorer checks the home server to see if its version is up to date and to get the default map to display.  It does this by referencing information in the ‘explorer’ folder on the server. This folder is part of the ArcGIS Server install.

Download the folder before continuing with this example.  This folder contains example resources for all of the ArcGIS Server Home Server settings that can be applied to ArcGIS Explorer.

Once you have downloaded the file unzip it and place it in the appropriate location on your server.

Default Locations:
.NET  C:InetpubwwwrootArcGISExplorer
JAVA   C:Program FilesArcGISjavaweb_outputExplorer

To display your own custom logo on ArcGIS Explorer applications that specify your server as their home server you will need to modify the following files:


In the e2config.xml file the <skins> section is commented out (highlighted in bold below).  Un-comment this section by removing “<!- -“ prior to the section and “- ->” following the section.  You will also want type in the appropriate URL for the <skinurl> tag. 

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>

<E2Config xmlns:xsd="" xmlns:xsi="">













  <!– To use a server skin file, uncomment the 'skins' tags below by removing the exclamation point and hyphens.

     Type your server name into the skin url. The server install contains an example skin file and resources –>







Next we will modify the skin file referenced in the <skins> tags to display the desired logo. In the explorerskinsexample_skin_file you will see that all of the lines are commented out.  You will also notice that there are two sections to the file.  The tags in the upper section all control the ArcGIS Explorer color scheme.  The tags in the lower section pertain to ArcGIS Explorer graphic elements.  Un-comment this section.

    <HUDFont type="Font">Verdana</HUDFont>

    <ApplicationTitle type="Application">ExampleApplicationTitle</ApplicationTitle>

    <DocumentPrefix type="Application">ExampleDocumentPrefix</DocumentPrefix>

    <SplashScreenBitmap type="Application">example_splash_screen.png</SplashScreenBitmap>

    <SplashMessageCoords type="Application">15,200</SplashMessageCoords>

    <DisplayCustomLogo type="Application">example_custom_logo.png</DisplayCustomLogo>

    <DisplayCustomLogoMaskColor type="Application">RGB(255, 0, 255)</DisplayCustomLogoMaskColor>

    <DisplayCustomLogoTransparency type="Application">75</DisplayCustomLogoTransparency>

    <NorthArrowBitmap type="Application">example_north_arrow.png</NorthArrowBitmap>


At this point if you start ArcGIS Explorer and specify your server as the Home Server, you will see the Example Custom Logo.

Setting the Home Server to your server.
1.    In Explorer, click File > Set Home Server
2.    Click “Connect to the Home Server located at”, then enter the URL to your server in the format http://<server name>/<instance name> (Example: http://myServer/ArcGIS)
3.    Click the Test button to test the connection, or click OK to return to Explorer.
ArcGIS Explorer will restart and apply the Home Server Settings.

To use your own logo you will want to change the .PNG referenced in the <DisplayCustomLogo> tag.  Your .PNG file should be located in the skins folder with the skin file.

Notice that the example logo makes use of the <DisplayCustomLogoMaskColor> to display a rectangular logo.

Actual .PNG on the left, what is displayed on the Right.


For more information on making your server the ArcGIS Explorer Home Server see "Making your system an ArcGIS Explorer Home Server" under "Administering the Server" at:  

Posted in Services | Tagged | Leave a comment

Deconstructing the ArcGIS Server tiling schema

ArcGIS Server 9.2 introduced the concept of map caches.  A map service that fullfills requests with pre-created tiles from a cache instead of dynamically rendering part of the map is called a cached map service.  The map cache is generated according to its tiling scheme.  A tiling scheme consists of a coordinate system, a tiling grid (which is made up of a tile origin, image size in pixels, and the DPI), and the levels of detail (cached scales).  Lets look at each of those properties in a little more detail.



Posted in Services | Leave a comment

Working with the Web ADF resources in a custom task at run-time

Rex Hansen of the .NET SDK team wrote this very useful post on accessing the Web ADF components at run-time.


When creating a custom task, you may often find it necessary to access Web ADF components (controls, resources, etc.) that share the same Web page with the task.  How the task will be used will dictate the technique for accessing Web ADF components.  There are three situations to consider:

1) Task run-time

2) Visual Studio design-time

3) Manager run-time

“Task run-time” defines the technique for working with Web ADF components in a Web application at run-time.  “Visual Studio design-time” uses custom verbs on the custom task control to access to Web ADF components during task configuration in Visual Studio.  “Manager run-time” defines a situation where a custom task must utilize Web ADF components when configuring the task in the Manager Web application.  

This discussion and walkthrough will focus on working with Web ADF resources in a custom task at run-time (situation 1).  The custom task will do the following:

  • Extend FloatingPanelTask
  • Expose a public property to buddy the task with a MapResourceManager
  • Use a DropDownList control to list the map resources in the MapResourceManager control
  • Populate the DropDownList upon initial load of the task

1) Create a custom task control that extends FloatingPanelTask.  As a composite control, the custom task will contain multiple controls.  Define two private member variables to store references to controls that will be used in the custom task.  In this case, a Label will be used to display the MapResourceManager id and a DropDownList will display a list of map resources.

namespace TestTask_CSharp


    public class TestTask_Demo : FloatingPanelTask


        private Label label = null;

        private DropDownList dropDownListResources = null;

2) Create a property on the custom task control to store a reference to the unique id of a MapResourceManager control.  The property will be maintained in state for the duration of a user session.  The property can be set declaratively in the aspx page using the MapResourceManagerID attribute (source view). 




        public string MapResourceManagerID




                object obj = StateManager.GetProperty("mapResourceManagerID");

                if (obj == null) return "";

                return obj as string;




                StateManager.SetProperty("mapResourceManagerID", value);



3) As with other composite Web controls, override the CreateChildControls method to construct the custom task control at runt
ime.  The Label and DropDownList are created and added to the custom task.  Note the DropDownList is empty.  The Controls property is inherited from the System.Web.UI.WebControls.CompositeControl class and maintains a collection of controls to be rendered at runtime.   

        protected override void CreateChildControls()




            label = new Label();

            label.Text = MapResourceManagerID;

            label.ID = "label_mrm";

            dropDownListResources = new DropDownList();

            dropDownListResources.ID = "dropdownlist_resources";




4) At this time we only need to populate the DropDownList (dropDownListResources) with a list of map resources upon initial load of the custom task.  Since interrogating the MapResourceManager to get a list of resources during every callback is not necessary and can be expensive, we only want it to occur during a full page postback.  The PreRender step in the page lifecycle offers a good location to modify page or control content before the final rendering step, and it's only called during a full page postback.  Override the OnPreRender event of the custom task control and call the UpdateResourceDropDownList() method.  The method will contain the logic to update the dropDownListResources control.  This logic was placed in a separate method so it can be called from other locations in the custom task code, if necessary in the future.        

        protected override void OnPreRender(EventArgs e)






5) The UpdateResourceDropDownList() method uses the MapResourceManagerID property to find a MapResourceManager control on the page and get a list of map resources.  The GetMapResourcesFromMRM() method returns a collection of resources from the MapResourceManager.  If the MapResourceManager has not been initialized, it will be explicitly initialized. Including code to return a collection of resources from a MapResourceManager was placed in a separate method so it can be called from other locations in the custom task that do not involve the dropDownListResources control.  Once a list of resources is returned to the UpdateResourceDropDownList method, only those that are an ArcGIS Server map resource are added to the dropDownListResources control.           

        protected void UpdateResourceDropDownList()


            if (!string.IsNullOrEmpty(MapResourceManagerID))


                GISResourceCollection grcoll = GetMapResourcesFromMRM(MapResourceManagerID);

                foreach (IGISResource gr in grcoll)


nbsp;                  if (gr is ESRI.ArcGIS.ADF.Web.DataSources.ArcGISServer.MapResourceBase)







        protected GISResourceCollection GetMapResourcesFromMRM(string mrmID)


            GISResourceCollection gisRC = new GISResourceCollection();

            MapResourceManager rm = (MapResourceManager)Page.FindControl(mrmID);

            if (rm == null)

                return null;

            if (!rm.Initialized)


            foreach (MapResourceItem mri in rm.ResourceItems)




            return gisRC;


6) Override a set of methods associated with task implementation to complete the custom task control.  Further details on these methods will be provided in future blog posts.

        public override string GetCallbackResult()


            return base.GetCallbackResult();


        public override void ExecuteTask()


        public override List<GISResourceItemDependency> GetGISResourceItemDependencies()


            List<GISResourceItemDependency> list = new List<GISResourceItemDependency>();

            return list;




7) Build the custom task in Visual Studio and add a reference to it in a Web page.  Add the custom task and a MapResourceManager control to the page.  Add one or more ArcGIS Server resources to the MapResourceManager.  Run the Web app and you should see a drop down list containing the name of each ArcGIS Server map resource item in the MapResourceManager.    It should look similar to the following screenshot:




You can download the sample here.


Posted in Services | Tagged , , | 1 Comment

Extending the QueryAttributesTask to highlight selected features

Bryan Baker of the .NET SDK team wrote the following great post on extending a task to modify its behavior and then adding that task to the .NET Global Assembly Cache so it can be reused across applications. 

Note: Also see the follow-up to this post from May 3, 2007.

Highlighting all task results


With the Web ADF at 9.2, a website can have one or more tasks to allow users to find features or locations on the map. For example, you might add a QueryAttributesTask to the website to allow users to find cities by name, by typing the first few characters in the name.


The out-of-the-box tasks at 9.2 do not automatically highlight found features on the map. Instead, the user can highlight features by clicking individual check-boxes of features in the task results (see graphic, with two cities selected).



What if you want to have all features highlighted immediately when the task runs? There is currently no setting to enable this. But such immediate highlighting is possible through customization. Let’s look at one relatively easy approach for someone with modest programming skills. We’re looking specifically at the Web ADF for the Microsoft .NET Framework here, by the way.


We will extend a task control to modify its behavior. The object-oriented nature of .NET allows us to create a new class (a task, in this case) that inherits all of the behavior and properties of the original class. We only need to add or modify the original class where we need it to act differently from the original class (task). Some aspects of tasks may be difficult or impossible to change, but modifying the task results output is not difficult.


The approach we’ll look at can apply to any task that produces a task result in the form of a graphics layer, where users can click on feature check-boxes to highlight them on the map. This includes the SearchAttributesTask, FindAddressTask, FindPlaceTask and QueryAttributesTask. We’ll look specifically at the QueryAttributesTask in this example.

Extending an out-of-the-box task


First, I open Visual Studio 2005 and create a new project (not a new website), a Class Library project to be specific. One great thing about .NET is that even though the original class was written in C#, we can extend it in any language—we’ll use VB to show this here. We can put our new class library anywhere; it doesn’t have to go into a web folder. I’ll call my project QuerySelectTaskVB. By the way, we could also use either Visual Basic Express or Visual C# Express, but it’s more difficult to debug with a linked web application.


Visual Studio creates the project and adds a new class called Class1.vb. I right-clicked on it in the Solution Explorer and renamed it QuerySelectTaskVB. This also renames the class in the code view—nice.


We need to add some references to the project to the libraries we’ll be using. I right-click on the project in Solution Explorer, and chose Add Reference, and in the pop up dialog I select these libraries and then click OK:

  • ESRI.ArcGIS.ADF.Tasks
  • ESRI.ArcGIS.ADF.Web.DataSources
  • ESRI.ArcGIS.ADF.Web.UI.WebControls
  • System.Web


To start creating our class, we add some Imports statements at the top of the class file so we can use classes without having to type the full path. We’ll also add a Namespace around our class to better identify it.


Imports System
Imports System.ComponentModel
Imports System.ComponentModel.Design
Imports System.Data
Imports System.Web
imports System.Web.UI
Imports ESRI.ArcGIS.ADF.Tasks
Imports ESRI.ArcGIS.ADF.Web.Display.Graphics
Imports ESRI.ArcGIS.ADF.Web.UI.WebControls


Namespace QuerySelect
  Public Class QuerySelectTaskVB
  End Class
End Namespace


We tell the class to extend (inherit) the existing QueryAttributesTask by adding to the class declaration we’ve already created:


Public Class QuerySelectTaskVB
    Inherits QueryAttributesTask

End Class


The only thing we want to change in the task is how it outputs the results to the TaskResults control. The task creates this output in the standard task method called ExecuteTask(). We want the QueryAttributesTask to create its output as usual, but we will modify the results once they’re created. So we create our own version of ExecuteTask() that overrides the original. Our version will call the base (parent) class version of the method, then modify the results that have been created. The following code goes inside the Class definition we saw above:


    Public Overrides Sub ExecuteTask()
        ' QueryAttributesTask creates its results
        ' We'll modify the Results next
    End Sub


Once the parent class has run its ExecuteTask, we can modify the results. The results are stored in a task property object called Results (makes sense, eh?). If the query finds features, those get stored in Results as a standard .NET DataSet, which contains one or more DataTable objects. But if nothing was found, the Results contains a different type, a SimpleTaskResult. Also, if the map service isn’t working right, the results might have a DataSet, but the name (caption) just has an error message. Let’s make sure we have valid results:


        ' We'll modify the Results next
        ' Make sure features were found

        If TypeOf Results Is DataSet Then

            Dim resultsDS As DataSet = CType(Results, DataSet)
            ' Check for errors during query

            If resultsDS.DataSetName.IndexOf("Error") > 0 Then
End If

             ' Next we can get the table of results
End If


Now we can get the table from the DataSet. With the QueryAttributesTask, only one DataTable will be in the DataSet (other tasks may have multiple tables). One more error check: after the task performs its query using the Web ADF common API, it converts the data table to a GraphicsLayer object, so that it contains rendering (symbology) information. If it had any problems making that conversion, then features can’t be selected on the map.


            ' Next we can get the table of results
' Get the one table in the QueryAttributesTask result
Dim resultsTable As DataTable = resultsDS.Tables(0)

            ' Make sure no problems creating a GraphicsLayer from the results
If resultsTable Is Nothing OrElse Not TypeOf resultsTable Is GraphicsLayer Then


            End If

            ' Now we can modify the table results


Now we can actually set the selection for the features found! The GraphicsLayer, which extends the DataTable type, will have a Boolean-type column indicating whether the row (map feature) is selected. We can get this column, and use it when looping through the features to set all features as selected:


            ' Now we can modify the table results
Dim graphicsLayer As GraphicsLayer = _
(resultsTable, GraphicsLayer)

            ' Get the column that holds the selection attribute
Dim selectedCol As DataColumn = graphicsLayer.IsSelectedColumn

            ' Set each feature to selected
For Each row As DataRow In graphicsLayer.Rows
                row(selectedCol) = True



That’s it! Now each feature will be selected in the TaskResults tree and also on the map….that is, once our task is inside a website, which we’ll show shortly.


We compile the code by choosing Build—Build Solution f
rom the menu (no errors, of course!). This puts a compiled .dll file into the Bin directory of the project (you can see that by looking at the project folder with Windows Explorer).


Adding the task to Visual Studio

The easiest way to add the task to a website is when it’s in the Visual Studio toolbox. It can be added manually to an individual website, but requires more editing of the source of the page. Let’s get the task into Visual Studio. There’s more on this in the Developer Help for ArcGIS Server and ArcIMS, so we’ll go over this quickly.


First, in the code for the task itself, we add some information so Visual Studio sets properties for task when we drag it onto the page from the toolbox. This goes just above the class declaration:


<ToolboxData("<{0}:QuerySelectTaskVB runat=""server"" Width=""200px"" Transparency=""35""" _

        + "BackColor=""White"" TitleBarColor=""WhiteSmoke"" TitleBarSeparatorLine=""False""" _

        + "TitleBarHeight=""20px"" BorderColor=""LightSteelBlue"" BorderStyle=”Outset”’ _

        + "BorderWidth=""1px"" Font-Names=""Verdana"" Font-Size=""8pt"" ForeColor=""Black"">" _

        + "</{0}: QuerySelectTaskVB >")> _

Public Class QuerySelectTaskVB

    Inherits QueryAttributesTask


Next we will add the task to .NET’s Global Assembly Cache. This makes it easier to add the task to various websites. Before we can do this, we must create a strong name to sign the task. We open the .NET command prompt and type:


sn –k QuerySelectVB.snk


This creates a key file (QuerySelectVB.snk) in the folder indicated by the command prompt (probably C:Program FilesMicrosoft Visual Studio 8SDKv2.0). I moved that file using Windows Explorer to the project directory where my task .vb file is located. I then added that to my project by right-clicking on the project in Visual Studio’s Solution Explorer, choosing Add-Existing Item, changing the type to All Files, and clicking on the .snk file. I used the keypair to sign the assembly by right-clicking on the project and choosing Properties. In the Properties page, I clicked the Signing tab. I checked the box Sign the assembly and clicked my .snk file in the drop-down list. Finally, I recompiled the code (Build-Build Solution).


Now we can add the task to the Global Assembly Cache. Here’s the easy way: I opened two Windows Explorer windows, one to my project folder’s bin directory where the .dll is located, the other to C:WindowsAssembly. I dragged the QuerySelectTaskVB.dll into the C:WindowsAssembly window. This doesn’t actually move the file. It just registers the assembly so it’s available to all applications on the system.


Finally let’s add the task to Visual Studio. Open Visual Studio and then any .aspx page (you can create a website and page if necessary). Open the toolbox if necessary. Right-click where you want to add the task and click Choose Items (you can also create a new tab with New Tab).  In the Choose Toolbox Items dialog, click Browse, and navigate to your DLL’s location. Highlight it and click Open. This adds the task to the list of available assemblies. Make sure its check box is checked, and click OK. This adds the task to the Visual Studio toolbox.


Now the task is available whenever you design an ASP.NET web page. Notice it has an icon—the same one as the QueryAttributesTask. It inherits this along with other properties of the parent task.

Using the custom task in a website


We can use the task in any website. I created a new website using the Web Mapping Application template. I could have added this website to the same solution as my task code if I’m using Visual Studio (useful for debugging the task), or in a new instance of Visual Studio. I could also open a website created with ArcGIS Server Manager or ArcIMS Web Manager, and add the task to it.


Once we have our web page open in Visual Studio, we can drag our new task from the toolbox into the Task Manager. Once it’s there, we can click its “smart tag” in its upper right to set the task’s properties.




Yet another great thing about extending an existing task: we get its designer tools with no coding required. We can set the Task Results container and use Edit the Query to create the query, exactly the same as with the standard QueryAttributesTask (see the Developer Help for tips on that).


The demo site for this page uses this custom task to query a water-well point layer for estimated yield. Choose a value from the drop-down list, and up to 50 features will be found. Notice that all features are immediately highlighted. Our custom task acts like the standard QueryAttributesTask, except that it highlights all features returned.


Try it out:



You can do other interesting things with the Results of the task. For instance, you could add a hyperlink to the task results output by adding the <a href…> tag to the DataSet’s DataSetName property, or to the DataTable.TableName. The hyperlink is displayed in the tree of the task results. You can also modify the DataTable itself, such as by hiding columns you don’t want displayed. Another example would be to display the results as a table, by creating a GridView from the DataTable and getting the HTML output from the GridView. You’d then create a new TaskResultsNode, add the output to that, and set it as the Results instead of the DataSet.


You might be tempted to customize other aspects of tasks, such as modifying
the task’s user interface. Although some customizations are possible, you’ll probably find that in many cases it would be easier to author your own task from scratch. The Developer Help has a good discussion of writing tasks, and the Web ADF has a couple of samples that can help get you started.


You might have noticed that the task user interface does not have the look-and-feel of other tasks—the colors, fonts, etc. The standard tasks have properties defined in the Theme of the website, specifically in the file. If you want the same look-and-feel of the standard tasks, you can create your own control skin by copying properties from one of the standard tasks in


We haven’t discussed how to make our custom task available within Manager, so that non-programmers could use your task. This is a much more complex chore than adding the task to Visual Studio. Tasks are configured in Manager using a separate class called a web configurator. The standard task’s web configurators are hard-coded to output information specifically for that task. It is not possible to easily modify them to instead output tags and properties for tasks that extend them. It would be necessary to rewrite most or all of the web configurator in order for it to properly output your task from Manager. The Developer Help has information on writing web configurators, if you are ambitious.


Download the code for this sample:




Posted in Services | Tagged , , | 7 Comments