The SharePoint 2010 – Client Object Model is one of the most interesting new enhancements in SharePoint Foundation 2010 and SharePoint Server 2010. It includes 3 APIs (managed .NET, Silverlight and ECMAScript/JavaScript) allowing you to interact with SharePoint 2010 from a client which hasn’t to be run on a SharePoint server. I had some time to play around with it using resources which were available until the end of 2009.

I’m trying to become a SharePoint 2010 developer using the resources which are available. I will write down my impressions in different articles…

Resources

First of all here are the resources I used:

Videos

The first link includes 6 videos and the second link includes 3 videos with Andrew Connell.

• Get Started Developing on SharePoint 2010
• SharePoint 2010 developer – Client Object Model

Note: I think the first link provides videos based on SharePoint 2010 alpha or beta 1 and not the public beta… so there might be differences.

MSDN

• Using the SharePoint Foundation 2010 Managed Client Object Model
• SharePoint 2010: Getting Started with Development on SharePoint 2010 Hands-on Labs in C# and Visual Basic

Blog posts

Steve Peschka has written a really good series:

• Using the SharePoint 2010 Client Object Model - Part 1
• Using the SharePoint 2010 Client Object Model - Part 2
• Using the SharePoint 2010 Client Object Model - Part 3
• Using the SharePoint 2010 Client Object Model - Part 4
• Using the SharePoint 2010 Client Object Model - Part 5
• Using the SharePoint 2010 Client Object Model - Part 6

Using the SharePoint 2010 - Client Object Model

It’s mentioned in Steve Peschka’s series that you need to take care about the amount of data which is sent over the wire. To clarify things I tried it by myself and here are the results:

1. The first method I have used receives 22k and includes some bad coding.
2. The second method is optimized using a lambda expression and receives about 2k.
3. The last method is more optimized and receives only 710 bytes.

Method 1

This method includes some unnecessary code since you don’t need to load the ‘web’, the ‘lists’ and the ‘tasklist’. You only need to use the last ctx.load() statement.

ClientContext ctx = new ClientContext(siteUrl);
Site site = ctx.Site;
Web web = site.RootWeb;
ctx.Load(web);

ListCollection lists = web.Lists;
ctx.Load(lists);

List taskList = lists.GetByTitle("Tasks");
ctx.Load(taskList);

CamlQuery query = new CamlQuery();
query.ViewXml = "<View><Query><Where><Eq><FieldRef Name='Title'/><Value Type='Text'>
                 Task1</Value></Eq></Where></Query><RowLimit>3</RowLimit></View>";
ListItemCollection listItems = taskList.GetItems(query);
ctx.Load(listItems, items => items.IncludeWithDefaultProperties(item => 
                             item.DisplayName));

ctx.ExecuteQuery();

Because of unnecessary code the payload (22k) is high compared to the other methods.

Method 2

Removing unnecessary code results in a lower payload (about 2k)

ClientContext ctx = new ClientContext(siteUrl);

List taskList = ctx.Web.Lists.GetByTitle("Tasks");

CamlQuery query = new CamlQuery();
query.ViewXml = "<View><Query><Where><Eq><FieldRef Name='Title'/><Value Type='Text'>
                 Task1</Value></Eq></Where></Query><RowLimit>3</RowLimit></View>";
ListItemCollection listItems = taskList.GetItems(query);

ctx.Load(listItems, items => items.IncludeWithDefaultProperties(item => 
                             item.DisplayName));
ctx.ExecuteQuery();

The lambda expression used in the ctx.load() statement loads all fields of a list items. If you know which columns you need you can use the third method.

Method 3

The difference is the lambda expression where you can select the properties you want to receive before you execute the query.

ClientContext ctx = new ClientContext(siteUrl);

List taskList = ctx.Web.Lists.GetByTitle("Tasks");

CamlQuery query = new CamlQuery();
query.ViewXml = "<View><Query><Where><Eq><FieldRef Name='Title'/><Value Type='Text'>
                 Task1</Value></Eq></Where></Query><RowLimit>3</RowLimit></View>";
ListItemCollection listItems = taskList.GetItems(query);

ctx.Load(listItems, items => items.Include(item => item["StartDate"], item => 
                             item["DueDate"], item => item.DisplayName));
ctx.ExecuteQuery();

The payload is 710 bytes. 

Summary

The SharePoint 2010 – Client Object Model is definitely a really cool feature since it allows you to access data from SharePoint with an application that doesn’t need to run on the server. And compared to the SharePoint 2007 beta phase there is enough documentation available to get started.