Linq up that Webservice

Linq up that Webservice

As with many developers around the globe I downloaded the latest CTP of the Linq technology preview last week, but of course, work and other things got in the way.  As such it was this evening when I really had my first play (other than just opening the box and going wooow, which I dutifully did last week).  Of course, as Bill has already pointed out, one of the big improvements in the VB.NET space is the From, Where, Select syntax for Linq expressions.  This makes more sense if you look at how IDE support (eg intellisense) is vital to a feature being usable!

Despite giving a session at this year’s Code Camp I still felt that there was a lot more I need to play with to really test the boundaries of this new technology.  An interesting scenario presented itself over the weekend, and for those of you building distributed, ocassionally connected applications this scenario might sound familiar: 

Let say I have a database which contains a table called Customers.  My mobile application wants to be able to call a webservice (eg CurrentCustomers) which will return a list/array of Customer objects.  While offline, these Customers may be modified, then when a connection is available, updates are persisted back to the database via another webservice (eg UpdateCustomer, which takes a Customer object as the parameter).

Now, lets break this into Linq/Dlinq components

–   Create a new Linq ASP.NET application
–   Add a Dlinq object model and drag my Customers table across from Server Explorer
–   Create the Webservice methods CurrentCustomers and UpdateCustomer
–   Add code to CurrentCustomer to return the approprate Customer objects

Return (From c In db.Customers _
            Select c).ToArray

NB: Will leave code for UpdateCustomer for the moment as this is a bit tricky….

–   Add web reference to the new webservice
–   Call the webservices (ie retrieve customer list, make changes and call update webservice)

So, the only part that I’m left to implement is the UpdateCustomer method in the webservice.  And here lies a problem – the way that Dlinq works is that when you access objects using a linq expression it dynamically loads them into memory.  As you make changes to these objects it tracks what has been modified so that when you call SubmitChanges on the DataContext it knows what changes to persist to the database.  Since the Client side of your application needs to be able to operated when disconnected, there is no way for this persistance manager to keep track of the objects you are working with.  When the Client submits the changes via the UpdateCustomer method somehow the persistance manager needs to be notified so that it can handle the writing to the database.

Before we jump in and look at the solution lets just delve a little into how the Dlinq persistance model operates.  The persistance model starts with a class that inherits from DataContext.  Think of this correlating to a database. This class in turn contains a number of Table(of T) objects, where T is a mapping class for a table in the database (in our case the Customer table).  T also implements the INotifyPropertyChanging and INotifyPropertyChanged interfaces.  This point is significant as it is these events that are used by the persistance manager to track changed objects. When a changed event is raised on an object, a clone of that object is created and maintained by the manager.  At a later stage the user has the option of rejecting changes (in which case the object is reverted to the clone) or submitting changes (in which case the difference between the objects is used to update the database).

The solution to our scenario is that somehow we need to a) get an unchanged version of the Customer object we have modified from the persistance manager, b) update that object with the Customer object sent from the client and finally c) persist the changes to the database.  In doing this I actually worked in reverse since a) is actually more fiddly, but I will present them here in order so that they make more sense.

a)   Get an Unchanged Customer

To retrieve a Customer object from the persistance manager we need to access the CommonDataServices object that sits within the DataContext object.  As this is a protected property we need to us a bit of reflection magic to discover it.  Once we have a reference we can call the Requery method to retrieve the Customer object.  Note that this object may be retreived from the in memory cache, or will be pulled from the database if the object doesn’t already exist.

   Dim ser As Object = context.GetType.InvokeMember(“Services”, Reflection.BindingFlags.GetProperty Or Reflection.BindingFlags.Instance Or Reflection.BindingFlags.NonPublic, Nothing, context, Nothing)

   Dim currentCustomerState As Object = ser.GetType.InvokeMember(“Requery”, Reflection.BindingFlags.InvokeMethod Or Reflection.BindingFlags.Instance Or Reflection.BindingFlags.Public, Nothing, ser, New Object() {newCustomerState, False})

b)   Update Unchanged Customer

To update the managed Customer from the client’s Customer all we need to do is iterate through each of the writable Properties of the Customer type.  The appropriate properties will also be marked with the ColumnAttribute to indicate their association with a database column.

        Dim props As PropertyInfo() = gettype(Customer).GetProperties
        For Each prop As PropertyInfo In props
            If prop.CanWrite And prop.CanRead Then
                Dim attribs As ColumnAttribute() = TryCast(prop.GetCustomAttributes(GetType(ColumnAttribute), True), ColumnAttribute())
                If attribs IsNot Nothing AndAlso attribs.Length > 0 Then
                    prop.SetValue(currentCustomerState, prop.GetValue(newCustomerState, Nothing), Nothing)
                End If
            End If

c)   Persist Changes to Database

What remains is to use the DataContext object to persist the changes to the database:


And there you have it – how to persist a webservice object.  Now you can take this to the next level by extracting this out as an extension method for the DataContext class by removing all references to the Customer class, but I’ll leave this discussion to another time (or to Bill’s discussion on the topic).

Live Messenger gets another face lift and a KB article that laptop/tablet users should know about

Live Messenger gets another face lift and a KB article that laptop/tablet users should know about

As I didn’t sign into Messenger over the weekend I didn’t become aware of the refresh to the beta version of Live Messenger.  In my previous post on the leaked version of Live Messenger I talked about building Addins using the new managed API.  In the official update (version 8.0.0689, available via the beta program) this API is also now available for you to begin developing with.

The other piece of information I would like to share with all laptop and tablet PC users who have more than 1Gb of RAM on their machine relates to the occasional inability to hibernate.  If you have issues with hibernation then check out this knowledge base article as it might help.

Mobile Client Software Factory – Drop 7

Mobile Client Software Factory – Drop 7

So I’ve managed to allocate some time to take a quick look at the most recent community drop for the Mobile Client Software Factory.  From an initial inspection the part that I really like is that the Dynamic Resolution application block now works.  In my earlier post I discussed how this block can be used to provide different layouts for different resolutions/orientations.  At that stage there were some issues when a control was docked on a Form, as it didn’t seem to apply the new layout before docking. This resulted in the layout not being appropriately resized.  In the latest drop this issue seems to have been fixed. 

I also noticed that not only can you change the resolution/orientation of your control while you are building it, but when you add your control to a form it applies the appropriate layout.  For example if you rotate your form in the designer (using the design skin introduced into VS2005) your dynamic resolution control will apply the horizontal layout so you can see what it will look like. 

Well done team, keep up the good work!

WM5 and Embedded devices @ MEDC

WM5 and Embedded devices @ MEDC

If you have seen the blogs by Dr Neil, Mike Hall and others about the Australian MEDC you would of course know that I’ve been invited to help the crew out and present again at the event this year.  During one of the many discussions about the event I was privy to some news that I thought you would all be interested to hear:  The team have secured a number of SP5 Windows Mobile phones and Via Single Board computers (which run Windows Embedded CE/XP).  At this stage I’m not sure of how or to whom these devices will be given, but hey what’s one more reason to attend the premier Mobility event of the year (don’t wait for TechEd to get your annual injection of developer goodness) !!!

Places at MEDC are limited, and rapidly filling up, I would recommend that you take advantage of this early notice and register as soon as possible.

Look forward to seeing as many of you as possible at the event.

Mobile Client Software Factory – Drop 6

Mobile Client Software Factory – Drop 6

Good news fellow mobilites, community drop 6 of the Mobile Client Software Factory (previously the Mobile Baseline Architecture Toolkit) has hit the shelves.  Actually the news is a bit old as I was trying to allocate some time to review the changes between drops.  Unfortunately time (particularly when there is work to be done) does get away from you some times.

In summary some of the changes that I have had time to notice are:

  • New Configuration Manager application block (which provides basic configuration functionality, similar to what’s found in the full framework)

  • Connection Manager – now supports removing networks and connections after the manager has been created

  • Data Access – the concept of an abstract DatabaseService has been replaced with an abstract Database.  It also uses a DbProviderFactory, which should also be inherited for each database type to implement functions, CreateCommand, CreateConnection and CreateParameter.  This factory needs to be supplied in the constructor of the concrete Database implementation.

  • Data Access – the SQL Mobile implementation of the DatabaseService class also provides support for the SQLCeResultSet, which is unique to SQL Mobile.

More information will be available as I start to work more closely with this new framework.  Also, register for MEDC to see some of these application blocks in action when I take to the stage to assist fellow presenters Dr Neil Roodyn, Dr Peter Stanski and Dave Glover.

Vacant Exception Swallowers

Vacant Exception Swallowers

So, I guess the question raised in Mitch’s counter-post following my thoughts on checked exceptions in Java is whether we should have compiler errors for vacant exception handlers or whether we need to cull the developer population of vacant headed developers who think that empty exception handlers are acceptable. 

In the same way that enforcing coding standards (eg using code analysis/FxCop) doesn’t ensure developers write good code, neither does having checked exceptions.  I guess one reason I like checked exceptions is that I have an idea of under what conditions a method call will throw an exception (is is of course assuming that framework developers use the concept properly).

More MEDC Matters

More MEDC Matters

In case you haven’t already seen the blogs from others, such as Dr Neil Roodyn, Mike Hall and Frank Arrigo, the list of speakers is now available for the Australian MEDC:

This is set to be a fantastic event and I highly recommend attending.  The level of presentations will be more advanced, based on customer feedback, than in previous years so it will be worth attending.

Oh to be an Exception….

Oh to be an Exception….

After reading Mitch’s discussion on throwing Exceptions and quickly refreshing my brain as to what the official Microsoft position is, I went away and was rethinking how I determine if/when I’m going to throw an exception.

I started off with the following concepts:

–   An exception should be thrown when a method/operation cannot be completed successfully.

–   Exceptions should no be used for application flow control.

–   Return values should not be used to indicate if a method/operation has been completed successfully

Now I must say that I agree with all of these.  However there is one concept that I definitely think has merit, which is:

–   An additional set of functionality should be provided to allow a developer to query whether a method/operation will be successful if called (eg “CanConnect” to be called prior to an attempt to call “Connect”).

This last concept does cause me some concern though.  The main reason that you would want this functionality is to reduce the number of exceptions that are used for application flow control (see concept above) – for example in the case of the Connect method you would use a condition based on the CanConnect method to determine if Connect should be called.  So, what are the downsides:

–   Firstly, I see having these guards might lead to sloppy programming where exceptions don’t get managed.  For example, just because I have called CanConnect, doesn’t mean that Connect won’t fail.  If it does and an exception is thrown, I still need to handle it!

–   Secondly, the query functionality may increase the time taken to call the method.  For example the CanConnect method might take as long as a call to Connect, in which case if this is routinely called everytime you need to connect it may well negatively affect performance. I guess there is a trade off here the again has to be managed.

My last thought regarding exceptions is that I still believe that .NET is missing a feature that I loved/hated in Java (and yes for my sins I did do a bit of Java programming way back when it was all the rage).  This feature is the way that methods specify which exceptions may be thrown. The reason I say loved/hated is that I loved it when I was building robust code, as it forced me to think about how I was using that method and whether I had handled the different responses (including exceptions) from the method. 

I hated this feature for two reasons.  Firstly, that it was a pain cause the compiler insisted that I rethrow exceptions that I didn’t handle in my code (for example if I used a method that could throw an  IOException I either had to catch this or specify that my method could also throw this exception) – why can’t the compiler be sensible and deduce this for me, after all when I use that method it’s not that I look at the method itself to work out what exceptions are thrown, that’s again the compilers job (or intellisense – does Java have that now??).  The second reason is that some developers are sloppy and just say that every method could throw Exception – kind of defeats the purpose of having this feature, right!!

Now it would appear that we may have missed the boat as it would be too difficult to go back and force every .NET method to specify which exceptions are thrown.  However, this is an area where tool support might provide a good compromise.  Now that both C# and VB.NET support rich xml commenting it may be an opportunity to really make use of this functionality.  By specifying in the comments which exceptions a method may throw, this could be used by the IDE to warn developers if they haven’t handled theses exceptions when a method call is made.  As with most othe IDE support this could be an option that can be configured by the developer or dev team according to their preferences.