Assume, for example, that you are developing your own mobile app (or apps, in my case) and want to be able to send push notifications to these apps every once in a while.
You would require a server that will run somewhere in the background, collect new information every given interval and send relevant notifications to the client apps. You would also need to allow the mobile apps to register and unregister from that server, and supply a simple management console user interface for operations such as browsing though your logs, changing intervals and behaviors and so on.
Thinking it through you realize that you also require a SQL database, that the server has to be fully secured (you would not like anyone hacking in and pushing messages in your behalf into your mobile app, used by unknown users..), and that it has to be reliable and always ready.
Now, you could start by trying to locate a suitable and available server(s) and get it allocated for your purposes, or go out and purchase one. You would also have to make sure that the server (or servers) is licensed and installed with all that is required, from operating system through database engines, all the way to the web-server and other packages with the latest version. Your server should also be exposed to the internet but yet secured.
But even when you finally have that server running, who is going to make sure that the server is always up, operational and taken care of? What about variable bandwidth and storage capacities?
A simple and great solution for your background processing – Windows Azure.
For my server, I start off by creating a Windows Azure SQL Database (No need to put up a fully secured, reliable, always ready SQL Server – CHECK!).
Back in my Visual Studio 2012, I use the “Server Explorer” tab to further create the tables, relations, keys etc., as Windows Azure SQL Databases act as any standard remote SQL server databases, just in the cloud. (No special tools – CHECK!)
Two pieces were still missing from the puzzle: the background service and a management console. Simple solution there – I create a Windows Azure Cloud Service composed of a Worker role and Web role. The Worker role would be used for background processing and the Web role would serve as the management console and Web-API host.
NOTE A Windows Azure Web Site would not be a good enough solution for these purposes as it becomes dormant after a short timeout of not receiving new calls, thus any background service will stop working.
You have two options: create both elements through the Windows Azure console and download a ready-made Visual Studio project, or create what you need in Visual Studio using wizards and have Visual Studio take care of creating whatever is required on Azure for you.
I usually choose the latter as it is easier for me to do it all from the Visual Studio environment, where I have everything I need. You would need to install the Windows Azure SDK in order to have the cloud options in your Visual Studio.
Creating a cloud service is as simple as “File/New/Cloud/Windows Azure Cloud Service” (assuming you already have the required live account):
and selecting the desired roles:
The wizard creates ready-made solution and projects for the selected roles, already set to work with Entity Framework 5. The Worker role project already contains stubs for the “Run” and “OnStart” method overrides, allowing you to inject your business logic into the background service, while the Web role (if the preferred MVC 4 pattern selected) is already a full-blown website, with REST support, controllers, models, and everything that you need to form a secured websites, including pre-defined ASP.NET membership.
At this stage, both projects can already be published to Windows Azure. You only have to worry about your specific code. You can also debug both projects using the Visual Studio and the “Windows Azure Compute Emulator” and “IIS Express”, that are installed as part of the Windows Azure SDK install.
So, what do we have here?
We just setup a WebAPI server, a website and a background service running against a SQL server in record time and with minimal effort. Having all of them running on Windows Azure, we also ensured reliability, manageability and consistency for our services.
Using the above steps, it took me less than a day to complete the rest of the code logic and get a push notifications server up and running. Want to see it in action? Check out the Toronto Events and Festivals Windows 8 and Windows Phone 8 apps as well as the Trashswag Windows 8 and Windows Phone 8 apps. Both use this background processing cloud service architecture to collect new information and send notifications to the apps.
If you’d like to go deeper into this concept, I’ve put some demo code on my SkyDrive. Going through the examples, you’ll see that there was very little code required in order to get Windows 8/Windows Phone 8 apps connected to the WebAPI based Windows Azure app. Please note that the WebAPI MVC 4 project uses a bunch of NuGet packages, automatically installed by the wizard within my project. Write me or comment below if you have any questions or need some help.
Featured on MSDN Canada’s Canadian Developer Connection
For those of you who already have some experience developing Windows 8 apps, I compiled a (partial) list of changes that I find notable on the UI side. For those that haven’t, the below are things that you should absolutely not think twice about implementing in your apps.
- Tiles now have two new sizes, while older ones has been renamed. The new sizes : “Small” and “Large”, while “Normal” was renamed to “Medium”.
- This does not mean that your old code would stop working, as the older names are still mapped to new ones in the enums for backward compatibility, but if you want to support the newer tile types, you would have to update your code.
- The only restriction is that in order to support the Large tiles, you will also have to support Wide tiles in you app.
Multiple windows/No more snap-view
- Snap view was replaced by a variable size window mechanism allowing the user to have multiple apps and windows running on the screen at the same time, with different widths, while an app may also spawn other apps and share the available space with it.
- You now have to the following new properties in the ApplicationView class: AdjacentToLeftDisplayEdge and AdjacentToRightDisplayEdge used to determine if your app window is adjacent to one of the edges, IsFullscreen and Orientation.
- The Search charm, while still there, will now act more of a global (computer-wide and/or web) search, and less of an in-app search. Visual Studio 2013 now offers new SearchBox control to be included in the apps themselves and used for in-app searches. Older, search-ready Windows 8 apps will still work with the search charm as they did before, but in order to search within the app the user would have to run it first and select it as target on the search target combobox. Not as intuitive as it used to be, or as using the in-app SearchBox control.
- Flyout: No need to use third party vendors’ flyout controls anymore. The Flyout control allows you to display a floating window which is not a dialog (unlike MessageDialog). Deriving from the Flyout control, you now also have the MenuFlyout and the SettingsFlyout controls.
- Hub: New navigation control allowing you to to aggregate different types of information controls within one content control, therefore controlling the width of your app’s main screen. For example: Displaying a GridView of images in one hub section, a ListView of detailed information in another, and a form on a third section.
- AppBar: Windows 8 already allowed you to include various controls in the app-bar, but Windows 8.1 offers the AppBarButton, AppBarToggleButton, and AppBarSeperator controls right out of the box to be used as part of the new CommandBar control within your app-bars. So, what is the difference? It’s all about styles and conventions. Using the new controls allows you to have your app-bars look and behave in a more conventional way while the CommandBar control automatically lays out the commands on the bar,
As you can see, developing for Windows 8.1 is definitely not different than developing for Windows 8. It is actually easier. Migrating your apps to Windows 8.1 is mostly performed automatically, and you can make the required changes to support the new features in a fraction of the time it took you to develop your app originally. I would definitely suggest migrating your apps to allow your users to benefit from these improved UI capabilities.
While I did not cover all of the changes in Windows 8.1, and not even all of the changes to the UI, what you do have here are the ones that will make your development easier, while, and most importantly, enhancing the experience of your app for your users.
Featured on MSDN Canada’s Canadian Developer Connection
Have you ever tried to run/debug your Windows Phone 8 apps when using Hotel mode Wi-Fi?
Hotel-mode Wi-Fi are guest access Wi-Fi routers that do not require a password to connect, but rather the first time you attempt to open a web-site you are routed to a login form, where you’re prompted with either a confirmation page or a user-id/password page.
So, where’s the problem?
You have your Visual Studio 2012 open, you already connected to the Hotel-mode router and made sure to open a page (google.com/microsoft.com/whatever) in order to go through the intermediate page and get internet connected.
Now, you launch your app, causing the emulator to load (takes about forever…) and the app to run just to find out that your app cannot connect to the internet!…
Checking again, you discover that your computer somehow reset the connection with the router and you have to go through that login page again. But, that’s not enough, as the connection will keep resetting every few seconds, making it almost impossible for you to run and debug your app, assuming it requires internet connection.
So what can you do if you just have to debug or demo your app from the emulator on such a network, assuming you do not have a
After many tries (and failures), I found only two possible solutions:
The first is to use your phone’s mobile data and tether the internet connection through Wi-Fi to your computer. But, mobile data doesn’t come cheap and you would rather use that complimentary Wi-Fi for your tests.
The solution I found was to run a USB tethering app on my Android phone, have the phone connect to the Hotel-mode Wi-Fi and tether the connection to the computer using cable.
That way, the connection is not being reset when using the emulator and you can run/debug your Windows Phone 8 app.
Thanks to everyone who attended my “Azure as the backbone” web-api based talk today @DevTeachConfere in Montreal.
Demo code can be found on my sky-drive here.
See how easy it is to create a web-api based app using the Visual Studio 2012 wizard, deploy it to Azure and then create Windows 8 and Windows Phone 8 apps that uses the service from within Azure.
Going through these examples you’ll see that there was very little code required in order to get these Windows 8/Windows Phone 8 apps connected to the web-api based Azure app.
Please note that the web-api MVC 4 project uses a bunch of NuGet packages, automatically installed by the wizard within my project.
Write me or comment for questions!
Thanks to everybody who attended my session about practical kinect at the Dev-Teach conference (@devteachconfere) today, to hear about the Kinect SDK, the different types of interactions and the way to implement them in Windows.
The demo code can be downloaded from my SkyDrive: here
The demo code includes two solutions:
KinectPointers: demonstrating how to react to hand movement by moving the mouse and generating a mouse click upon a 1 second hover at a certain position
KinectWPFDemo: demonstrating how to create a media-player that uses postures to play, pause, fast-forward, rewind and alter the playback volume.
I’d love to hear your suggestions for practicle apps using the Kinect SDK!
Microsoft Kinect SDK is claimed to work under Windows 8.
Well, before you run-off starting to develop your next Kinect multi-touch experience for the Windows 8 UI, it is important that you know that there are limitations to that.
Although using the latest Kinect SDK driver, released October 2012, you can develop and run apps under the Windows 8 operating system, it only works for developing traditional desktop apps, meaning what you would call “Windows 7 apps” and not what is called: “Windows Store apps”, also formerly known as “Metro-style apps”.
The Kinect libraries are not compatible with Windows Store apps and therefore you cannot add a reference to a legacy library that works against the Kinect SDK. Visual Studio 2012 prompts you with an error message saying: “Unable to add reference”.
I’m currently working on a work-around that may enable a small portion of the Kinect features in Windows Store apps and will surely update when I come-up with something.
Latest Kinect SDK information and download to be found here: http://www.microsoft.com/en-us/kinectforwindows/develop/new.aspx
This release puts Microsoft’s flagship cloud offering in great competition with Amazon in the IaaS market and with Google App Engine in the website market.
On the coming Sela Dev Academy I will be training a whole day tutorial in which I will present the new Azure portal and demonstrate the new features in the SDK.
Are you already using Azure? Are you developing against the SDK?
If you’re using Amazon or Google’s platforms you might want to hear this one out.
So, you’ve created an amazing new app for the Windows 8.
Everybody loves it and tell you that you should make it public and that people should pay you to use it.
But, how do you do that? It sounds complicated, isn’t it so?
It sounds like something companies does, not private people or small groups!
Well, it’s not that complicated and surely everybody can do that. It just takes some learning and a few correct steps and you’re there, at the app store, with your brand new app waiting to be downloaded and purchased!
So, to help you jump the hedge, I’ll be giving a lecture @devacademytoronto titled: “From a New Windows 8 Project to the Store“, in which I’ll show the list of actions to be taken in order to have an app in the store, explain the methods and show the pitfalls to avoid along the way.
See you there!
I’m very excited to announce the “SELA Dev Academy” conference that is going take place in Toronto on Nov 5 through 7!
The Dev Academy will host Sela’s professionals, as well as Microsoft Keynote speakers, for a one day of breakout sessions, followed by two full tutorial days.
Continuing a long lasting tradition in the SELA Group of local technological conferences, this conference, hosted by Microsoft Canada, is going to cover Client, Server, and ALM issues, such as: Windows 8, the coming Windows Phone 8, Parallel programming, Azure, WCF,Debugging, Agile and TFS, and will bring the latest news from the Microsoft front.
Registration is now open!
See you there!
Another session to prepare for the coming “DevTeach” in Montreal- “Developing Kinect interface for common tasks”
I was just scheduled to give another session at the DevTeach Conference in Montreal (Dec 10-12 2012), titled: “Not only for games! Developing Kinect interface for common tasks“.
Microsoft Kinect hardware is a fun way for playing games, but not only games.
Using the Kinect SDK for Windows allows developers to use the same hardware for enabling exciting new ways to interface with their apps, in addition to the mouse and keyboard.
In this session we will go over the different ways to interact with the user using the Kinect and demonstrate how to develop an interface that will extend an existing app to accept user commands via the Kinect.
I’m going to enjoy waving my hands up and down in front of all of these people, and for strictly professional reasons! 🙂