We just started a brand new year that will surely be very exciting. Looking into my crystal ball, here are a few predictions for 2009.
ASP.NET is dead
You should drop ASP.NET and start using Silverlight right now because it is clear that Microsoft has no roadmap and future plans for that technology. At the last PDC, nothing new ASP.NET was announced except a few tweaks for AJAX and the fact that Microsoft released the MVC Framework proves that but IMHO, the MVC Framework is just a toy technology to shut the Agile loudmouths that wants to code everything by hand and reinvent the wheel.
BizTalk is dead
With Dublin coming up with the Framework 4.0, it makes no sense to purchase and invest in BizTalk anymore. Why spend gazillions of dollars when you’ll have the exact same services for free right in the Framework? Why being locked with a product when you get the API for free?
LINQ 2 SQL is dead
Microsoft recently stated that they have stopped working on LINQ 2 SQL to focus on the Entity Framework so why would you use LINQ 2 SQL anymore? Once again, Microsoft has released a new technology, people invested in that technology and Microsoft abandoned it without warnings. I’m not happy because I invested so much in LINQ 2 SQL so I think I’ll stick with ADO.NET fearing that Microsoft will eventually dropped the Entity Framework as well.
VB and C# are dead
At the last PDC, Microsoft showed Oslo, its modeling platform along with M, a modeling language. This will revolutionize the way we design and build applications. In the near future we will no longer code using low level languages like VB and C#. I would stop investing in these languages right now and train my whole team in M.
Azure is DOA
Cloud computing, talk about something so stupid its name is stupid. Why the heck would I push my apps and data in Microsoft’s data centers? I don’t get it and I think no one will use that technology and trust Microsoft. This thing is dead even before it will be released (.NET My Services anyone?).
OK, you must have seen similar comments everywhere in forums, blogs or talking with colleagues and other developers but is any of this making sense?
AP.NET vs Silverlight vs MVC Framework
When I was at the last PDC, one thing that struck me was the fact that no major new announcements were made for ASP.NET. Clearly, this is the sign of a mature technology but at the same time, Microsoft is putting a lot of resources and efforts in building a new platform in Silverlight. Does that mean that you should drop ASP.NET? Of course not: it’s mature, it’s stable, hundreds of 3rd party controls are available, it’s proven and you can find thousands of developers familiar with that technology. Why would you stop using it all of a sudden? But why is Microsoft working so hard on Silverlight? Because Silverlight fills the RIA gap in the Microsoft offer because developing complex RIA apps with AJAX is way too costly right now. So why is Microsoft working on the MVC Framework, a direct competitor to ASP.NET? Again, another model to solve different problems. It’s easier to use if you’re deep into TDD, it’s somewhat lighter but at the same time, it’s somewhat a return to Classic ASP. Well, it’s another tool in your .NET toolbox and this creates confusion for people who think in the “one size fits all / one ring to rule them all” way. Again, look at your project/assignment, analyse the problem/need and select the right technology to do the job. But in the long time, could Silverlight be so successful that it would replace ASP.NET? Maybe or maybe not. It certainly has the potential to do so but who knows at this point in time!
2009 prediction: ASP.NET will dominate and we will see a slow but steady Silverlight adoption. The MVC Framework will remain a niche tool for Agilists for now (and I’m not saying this in a pejorative way).
BizTalk vs Dublin
The .NET Framework 4.0 will include a set of extensions to Windows Server that will add new services to Microsoft’s app server. Features that will simplify the deployment, configuration, management, and scalability of WCF and WF applications. Basically, some of the same features found in BizTalk Server. So why would I bother with BizTalk and spend big money when I can get the same services for free? Well, BizTalk is an integration server and Dublin is a set of management tools extending Microsoft app server. Dublin will allow you to deploy more easily your WCF apps so instead of writing a Windows service to host your WCF services, the app server will be able to host them directly.
So you think you can write a BizTalk clone? Remember how BizTalk 1 was a piece of ****? And remember BizTalk 2? Microsoft only got it right with version 3 so yes, you can write a BizTalk clone using Dublin but it will cost you a lot more than purchasing BizTalk because Dublin and BizTalk are two different animals.
2009 prediction: Dublin will be widely used right away to help host WCF apps. BizTalk will not die and some people will get burned by trying to replace BizTalk with Dublin in a scenario where the use of BizTalk would have make more sense. FUD will run loose on the Web.
LINK 2 SQL vs Entity Framework
So Microsoft created some confusion by releasing two similar ORM technologies? Yes of course but if you look closely, they are somewhat similar and different at the same time. LINQ 2 SQL is RAD against SQL Server and the Entity Framework is the full blown ORM thing. Both use LINQ as the query language so if you invested in LINQ 2 SQL, your investment is not wasted and for God’s sake, LINQ 2 SQL is in the Framework; it cannot die!
One suggestion that I saw on the Web was to release LINQ 2 SQL on CodePlex so that people can update it since Microsoft has “abandoned” it. First, Microsoft never said that it has abandoned it. They will keep it as it is right now and they will put all their efforts on the Entity Framework. My hope is that when we’ll see new SQL Server releases, Microsoft will update LINQ 2 SQL to match some of the new features. So back to the CodePlex idea: I think it’s a bad one because people will try to match the features of the Entity Framework or Nhibernate and that will create more confusion because we will have similar products competing against each other.
LINQ 2 SQL = simple 1 to 1 mapping against SQL Server
Entity Framework = complex mapping against SQL Server and more
2009 prediction: people will still be confused and FUD will run loose on the Web
C# and VB vs Oslo
Microsoft have finally abandoned the traditional design surface where we connect little boxes as the way to model apps and I think that Oslo and M are very good concepts. Is this the modeling Holy Grail? I have no idea but it looks good. Can Oslo and M replace VB and C#? Of course not!
Every time that Microsoft launched a new way to model apps, these technologies never lasted more than one version. I would tend to have a wait and see approach. Take a look, experiment and wait for version 2.
2009 prediction: Olso and M will be curiosities. Let’s hope that they will not die as the other Microsoft modelling oddities.
Azure vs apps hosted in the enterprise
Microsoft has taken a bold approach to cloud computing. Instead of hosting your virtual machines running your apps, they will host your apps in a new cloud OS running in their virtual machines. But why would you do that? To drive IT costs down of course. Look, there’s something called a recession going on right now and it’s a nasty one. At some point, your boss will be asked by his/her boss to cut expenses drastically and maybe one way to do it is to go the cloud way. We’ll see but there’s one thing that I’m sure: this will be another tool in your .NET arsenal and it will not make sense to write all apps to run in the cloud. Again, look at your project/assignment, analyse the problem/need and select the right technology to do the job. I’m sure that people will get burned by using the cloud model on projects where it doesn’t make sense.
2009 prediction: Microsoft has a lot of work to do to convince people to move to cloud computing.
[Edit] Of course, the first part of this post was a little provocative and it was meant to shake things up and create some discussions. All ideas are welcome and remember....there's nothing like "one size fits all".
So these are my predictions for 2009. What are yours?