Silverlight is dead. F****** dead

Or maybe it isn’t that dead? OK…you expected me to say that, and you’re thinking, “this guy is just trying to sell me Silverlight tools and save his business.” I will not argue…you’re almost right. There’s one small correction to be made - we are not trying to save our business. :) Rather, we are hard at work on all fronts, be it HTML5 or Win8, so it doesn’t make a difference for us what technology you’ll ultimately choose as we have all bases covered.

As I’ve said many times, use whatever makes most sense for your business. That said, I do want to share our thinking on Silverlight so that you can make a more informed decision.

At Telerik we still believe in Silverlight, and we are betting on it. We are not only continuing to sell our existing Silverlight controls, but we are also continuing to invest in the ongoing development of new controls for Silverlight. We’re doing this because we believe Silverlight is not dead…at least not yet and not for a long time.

All the Cool Kids are saying that Silverlight is dead, or at least assuming so because it got zero sessions at BUILD. I think a product is dead when most of the things below are true:

• When the platform vendor drops it and leaves it without any updates for a long time
• There are significantly better alternatives on the market
• When all ISVs that focus on a platform abandon it and the ecosystem dies
• Everyone stops using it

Let’s think about each of those points.

Microsoft’s Strategy

Let’s say that Microsoft stops adding features to the Silverlight plug-in after Silverlight 5, which is still slated to ship later this year. With SL5, developers finally get a super mature platform to build line-of-business apps. The tooling is there. The performance is there. The stuff will run on Windows 8 (rumor has it that desktop mode will run even on ARM devices). Add this to the Microsoft support guarantee and you have viable platform for many years to come.

Say an app has a 5 year lifecycle. If you start today with SL/WPF, the underlying platform will still be very relevant past 2017. In fact, even with no additional attention from Microsoft, Silverlight and WPF are well positioned to serve businesses for years (decades?) to come. The best testament to this effect is WinForms.

Windows Forms has received pretty much zero love from Microsoft since 2005, but that hasn’t stopped many businesses from using it as their preferred platform for building LOB apps. Even today, many people use Telerik tools to build WinForms applications, and I don’t see this number drastically decreasing. Remember, that’s 6 years after the platform got its last big update.

To further clarify the support issue, Silverlight has a 12-month support commitment from Microsoft. That means if Microsoft formally announces the end of Silverlight, you still get direct support for 12-months. Visual Studio, including its tools for Silverlight development, has a 10-year support cycle.

The Alternatives

Let’s talk about the second point – the alternatives to Silverlight. If you are building a LOB app, what is your best option? Flex, HTML5, Silverlight, WPF, WinRT+ JS/XAML?

• If you are looking at Flex, you are probably not the ideal customer for Silverlight anyway.
• WinRT is nice, but it has a couple of major obstacles:
  • First is that Win8 is not just around the corner, and for now, you don’t have anything to build against. As Win8 progresses through beta 1/2/3, many things will change, requiring you to invest countless hours to keep your stuff running. In short, WinRT just isn’t ready.
  • Second, if you want to build a Metro app, you really need to spend a lot of time understanding the Metro guidelines and the target application scenarios appropriate for Metro. You can’t just port your existing SL/HTML stuff to run in Metro mode. It’s useless and it will kill the whole experience. And if you want to run in Win8 desktop mode, “normal” Silverlight is waiting for you. :)
  I won’t even touch upon the fact that Win8 targets mostly consumer apps in Metro mode and that it will take 3-4 years for it to get ANY meaningful adoption in the Enterprise space. Telerik EVP Doug Seven has a great blog post about the target app types for Metro.
• Another option is to use HTML5 plus .NET on the server. That’s a viable alternative, but unless you need your application to be accessible on every device imaginable and you have control over your environment, it’s much easier to achieve the same result with Silverlight. No browser compatibility issues, no lack of tools (yes, I know HTML/JavaScript tools will catch up, but we are talking about your options today and 2-3 years ahead). HTML5 is great, but it’s still maturing.
  Silverlight gives you the best of both worlds today – a rich presentation platform + the deployment/update story of an HTML application. You also have a clean logical separation – pure data is coming from the server and a presentation layer that is built on top of that data. Add to that the fact that Silverlight lets you easily integrate your existing .NET codebases (or even native code via PInvoke) and it makes it the best choice for LOB apps.

The Ecosystem

Moving on to the third point: the Silverlight ecosystem. This one is pretty important. Even if Silverlight 5 is the last version of the plug-in (which, for the record, we don’t believe it is), at Telerik we have a long list of things that we plan to do for the future that will improve your Silverlight development experience. If you look at our roadmap, it is pretty packed. There are also many other Silverlight-based things that are super big chunks of work and will ship in 2012 (I can’t share more as I will spoil the surprise but it is MAJOR stuff).

We don’t plan to stop, and this will only make it easier to build anything using Silverlight in a fraction of the time of other platforms (such as HTML5). From what I can tell, all other major vendors also plan to release a lot of new things in Silverlight moving forward, too. This means that when you are using SL, you are in good hands as the ecosystem that enriches the platform is standing behind it, even if Microsoft is slowing their investment in the platform. Our work WILL have a very tangible effect on your productivity.

The customers

Finally, point 4, customers and platform adoption. Ultimately you are in control. For as long as there is customer demand for Silverlight, we will keep on creating the best tooling for SL. And as exciting as BUILD is for the long-term future of Microsoft, we don’t see it having any impact on customer demand for Silverlight. Of course, we will also provide a complete toolset for WinRT when Windows 8 ships, but we very much believe that Silverlight is the way to go today if you are building Enterprise LOB apps.

Silverlight is not dead for us. How about you?