This post was imported from my old blog and had 54 comments which are included as a screenshot at the end of this post.
Latest update (and hopefully last update): Microsoft has published a few posts to clarify the debate. Please consider the following:
- PDC and Silverlight by Bob Muglia
- PDC Thoughts by Steve Ballmer
- *NEW* Silverlight Questions by Scott Guthrie
- Silverlight is dead. Long live Silverlight! by Tim Heuer
- Committed to Silverlight by John Papa
Disclaimer: I am a Silverlight MVP and book author, and as such I have an interest (some would say a passion) for Silverlight. Also, I do have a privileged relationship with Microsoft, giving me early access to tech previews and confidential information. Obviously I cannot talk about these in public forums.
Update it was pointed out to me that it is not fair to MaryJo Foley to use the word “incorrect”. I agree and changed the following paragraph a bit.
Another update: MaryJo commented here.
That said, I would like to encourage the public to think with their brain instead of merely rehashing information posted by tech journalists. We all know how tech journalism works today. The goal is to publish as fast as possible, even if the statements are taken out of context. The only thing that matters is audience. I saw this personally a few times, with statements I made to tech journalists being distorted in one way or another, and sometimes even plain wrong statements being attributed to me.
One fact is that HTML5 is in its infancy. At this stage, if we consider it carefully and with a cold head, anything can happen. It could die (I certainly hope not but it is a possibility). It could be hijacked by web browser makers into multiple platforms with proprietary extensions (if there is anything that the past taught us, it is that web browser wars have hurt us many, many times). It could, of course, evolve into a beautiful butterfly running everywhere. However the point remains: At this stage, it is simply not ready for prime time.
Update: The W3C also thinks it is too early: http://mashable.com/2010/10/07/w3c-stalls-html5/
For instance, let’s consider the demos we saw in the past few months: At PDC Microsoft underlined that the demos they showed run well only on IE9. In fact, many of us thought it was a weird way to push a supposedly standard technology, by showing how badly it runs on other platforms (see http://twitter.com/#!/LBugnion/status/29003780660). Similarly, the demos made by Google a few months ago ran only on Chrome. The video streaming during the latest Apple conferences required OSX or iOS. This is not a standard technology yet, by a long shot. If it remains this way or not needs to be seen.
Another fact is that there is no good story around development tools for HTML5 at the moment. I am sure things are coming, but the fact is that it is still much too early. On the other hand, the development tools for Silverlight are simply fantastic (pardon the bias, but really they are. I am happy to demo these tools to anyone interested, just ping me).
The web is made of a multitude of platforms, browsers, cultures, and of course technologies. In some cases, it is necessary to select a technology over another, depending on technical or commercial considerations. For instance, in the current time, Netflix simply cannot be written in HTML5 (notably because of the DRM issues). Silverlight has other strong advantages, which explains why we saw many web applications running Silverlight (for example the recent advance in smooth streaming allow to stream videos at conferences and events adapting the image quality to the bandwidth available, to pause, go back in time, etc…). Finally, let me mention that at the moment, if you want a rich app to run on a very large number of web browsers in the exact same way, plug-in based solutions like Silverlight have a clear edge. We may debate endlessly if that it a good or bad things, but the fact remains.
Also, let’s not forget the fact that Silverlight is not just a web technology anymore. While it does run in web browsers (and does it really well), we saw a shift (ah, this word…) in the demand, and we are developing more out-of-browsers applications in Silverlight 4 than in-browser apps. I expect that this trend will continue, especially with the convergence of Silverlight with the other large desktop application technology that Microsoft has, Windows Presentation Foundation. In addition, Silverlight runs on other platforms, the best known being that it is the major development platform for mobile devices at Microsoft.
In conclusion, before burying Silverlight (which Microsoft certainly did not do), I think we need to pause for a moment and consider the options. Any tech prediction for the next 2 years is hard. Any prediction for the next 5-10 years is almost impossible. We all have things that we wish and hope for (I do too), but in the mean time, it is best to keep our feet on the ground, keep a cool head and just, you know, think.
I want to finish with a thought that ran through my mind many times yesterday: By a funny coincidence, a Rally for Sanity was organized in Washington DC by Jon Stewart and Stephen Colbert (two US satirists and observers of the political life, http://www.comedycentral.com). Now don’t get me wrong, I am aware that the stake is much higher in the case of US politics, but still, it was amazing to see how much the mechanisms that this rally attempts to fight are similar to what happened around Silverlight. The role of the press, the “out of context” statements, the immediate rage that ensued on Twitter and then in blogs, the comments full of bile (most of the with words such as “dead” or “die”)… this is exactly similar to everything that Jon Stewart has been showing (and fighting) on his show and at the rally. The Internet (and humanity in general) is prone to such excess. It is time to go back to reason and sanity.
Thanks for your attention,
Laurent Bugnion, Silverlight MVP of the year 2010