Lights out for Silverlight

Editor's note: ZDNet Australia is running duelling perspectives on the Microsoft Silverlight vs. Adobe Flash debate. To find out why project manager Scott-Bradley Pearce thinks the Flash is out for the count, click here.

commentary It's been a couple of weeks since the full announcement of Silverlight took place -- now that other players have shown some of their cards and the dust has begun to settle, what can we take from it?

The one point that is screaming out and demanding attention is that no matter how much good Microsoft does, invariably someone, somewhere and ultimately from within, will shoot them in the foot.

Allow me to explain. Despite the rhetoric out of Redmond to the contrary, it is clear that Silverlight is Microsoft's equivalent to Flash -- and when it comes to video, it rivals Flash in capability. The video leadership is a good thing for Microsoft and something that they are trying hard to get across.

Unfortunately, it's not all that easy.

Firstly, Silverlight isn't cross platform. It is currently only available on Windows, OS X and Windows CE -- this reduces it down to essentially the Windows platform plus one. "Plus One" is technically enough of a basis to claim cross platformness but it is hardly a convincing argument.

Silverlight is only in beta at this point and more platforms are planned, but they are mostly on mobile devices. The yawning gap here is the non-availability of Silverlight on Linux. Microsoft's reason for this is one of "priorities" -- a justification that seems to be recycled directly from the Adobe camp circa Flash 8. Adobe appears to have learned the error of its ways with Flash 9.

If Silverlight were to appear with Windows, OS X and Linux versions available then it would be a concrete sign of Microsoft's wish to be all about "choice and interoperability". It would also quieten a vast number of potential detractors from the Linux camp. At a worst case, Microsoft may not gain anything positive from it, but it would alleviate much of the current and future criticism that will be thrown at Silverlight for its Windows and OS X only stance.

The irony here is the technology is so close to being on Linux anyway. There is no hardware acceleration needed for Silverlight, and it is merely a browser plug-in. What is the difference between a plug-in for Safari or Firefox on top of Darwin compared to a plug-in for Konqueror or Firefox on top of Ubuntu? For a company with the resources of Microsoft, it cannot be much.

The next point is the availability of the tools. Even though you can view Silverlight on OS X, you will have no chance of being able to author Silverlight content without a Windows licence. This is a conscious decision. On the one hand, Microsoft wants to get designers using their Expression toolset yet designers will have to come across to the Windows platform in order to do it. It may not be such a large hurdle but it is a hurdle nevertheless. Ever tried to force an OS X user onto Windows? They cry, they scream and they want their (at times) consistent GUI back.

And what is stopping Microsoft from porting Expression to OS X? Only that they will then have to port WPF and a complete .NET runtime onto OS X. This is currently regarded as akin to handing over the crown jewels, and will not happen soon.

However, Silverlight is currently shipping with a cut down .NET runtime on OS X. In the MIX keynote, demonstrators gleefully showed how to debug and break the running of Silverlight inside Safari from within a Windows machine running Visual Studio. The financially minded members of the audience would have immediately noticed the problem with this picture. If you are a developer or designer on OS X, you will need to have a new computer that runs Windows to truly create and debug the full gamut of Silverlight possibilities; or upgrade your Mac to Intel and install virtualisation software.

The endgame means that if you want Silverlight, you will need Windows, no ifs ands or buts.

The final problem that faces Silverlight is one of distribution. Microsoft is adamant that they will not leverage Windows Update in order to push out or give users the optional ability to install Silverlight. Perhaps they are wondered if the US Department of Justice would not look favourably upon it, though the .NET Framework is an optional download for users and no one seems too concerned about it.

Instead, Microsoft will be relying on its partners to make users install Silverlight. For the near term that means unless you visit mlb.com, cbs.com or perhaps a Silverlight technical demo site, you will not need Silverlight. For most people this will mean that Silverlight will not exist -- the mindshare will only be with developers. Compare that to Flash or Java. If you are coming to the game late you need a big entrance to overcome the inertia -- slowly waddling out of the gate will not cut it.

The one saving grace for Silverlight is if a YouTube-esque site creates an amazing video experience with Silverlight and it really takes off. But why would someone do that when there is no need? Flash and Java are everywhere, why would you go against what you have been using until now?

Betamax has shown that technical superiority can be beaten with a good dose of distribution. Make it good enough and ubiquitous, and that's all that is needed to show the technical upstart the door.

Therefore what is left? We have a video platform held back by a lack of distribution issues, an enforced lock-in for the tools and a lack of true cross-platformness.

When it is put that way, it begins to sound an awful lot like Microsoft's previous video strategy with Windows Media.

There is a reason that YouTube is built on Flash. Until Silverlight presents a compelling reason for YouTube to move away from Flash, it will not be the front-runner -- even if does look cute as a button and is technically superior.

For the opposing viewpoint, see this story by Scott-Bradley Pearce.