Will this one work right with Firefox, though?

The world doesn't need or want silverlight, we already have Adobe Flash, thank you. Its just more advertising pain from Microsoft trying to get one more "service" running on your computer while surfing, and slowing down the internet. Not to mention, if silverlight will be another spy vehicle for Microsoft, and they really like to spy on their users.

A good friend who is a superb developer, has a long history of building, selling, and supporting his own products, and now has his own successful and growing software development company, has told me that "Silverlight rules." He simply loves what it can do, and even more, he loves it from the perspective of what he can deliver as a developer and vendor. To him, Microsoft has a clear and solid winner in Silverlight.

And I am a Unix and Linux adherent. For me, they pay the bills. Not everyone in the world thinks servers and applications should be rebooted once a month--or sooner--to keep them running and to keep memory usage low. They expect their critical software to just work, and to keep working without on-going maintenance, restarting, or rebooting.

My use of Linux as a desktop and laptop OS gives me a vastly better development environment for Unix and Linux based products than Windows ever could. And, over the years it is also increasingly closer to the ideal desktop. Even getting Flash 9 to run within a 64-bit Ubuntu 8.04 WITH 74-bit Firefox was a walk in the park; and vastly easier than on Ubuntu 7.10, and not worth the effort on earlier versions of Linux.

Which is why I believe that Microsoft desperately needs Silverlight to be nothing less than compelling. It must sparkle and shine. It must run flawlessly, beautifully, and make developers think they have gone to heaven. Because on nearly every other front, Microsoft has failed to shut the door tightly on Linux desktops in businesses. But if people equate "Silverlight" with "gotta have it to live", then they will win at least one major battle against their Linux-based competition.

Winning this battle will take technical excellence, business excellence, and the full weight of their monopoly. With Silverlight, I believe that they finally have the first two nailed down perfectly. The rising groundswell against the abuses of their monopoly position might be the only weakness in the third leg of their strategy. But with their huge market cap, they can purchase a lot of government officials, and shouldn't have too much trouble for a while.

Well those OSS dudes are making it happen with Moonlight. Microsoft doesn't even need to lift a finger to cross-platform it's er.. platforms these days with Moonlight and Mono.

I can't help but wonder what these same obviously smart guys could have done instead of copying a moving Microsoft target. I also have to wonder what they think they're achieving with this...

I hope that the "OSS dudes" you are referring to are better than those who put together the Evolution email client. That client is basically unusable, even though it has a lot of potential and a really nice UI paradigm.

Two reasons why Silverlight will be not rise beyond a Flash competitor.

1) cost - the cost of implementing an application in Silverlight is many multiples of a traditional webform or winform application. Business application buyers will not see the benefit of the added cost and will bypass it.

2) performance - WPF is sloooow. Users will not accept a sluggish interface. The graphics fly (hey its still MS Media Player at the heart of it) but the business apps created from WPF controls crawl.

Sam, you say the world doesn't want Silverlight. You can make a claim like that, but the numbers say otherwise. Last month, the Silverlight plugin was downloaded an average of 1.5 millions times per *day*. While you and the people you know might not want it, the numbers show the same isn't true for the rest of the planet. (I'm not making a judgment call here; I'm just pointing out the numbers.)

Phil, about the cost, if you're already developing webforms and winforms, that implies you're using Visual Studio, which is also what you can use for developing Silverlight apps. But I'm not sure what you mean about "business application buyers." There's no additional cost to executing a Silverlight application; all you do is download the plugin. So if you develop an application and a business purchases it, there's no additional burden on them if the application uses Silverlight.

Also, I forgot to add -- for other people reading this, contrary to what was implied in a comment, Silverlight isn't just a front-end for Windows Media Player. That's not correct.

Jeff,
When you post an opinion don't you think you should identify yourself as being sponsored by Microsoft?

As to the cost of buying business applications; nice obfuscation buddy! As commonly used "Business Application Buyers" is understood to include those businesses who develop for internal use and not for resale. The cost being addressed is obviously the development cost and not the cost to run the application. The fact that you chose not to address the increased development cost I will take as confirmation of the fact that the cost for these people will be multiples higher. Honestly, for someone calling themselves DevSource you've made a ham handed attempt to deflect a legitimate concern of businesses when it comes to using Silverlight.

As to Silverlight and Windows Media Player, I didn't not say that Media Player was simply a front end for Media Player but nice try at twisting my words. Obviously with the much ballyhood addition of a subset of the .Net libraries Silverlight is more. How much more the market will decide. My comment was that the graphics (ie video) piece was based on Microsoft Media Player a product that never enjoyed commercial success.

Phil,

I'm not personally sponsored by Microsoft, although I doubt that's what you meant when you said I'm sponsored by them. One of the websites that I edit here at Ziff (the one I linked to, DevSource) is sponsored by Microsoft. But here at Ziff, the websites have several sponsors; go to eweek.com and you can see the different sponsors that our company has. And since I gave full contact information, unlike most of the people who comment, anyone can click on my name and see who I am and that I'm a Ziff employee, just as Joe is.

But I'm a bit surprised at how defensive you sound in your response. I wasn't trying to start an argument; rather, I was pointing out my perspective. I apologize if you took offense at my comments.

That said, it sounds like you're talking about the cost in development as opposed to the cost in purchasing applications that use silverlight. (I googled the exact phrase "business application buyers" that you used and it got a total of four hits, so I wasn't familiar with the term.) Development in general has costs whether you use Silverlight, Flash, ASP.NET, PHP, whatever, as you certainly know. Whether the cost is worth it is up to the company; they need to do a return-on-investment analysis to determine whether it would be worth it to develop in Silverlight. But will it be "multiples higher" as you say? Well, first one needs to determine what it's being compared against. Again, if you're using Visual Studio to develop WinForms and WebForms apps (i.e. .NET and ASP.NET), then you already have the tools, and from there it's a matter of getting the programmers up to speed. I'm not convinced, then, that it will be multiples higher, at least not without some solid numbers from such an analysis.

"Two reasons why Silverlight will be not rise beyond a Flash competitor."

Have you ever tried to develop a full blown application in Flash? That's not the right platform. Flex is a closer cousin to Silverlight. I've tried both Flex and Silverlight and I find the application model and code-behind model in Silverlight a lot cleaner.

Multiples higher in cost? How about not having to deal with cross browser rendering issues? I spend an overwhelming amount of time trying to get the same XHTML based web app looking the same between IE6, IE7, Firefox and Safari. How about truly stateful apps instead of the stateless pigs we are developing today? Wouldn't you like to build more scalable apps? Silverlight, in my opinion, is a happy medium between ease of deployment of web apps and ease of development of winforms. In fact, the time saved worrying about and developing in the traditional web-based model could be used to create additional features or focused on better activities like usability.

The following are the reasons I can see that would raise the cost of a custom application developed with Silverlight:

1) Developer learning curve (just like when you started to learn WinForms and ASP.NET)
2) Lack of a complete set of controls (it IS in BETA, and the third party control developers like ComponentOne and Telerik are already developing control suites)

After you get passed these, however, the opportunity to create an extremely flexible, extensible, reusable UI framework is realized. Many enterprises already create their own frameworks, so I see Silverlight being perceived as added value to these frameworks.

As for development costs related to tools: Notepad and the few minutes of bandwidth it takes to download the FREE Silverlight SDK are all you need - but Visual Studio provides so much productivity it quickly becomes a better business choice.

One more note - you don't need to roll your own framework to release an application - but if you discover common patterns between applications that you develop, it is better to encapsulate these patterns in a separate library. Silverlight makes it easier to develop your own custom UI library.

"Visual Studio provides so much productivity it quickly becomes a better business choice."

Yes, it does. It allows one to mash together a great looking demo or cool-looking production application very quickly.

But.... why oh why then, can I use GNU Emacs on Linux and produce much more, and with vastly better quality, than those who use Visual Studio? Granted, my applications are behind-the-scenes search engines and other server applications and don't involve direct GUI development. But even those who use Visual Studio to develop parts of the behind-the-scenes infrastructure are not as productive. And those who use Visual Studio to produce our GUI applications are consistently stuggling against slip after slip and never ending stream of bugs.

And no, I am not talking about quickly-implemented server applications that leak memory, coredump, or need to be periodically restarted. I am talking about server applications that you can walk up to and beat to a pulp with a crowbar, and they just look at you and smile and keep on working (in a manner of speaking, of course!), while the corresponding Visual Studio crowd struggles with bug after bug after bug.

And... why oh why, using that ancient primitive Emacs-based environment, do I get more done, get it done better, and still have most evenings and all weekends free for myself and my family. And the Visual Studio crowd gets less done, does it poorly, and works long hours and weekends and holidays. In fact, two of those "top" developers have 7 stents between them. What's up with that?

I guess that the points of my long-winded rambling are:

1. The quality of the user of the tool is much more important than the exact choice of tool.

2. C# and Visual Studio and Silverlight are compelling and seductive, but they are also akin to selling your soul to the devil. Yes, it feels great in the short run. But in the long run, idiots with great tools do NOT produce great results, and a pact with the devil always ends badly... eventually.

Working for a while in the same market, one becomes of two things: very happy, or a philosopher. Hmmmmm......

For one, the company I work for creates applications that run a multi-million dollar mortgage company (one that survived the market's current state). The non-gui based services that we build are complex, integrate many third party products, are really stable, and all coded in C#. My team gets called when there is a production emergency, and to be honest, I can't remember when I had to take a call last. I attribute much of this to both great developers and great testers. As mentioned in the previous post, the majority of bugs I get are usually cosmetic and are non-ASP.NET cross browser issues. The other ones are usually my mis-interpretation of the requirements. :-)

Second, you really can't compare data services or other services that do not have a UI with those that do have a UI. There are many different considerations when discussing development. I do know that I prefer C# to PHP, and now also Silverlight to ASP.NET. I also think the designer in Visual Studio is not optimal for production end-user applications, and as a result I don't use it. I do use VS's code editing and project management features. All this does not prevent me from creating production applications, nor does it make me take longer to develop it. Every tool has its shortcomings, and the smart developer will know how to work around them.

That being said...I think you hit something RIGHT ON THE HEAD:

"...idiots with great tools do NOT produce great results..."

It's not the tool in this case that makes bad applications. It's the uninformed person behind the keyboard.

"And those who use Visual Studio to produce our GUI applications are consistently stuggling against slip after slip and never ending stream of bugs."

It amazes me how often in Joe's blog I see people trying to make a point by relying on anecdotal evidence. Probably the worst I've ever seen for the slip-after-slip and ending-stream-of-bugs problem was a company that was producing software using J2EE and WebLogic. The thing was a disaster. Does that imply that J2EE and WebLogic are poor choices? Of course not. The problem was the quality of the work being done within the organization.

Anecdotal evidence doesn't go far. Chief Financial Officers want to see numbers, data, and statistics in making decisions on what to buy, and don't want to just rely on "I knew a guy who had a problem with..." statements.

The fact is, ASP.NET and SQL Server do just fine in a production environment when put together properly. Specific case studies include the ComputerJobs website, the USAToday website, Home Shopping Network, Buy dot com, Allstate insurance, CNET Channel, Carnival Cruise Lines. And yes, you could produce a similar, even larger list of Unix-based production systems. But my point is that a blanket statement claiming that VS-based apps are more bug-ridden and less reliable is simply without basis, just as my anecdotal story about the J2EE project I witnessed cannot represent all of the J2EE projects.

I have developed software for many years now and I have worked with several different tools and platforms.

The fact is that Silverlight is one of the most compelling new technologies we have seen in a while. To compare it to Flash is almost like comparing "apples and oranges" since it already does so many things that Flash cannot do well.

Focusing on streaming media and ads is missing the point. You cannot limit your thinking to how Flash is used today.

The niche that Silverlight fills is that you can actually build front-ends for solid line of business applications with it, which is something that is just not cost effective or practical to do with Flash or Flex. With a few lines of code you can create a superior user experience to traditional HTML/AJAX apps with much lower development and ownership costs.

Adobe has the market with Flash but they miss the point of where their technology may actually fit best. Also, their development tools are very weak compared to Microsoft, which means cost to build and own will almost always be higher. It is more cost effective to use C# with VS.

I know there are fanboys out there that love to hate MS and everything they produce, but you have to think objectively when it comes to business value. In this area Silverlight is already ahead and the product is only in beta. If Adobe does not change direction it is inevitable that they will lose massive market share here.

Silverlight ia a real RIA tool, Flash is only a animation-design tool, It's better for somebody who doesn't know how silverlight works to stop talking.

@otto kanellis:
Agreed. And it's also better for somebody who doesn't know exactly how cocaine chemically interacts with the human body to produce ruined lives and death to stop talking about why it shouldn't be sold to school children.