Well for my point of view this is given. When developing .NET applications it has always annoyed me that browsers interpret the common “standard” of HTML very uniquely. Adding AJAX into this equation only make matters worse.
Using Flash you can almost guarantee that the design is showed correctly on any browser but you need to make an interface for the server code to deliver data to the Flash component. Furthermore you have to make this interface secure – which can be quite difficult or even impossible. The same issues apply with the use of AJAX.
What Silverlight offers is a secure, sandboxed exchange between server side and client side code, making Silverlight a kind of fusion between traditional .NET and Flash/AJAX.
It support all the features of .NET meaning that the applications can be coded using the powerful C# which I prefer over any other language. And the interfaces are supposed to very secure even though they are based on existing technology like webservices.
And another great aspect of Silverlight is that you can add the client side code to your server side code in one project making them tied strongly together and making it far easier to make applications the correct way. When programming using the 3-tier model – which is HIGHLY recommended – you get all your layers included in one project with all the possibilities of .NET like inheritance of common error handling etc.
In my opinion Silverlight will make the development of AJAX web applications reach a more mature and fare more professional level and this is needed to make more serious web applications move into a richer client side. Imagine your net banking relying heavy on Flash or AJAX at their current state – not something that will happen likely.
Besides there is a rumor that around summertime the total lack of security in most AJAX sites will be the hot topic. This will put the whole AJAX development in a bad light and the demand for a much more secure way of making rich client side code will be high. Enter Silverlight – Get on the wagon as soon as possible because this is going to be the next big thing.
Hi Kasper, nice of you to drop by (again), you must be loving the new design :)
Again, I still think it depends on the context of things. Are you building a functional web application that requires desktop-like behaviour, or are you simply displaying (and linking) information of text! A HTML-centric solution resolves many of the design issues that otherwise needs to be defined explicitly. However, I agree its a difficult platform to work on (as with many none-proprietary frameworks), but it is getting better! A multitude of JavaScript libraries have seen the light lately, streamlining development across layout engines. The same goes for markup and stylesheets, a set of CSS-Reset definitions and a handful of guiding rules almost make web development consistent.
I believe that the main reason a plug-in module (as Silverlight is) hasn't taken presence over HTML is 1) because hypertext and browser navigation is an ingenious solution to most challenges, 2) because plug-ins haven't been implemented into the ecosystem of the web, specifically being searchable and page-centric.
The lack of security concerning AJAX, I dont see it as a big issue, for the most part (of what I have seen) AJAX is primarily used in none-critical security environments. Are you specifically thinking about POST form usage?
I used to bet on Flash. Then I finally succumbed to the far superior HTML.
This time, I'm not so sure.
Webapps on steroids seem to tickle me just right. Especially those that are so complex that they can compete with "real" apps. You know I dumped Thunderbird in favor of Gmail, so web-apps _can_ win.
Will I dump OpenOffice for Google Docs?
We'll see. Right now my money is on Adobe winning the first round. The second round will be decided by hardware acceleration.
Ecosystems aside, I believe this is a fight for developers, a fight for the best API and development tools. With that regard, lately I have been working with the Prototype and Script.aculo.us javascript libraries, in combination with ASP.NET, and I must say, its not the most efficient way of production. It's my upcoming Firefox questionnaire survey and new questions emerge as the first set have been answered. There is a lot of conditional show/hide/disable/enable and slide-up and -down. I was kinda missing the timeline as found in Flash. I think the Sliverlight solution of combining the two approaches (programatical and timeline based animation) is spot on. Another issue with Ajax (if not using Ajax.Net) is the extra layer of abstraction, I need to serialize and deserialize objects, check types, and setup validation on both client and serverside. And for what? Clean and structured JavaScript and HTML code :)
You forget: Adobe Apollo is not just Flash (or Flex). It's also HTML, CSS and AJAX.
In other words, Apollo is a browser :) Or is it an object within the browser that display HTML! That just doesn't make sense.
I am not sure that I agree with you guys. I think that both FLASH and SILVERLIGHT are obsolete. The future of dynamic web presentations lies somewhere else. I've stumble across this technology: www.qvikbutik.dk/.../index.html - that impress me far more than FLASH and SILVERLIGHT...
I kinda get the feeling that you are affiliated with that enterprise in some way Kristian? LOL I see your point, SilverFlashLight is the future!
Yeah he seems kind of biased towards this new revolution technology ;-)
Hehe, forget all I have stated earlier. SilverFlashLight is the obvious horse to put all your money on.