I could of given Damian the usual workflow, video codec's etc response that I'm sure many have heard umpteen times before. Instead I just deferred my answer and thought it over some more as I really hadn’t thought of the two being better than the other. In that, as a Product Manager for Silverlight to create a debate between these two approaches to the web is what we’d commonly label “breaking into jail”.
The simple answer above all is, that you’re working in a controlled environment. That’s why Silverlight could be a better fit and before you raise your jQuery swords and lead the charge against the arrogance of the few that which is plug-in’s let me ask you to pause ever so briefly and hear me out.
On top of this, you have the browser related issues where if folks don’t do their homework with regards to CSS they can in turn cause a number of visual issues, resulting in a degraded experience (something Damian is quite vocal about and so he should be and agree with his arguments).
Having absolute universal control over the XAML (DOM) and absolute pixel precession is definitely favored via plug-in, that being said it can still also be achieved with CSS / HTML (provided you’re skill set is capable of adhering to this).
To do this outside plug-in state, still requires a lot of trial and error when it comes to CSS production (jQuery though is technically brilliant at reducing some of this pain), as you’ll be adjusting everything from margins to padding along with injecting various DIV/SPAN tags to represent containers or inline elements that produce the visual result you’re after.
I still think however we’re clinging onto the few scraps of HTML we have today, and have noticed over the years many folks creating rules of engagement with these few tags – as if they were given to us as by our forefathers as being all that is great.
I simply think we’re holding onto them for the previously mentioned point, in that they are simply the first and probably the last agreed standard we’ve had all collectively blessed. I honestly don’t think when HTML was first drafted as a spec that we’d be seeing concepts like Virtual Earth or <insert your favorite AJAX / AJAJ > being used in the form it is today.
That comment was sure to be a redflag, so let me explain in detail what i mean “Holding onto the scraps of HTML we have today”.
My theory is that HTML as we know it today, seems to be lacking in depth (probably why there is so much conflict with the language). I’ve seen over the past 10 years many arguments arise over the correct use of this blessed technology (I remember the famous <B> tag vs <Strong> argument. I also saw the DIV vs. LAYER tag war arise…) I’ve witness HTML’s birth from beginning to now, and I’m still yet to see anyone camp win over 2/3rds of the other.
I vote for hybrid approaches for now, let’s just use both and evolve like we have been given today and explore in more depth what works vs.. what doesn’t.
In the end it’s all semantics and “it depends”.
You tell me what’s better and then tell me how Silverlight can improve. Please don’t beat me up, this ones just me thinking out loud after using both approaches since 1995.
Note: I’m not a Product Manager of Internet Explorer and my influence there is probably less than yours all said and done.