John Liu .NET

View Original

Silverlight is for Business Apps, HTML5 is for Flash

 

Silverlight has always been about Business Apps since version 2.
HTML 5 - well I don't think this will ever be suitable for Business Apps.

 

Most of us that have been actively working with Silverlight has never had the misconception that Silverlight was about reach.  Yes Silverlight 1 was rushed out the door because we don't really want to learn Flash, the Olympics was happening, Microsoft saw an opportunity and pushed it - while the rest of the world was going Flash because of Video and YouTube.

Silverlight evolved quickly, because the community responded and Microsoft heard what we're asking: business apps.  Ever since Silverlight 2 was available we have had great tools, support for connecting to services and hundreds of controls for building business applications.  We started using it for intranet and even Internet business applications (as long as you only support Windows and Mac - it works just fine).

Sure, some "might" have hoped we'd see Silverlight on the iPhone one day - but let's be serious I really doubt that day would ever come.  There's this big fruity company in the way. 

 

While HTML5 is great for the mobile devices, I don't see corporate upgrading to the latest browsers for another 5 or more years (so many are still on IE6/XP).  By contrast, Silverlight runs on IE6.  This makes Silverlight still the best choice for the next decade, even before you consider the possibilities for hosted online services (SharePoint, Exchange or CRM), or Windows Phone 7.

I would love to see HTML5 be successful, but people who only think Silverlight is about video or the canvas element is really missing the mark.  Technology-wise, I think HTML5 is still a few years behind Silverlight.

 

So I did some napkin Maths.

  • Say, it will take HTML5 another 2 years for the standard to mature.  (2012)
  • MS tools and JavaScript frameworks (JQuery-Mobile) should be ready pretty quickly.  (2014)
  • Based on current corporate adoption rates - let's add another 5 years to upgrade to IE9 (or later).  (2017)
  • Newest Mobile devices should move quicker.  (2013)
  • Except old devices will never work:  old iPhone, blackberry, windowsphone 6.5

 

  • So, 2014 or 2017 to be able to use HTML5 for less features than Silverlight 2.  Great. Fantastic.

 

Let's consider the scenarios to build applications:

Internet - any browser

  • HTML4 - best reach - shit "rich UI"
  • Flash - no mobile
  • Silverlight - less reach than Flash, works on most Windows and Mac, no mobile
  • HTML5 - adoption too low

Internet - mobile browser

  • HTML4 - best reach
  • Flash - bad bad
  • Silverlight - bad - Windows Mobile, Symbian (support woeful)
  • HTML5 - only for latest/future mobile devices

Intranet - MS environment

  • HTML4 - best reach, not rich
  • Flash - so so, but in a MS environment lack skillset to develop Flash/Flex
  • Silverlight - great, easy to deploy, installation independent of IE version
  • HTML5 - 2017

Intranet - MS + Mac

  • HTML4 - best reach, not rich
  • Flash - so so, but in a MS environment lack skillset to develop Flash/Flex
  • Silverlight - great, easy to deploy, installation independent of IE version
  • HTML5 - 2017

 

I wonder about the whole "HTML5 standards" - the IT industry is littered with corpses of different vendors coming together to put together different implementation of the same spec

  1. Java (write once debug everywhere)
  2. XHTML 1.1
  3. HTML 4
  4. Open GL
  5. SVG

At the very end - you can't trust ANY implementation - you end up relying on a framework to smooth out the differences and give you the lowest common denominator:

  1. Java Swing
  2. JQuery

 

The conclusion I came to is not pretty for HTML 5:

 

Microsoft-Corporate environment / Business Apps:

  • Silverlight (rich)
  • HTML4

Mobile Devices (latest / future)

  • HTML5 (reach) - I wanted to recommend HTML5, but really, the more I think about this, do I really want to use a web app on my WP7 or iPhone?  Really?  That would be pretty silly.  We're talking about rich interfaces here, may be in another 2 years when tools are better at faking it.  Does HTML5 even have multi-touch, GPS or camera support?  Do I have to code gesture-recognition in JavaScript?
  • Build 3 versions for (rich)
    • Java - Android
    • Silverlight - WP7
    • Objective-C - iPhone

Public Internet

  • HTML5
  • HTML4 (degrade for unsupported browsers - lose rich retain reach)
  • Build two versions
    • HTML5 optimized for mobile browsers
    • HTML4 for unsupported desktop browsers

 

UPDATE:
Intranet Application where Reach is important on new devices

  • Need to support Android, iPhone, Windows Phone 7 and iPad
  • Reach is paramount, and development budget does not allow 3 individual versions
  • Go for HTML 5, with HTML 4 graceful degradation
  • For this scenario, I would argue that you build the HTML4 experience right now, and wait to build the HTML 5 experience later - the later you can wait, the more mature the tools and browser gets.

 

Please let me know what you think - would love to be corrected.