How much should we worry about browser pentration when using Silverlight?

Question

I mean,, that would be the only issue to worry about when using silverlight compared to flash... do you think this issue could be solved in 2-3 years?

Answer 1

I'd say stick to the current norms.

Flash is very widespread, so if you want to make a public website that everyone should be able to enjoy, this is your best option.

Silverlight may not be as widespread as Flash, but in my opinion, it's much better for the development of actual applications, so if I were developing an application for working purposes, I'd choose silverlight.

As to where Silverlight will be compared to Flash in 2-3 years, that is anyone's guess.

Answer 2

It really depends, for anything that involves user acquisition I would steer clear of it unless your trying to do something very interactive and different. The install silverlight wall would lose you a lot of users.

On the other hand if you are developing an application for a known group such as a internal type app of rolling out something new to an established community then I think it is a great choice.

Most tech savvy people should have silverlight these days, it's just the average person who may still be using IE 6/7 that will have no idea about it.

Answer 3

Short answer: No, Silverlight adoption numbers should not stop you from developing any application with it.

Longer answer: No, it's a non-issue for most apps today and will become an exitent issue soon.

Fair disclosure:
This post is being written by former Microsoft Silverlight Program Manager - Justin Angel. My personal conviction in Silverlight was so great I left my entire life, family, friends and career behind and came to work for Microsoft on Silverlight.
So yeah, I'm biased. Given my background, I'll stick to a simple & basic analysis of where we are and back it up with facts.

Additional fair disclosure, this post will only exclusively employ knowledge already in the public domain, and is not the full extent of my professional knowledge.

Much longer answer starts here.

The numbers

1. Fact: Silverlight global adoption is currently at 45% of all internet connected machines, which is more then 500+ million people. (Statistic by Scott Guthrie at PDC09)

   1.1. Fact: 500M users is more people then the total of users who have Twitter, FaceBook, Myspace or Gmail combined. A lot more. a lot a lot more. (Statistic via Scott Burnes, former Silverlight Product Manager's, Twitter account)

2. Fact: In Mix09 Silverlight had 33% global adoption. (Mentioned is the Mix09 keynote)

3. Simple math: Mix09 was took place on March 2009 and PDC09 was on November 2009.
   12% growth / over 6 month period = 2% growth a month. Which is 1% acceleration from the previous growth/per-month period (Mix08->Mix09).

4. Conclusion: Following that math, Silverlight will hit near ubiquity in the next 18 months. I won't go into whether this is a log graph or a linear graph.

5. Where's ASP.Net in all of this? Well, here's another interesting graph for you: http://www.twitpic.com/lvv9l/full

Do the numbers really matter?

In my opinion, No. Enterprise, Corporate and ISV level customers with both internal and external facing apps are already hard at work or have shipped Silverlight apps.
Any end-users who doesn't already have Silverlight (the other 50% of users) can install it in less then 2 clicks and 10 seconds.

How many users have already successfully installed Silverlight in exactly that way? Half of all internet users. While your app might be unique, half of the internet is a pretty good sampling group.

Users don't mind installing an additional plugin for an application they'd like to use. The numbers don't lie.

Summary
Silverlight adoption numbers don't matter when choosing a development platform. That's due to the quick and painless install experience for Silverlight, The enormous number of users who've previously independently installed Silverlight, and the reasonable forecast for Silverlight ubiquity.

Caveat
That being said, There's one scenario where non-ubiquity penetration numbers can hurt user experience. Mainly, if your consumer base don't know their your customers. In "secondary markets" (like online ads integrated into pages), there's some truth that you'll limit your customer base since most people will not install a browser plugin for an ad.

Cheers,
-- Justin