Microsoft and HTML 5
8 views
Skip to first unread message
Ken Smith
Jan 30, 2012, 2:28:26 PM1/30/12
to Dot Net Startup
My current startup (http://alanta.com/) is heavily focused on
Silverlight on the front-end. This is looking to have been a mistaken
platform choice, though we're too far into it to change easily. And
for what we're doing, HTML 5 just isn't ready yet, i.e., Google's work
on WebRTC hasn't yet been widely adopted (see
http://www.tomshardware.com/news/Google-Chrome-WebRTC-DEv-Channel-HTML5,14526.html).
But when the time comes, assuming we make it that far, we're almost
certainly going to want to rework our client in HTML5.
Which brings me to my question. It goes without saying, I suppose,
that C# as a language is immeasurably superior to the ugly, obnoxious
hack that is JavaScript. But if circumstances are gonna force me to
leave C# behind, and if I'm gonna be diving into extensive HTML 5/
JavaScript/AJAX development, why should I want to use the MS platforms
and/or tools? I've grown up in the MS ecosystem, so Visual Studio
feels very natural to me (unlike LAMP, Python/django, RoR, etc.). But
apart from familiarity, what advantages do I get by sending my money
to MS for doing HTML 5 development in Visual Studio? Does anybody have
any insight, or even just speculation, about how MS is planning to
retain the loyalty of their developers when they've ceded control of
the underlying platform? Or has MS just found themselves painted into
a corner that they have no way out of?
Silverlight on the front-end. This is looking to have been a mistaken
platform choice, though we're too far into it to change easily. And
for what we're doing, HTML 5 just isn't ready yet, i.e., Google's work
on WebRTC hasn't yet been widely adopted (see
http://www.tomshardware.com/news/Google-Chrome-WebRTC-DEv-Channel-HTML5,14526.html).
But when the time comes, assuming we make it that far, we're almost
certainly going to want to rework our client in HTML5.
Which brings me to my question. It goes without saying, I suppose,
that C# as a language is immeasurably superior to the ugly, obnoxious
hack that is JavaScript. But if circumstances are gonna force me to
leave C# behind, and if I'm gonna be diving into extensive HTML 5/
JavaScript/AJAX development, why should I want to use the MS platforms
and/or tools? I've grown up in the MS ecosystem, so Visual Studio
feels very natural to me (unlike LAMP, Python/django, RoR, etc.). But
apart from familiarity, what advantages do I get by sending my money
to MS for doing HTML 5 development in Visual Studio? Does anybody have
any insight, or even just speculation, about how MS is planning to
retain the loyalty of their developers when they've ceded control of
the underlying platform? Or has MS just found themselves painted into
a corner that they have no way out of?
Marcelo Calbucci
Jan 30, 2012, 3:52:26 PM1/30/12
to dotnet...@googlegroups.com
Ken, I've been using JavaScript on the browser and .NET on the server for the last 8 years. It took me a while to accept JavaScript as a language, but with libraries like jQuery and others, it's a lot faster to develop rich HTML applications. My understanding is that Microsoft is investing heavily on Browser support for HTML5. They already have some great support for jQuery (with intellisense and all). I believe Adobe is also investing into HTML 5 "authoring" tools (as a way to target non-techies who were building Flash sites before).
-Marcelo
Ken Smith
Jan 30, 2012, 4:55:46 PM1/30/12
to dotnet...@googlegroups.com
I largely agree with you, Marcelo. jQuery is pretty good: it's almost good enough to make JavaScript reasonable. It's getting a lot faster to develop rich HTML apps. And all the resources that MS had been investing in Silverlight (and more) are clearly going to JavaScript/HTML5.
But I still don't like the situation.
jQuery is ultimately lipstick on a pig. The underlying language is a mess. Critical constructs like namespaces, inheritance, and data-hiding, that should be dead-simple 30+ years after they became commonplace, are complex, ugly and hacky, at best. Its equality rules are mind-bogglingly complex and completely counter-intuitive. The fact that you can do anything you want to any object means that static code analysis is all but impossible, and if there's any decent dynamic code analysis tools, I haven't seen them. Refactoring tools can basically be reduced to search-and-replace. Coffeescript and Dart seem to me like obvious directions to take, but until the IDE's and debuggers have better support for them, I can't quite see my way to using them to build a real system (maybe in 2-5 years, if they can get any traction).
OK, I get it, lots of folks manage to work around these limitations, and there are very good reasons, like platform independence, that make it worthwhile to do so. It seems like MS had been working hard to give me a reasonable alternative in Silverlight, but that's apparently not gonna happen, so now I have to do what everyone else does, and muddle along. I can live with that.
But that still raises the obvious question: why should I do it on the MS stack? Yeah, Visual Studio is awesome, but not when it comes to JS (and I haven't seen anything that leads me to believe the situation will be dramatically different with VS11). ASP.NET MVC is a nice step forward, but as more and more logic moves to the client, its benefits are less compelling - besides the fact that it's mostly playing catch-up at this point to other web frameworks, like Rails and Django. Alanta's client has roughly 3x as many lines of client-side code as sever code (30K vs. 10K), so it's the ability to solve client-side issues that's the most compelling. And I haven't seen MS offering anything that makes me think, "Oh, wow, I really need to stick with Visual Studio for my next project." I'm not saying I'm planning to switch - but for the first time in my career, I'm asking myself really hard whether I want to stay. And that ought to get MS very, very nervous.
Reply all
Reply to author
Forward
0 new messages