Silverlighters committed despite Microsoft's HTML5 love The "future of Silverlight starts now" — or, more accurately, it starts on December 2 according to Microsoft, with a day of webcasts from the company. Microsoft has promised a keynote from the chief geek synonymous with Silverlight, vice president for the .Net developer platform Scott Guthrie, who'll talk about what's in the …
It was funny to see all the Silverlight devs go mental about the fact that MS was mostly silent about Silverlight at PDC, whilst talking a lot about all their "HTML5" stuff. This is just a repeat the last couple of years, except then it was WPF getting "killed" in favour of Silverlight.
In actuality, of course, both WPF and Silverlight are doing fine. It's just that development and announcements scale back as the platform reaches maturity. Yes, Silverlight has a couple of features that I *really* wish were in WPF (as a full-time WPF developer), but in general it's a fantastic platform. If you look in the right places you can see that they *are* still adding features, they're just not great features to show off in a keynote (cross-thread collection binding! woo hoo!).
I'll be interested in hearing today's announcements, though. (And I do wish the WPF team wasn't *quite* as "dark" ATM).
Yeah I agree mostly with what you've said, Silverlight is doing fine and will continue to do so!
This quote from the article warrants clarification:
"Go back a month and it certainly sounded like Silverlight was being walked out on by its parent, Microsoft"
It was more a case of the press overreacting, and developers being led astray by misportrayal of what was actually happening with Silverlight (which turns out to be nothing but improvements - yay!).
...But everyone knows that, five years from now, the Silverlight apps will be those badly-neglected old dogs, that no one in the company wants to work on, but which have to be maintained because some small-time client in Idaho relies upon it. Getting all passive-aggressive about it, now, just means that your-particular-dog will end up sporting a load of bells and whistles that even Microsoft won't be properly supporting, by the time that happens. In fact, today's demonstrations will probably be a parade of all the super-whizz-bang things, that you absolutely should not implement, if you value your future.
"because running it is a one click "ordeal"..." Not if you haven't bought into the "Microsoft experience" it isn't. A sterling example of the crying need for open standards for the internet. Maybe if we can avoid an ISO ooxml ballot stuffing fiasco then a decent standard will evolve and hopefully do away with all the proprietary offerings. And yes, before you ask I do use Linux, not because I'm a fanatic but because it does what I want not what someone else thinks I want.
To make web apps, or any software for that matter, you need a programming language and a runtime framework.
So for web apps we have got:
- runtime framework: HTML5. Yeah, OK, not too bad, kind of works.
- programming language: Javascript. Hell NO on this one. Java script is a scripting language.
A programming language has at least strong typing and a compiler that makes sure that you cannot deply something with syntax errors in it. Java script has none of that. It is totally unsuitable for large projects. Maintaining or adding features to a Javascript app is a trip back to the 60'.
So what other options do we have for coding our web apps?
Java ? That would be nice. Java is a real programming language. Too bad Sun/Oracle totally screwed up their "applet" strategy so that nobody is using Java for this anymore.
Silverlight ? The underlying programming language is .NET, a real programming language so yeah, that works. It is only lacking adoption. And this is why I hope Microsoft succeeds.
I thought it was more the compiler that did the type checking, the language (certainly at runtime) would just go "WTF is this?" before vomiting a stack out to the console and going off in a huff.
JS may be far from perfect, but you can at least guarantee that the client can execute it. Not that I fancy doing much heavy-logic in JS (been there, got that t-shirt, it didn't fit)
This is not so for Silverlight, and until that is fully cross-platform it cannot succeed.
As for Java (and I am a Java dev) it never got anywhere on the client (due to it, basically, sucking donkey balls) and people need their heads examined if they use it on the web. It's found a niche in server apps and it does pretty good there; that's where it should stay. Use a known standard for comms to a client built in a more client-side friendly language and all is right with the world.
they wont slow down the development of it.
Not that anyone who actually knows how to program has the problems that clean_state alludes to for javascript. Syntax errors, strong typing etc are only a problem if your cavalier with your design, testing and deployment. You mentioned the 60's - when problems were addressed and solved rather than winged about.</bad workmen tools>
Actually Microsoft is doing fine with regards to Javascript these days. I'm not sure what inspiration they found emplaced in their NIH psyche, but they sure act enthused.
The new "use strict" will make many strong compiler users happy, and is being implemented now/soonish. The Javascript 'Harmony' effort has seemingly found rising air and all those features will make many more devs happy, though that's the future (maybe next year?).
And as for now, tests tests and more tests. It's not the language that's the problem, it's the commitment to error-free code. I'd not be comfortable working with anyone who 'relies' on the compiler for correctness. (and so agreeing wholeheartedly with "Tom 7" above)
Me thinks you have not been doing web development with non-MS tools if you think the choice is between Javascript and Silverlight.
For a start Google Web Toolkit (GWT) means you can write you code in Java and have it output HTML/Javascript on the client.
Silverlight & Flash are generally used embedded in HTML with lashing of Javascript as well.
I think even Apple & Google would advise you not to write complex applications in Javascript (though JS speed & tools have improved out of sight in the last few years), but as a scripting language it works well and runs everywhere. For the
Because if it does not run on Linux, it is a dead duck. Yes, I said "Linux". And no, I do not mean the desktop. The mobile market is stuffed with Linux (or Linux-like) devices, as well as set-up boxes. if you don't work on those (esp. set-tops moving forward) then you are boned.
It is available on a number of operating systems thanks to Mono and Moonlight. Roadmap for Moonlight is available at http://mono-project.com/MoonlightRoadmap - no doubt the first thing to mention is that Moonlight's main release is currently compatible with Silverlight 2. Moonlight 3 is almost finished, but compatibility with Silverlight 4 is still in the early stages.
I'm not sure what is going to happen to the Mono Project now that Novell has been bought. Discussion on the mailing lists suggests that the Mono developers will continue getting their paycheques and that licensing of Mono will not change. I guess we need to wait for further information from Novell or Microsoft about any changes the sale to Attachmate will have on the Mono Project.
...Silverlight is not cross-platform and cannot be run on Linux.
With all respect to Mono and Moonlight, they are either a few versions behind or can only offer a reduced subset of the functionality, so one cannot develop on .Net or Silverlight and guarantee with 100% certainty that it will execute on these frameworks - not with taking a lot of care over features and versions (or developing directly on those platforms).
It is true that Mono and Moonlight are lagging behind Silverlight, and that probably suits Microsoft down to the ground.
There are other problems with Moonlight and Mono though. The more apps that are developed using Mono and C# the more leverage is given to MS. RMS has written a piece on the dangers involved. See:
http://www.fsf.org/news/dont-depend-on-mono.
As I said in my previous post the sooner an open standard is reached the better.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
