Developers say Microsoft has "betrayed" them by changing Silverlight strategy | Hacker News Reader

Developers say Microsoft has "betrayed" them by changing Silverlight strategy

bbc.co.uk

16 pointsauxbuss30 comments

I've been working on a Silverlight app full time for the past 8 months. I was relieved when Microsoft made this announcement (even if they are now trying to back track on it a bit).

Silverlight isn't bad at all, really. In some ways it's just fantastic. And given another 3 or 4 years or so, it will mature to the point of being a great framework. It's a truly awesome framework for WP7 and if MS can gain some traction behind WP7, having Silverlight in their corner is going to be a boon for them.

But right now Silverlight is both mighty raw and almost pointless for webapps. There are so many bugs in Silverlight itself and its related tools that it can make working in Silverlight quite frustrating. When all this time I watch HTML5 blossom and take off and pretty much gain feature parity (and beyond) to Silverlight, I can't help but think to myself as I look at my app, "What's the point?"

Nothing is perfect, and of course HTML5 is going to have its warts, especially when dealing with every vendor's own implementation of JS and the new HTML5 components. But at least with this route you're swimming with the current and not against it. There's a lot to be said for that.

> But at least with this route you're swimming with the current and not against it. There's a lot to be said for that.

Exactly. It's disappointing how much time and effort has been squandered on various abstractions primarily popularized because they gave developers excuses to avoid learning HTML, CSS, and JavaScript.

When it comes to Silverlight, it's easy to see how so many devs were tempted to unnecessarily overuse it. Microsoft's PR efforts are often panned, but the Silverlight hype machine was incredibly persuasive. I even questioned my own focus on standards-based development[1] for a few brief moments during conferences like MIX.

[1] http://encosia.com/2009/09/14/is-silverlight-the-new-webform...

I thought this was a refreshing move. They've accepted the situation and adjusted their strategy accordingly. The only other option seems to be ignoring reality and continuing on their old course, which won't do anything to improve the success of Silverlight.

Even Adobe seems to have given up the fight and accepted html5 as the winner (given their demo's of html5 authoring tools recently).

I do wish MS would release a SL to HTML5 compiler though. That would mitigate a lot of the rancor.

That would be a hard cookie. In many ways, Silverlight has a lot more features than HTML5 and it is also compiled to binary. I guess it would be possible to compile the binary to some sort of JavaScript and emulate some of the Silverlight API, but I seriously doubt that it would yield very convincing results.

I agree, but in my mind it doesn't need to be perfect. Just good enough to get the bones to dance a little, and frankly, there isn't a lot you can do in SL that you can't do in javascript / HTML.

I could be wrong, but I don't think translating between Silverlight and actual web development with an acceptable level of fidelity is realistic. If it were, ASP.NET WebForms 3.5 and 4.0 would have implemented something like that to begin with and Silverlight would have never existed.

I can't see how this would work or be effective. It could possibly work for the smaller, more media oriented Silverlight apps. But those apps are relatively painless to port to HTML5 already. For larger, line of business apps, especially those that frequently talk to a server, I can't even imagine where a tool like this would begin.

Actually, I think Microsoft is doing developers a lot of good by keeping Silverlight in use in other areas, for Windows Phone development, for example. The technology is still useful, and they can transfer their skills to this new area.

So, rather than just dumping Silverlight, they are still using it in other key areas, areas they are looking to help developers make money with. Don't tell me Flash developers wouldn't love to have a mobile platform where they could use their skills to develop applications on.

Not exactly the pinnacle of BBC reporting there. No attribution, link or otherwise for the assertion that "Developers" say they've been "betrayed". Who are they, how many of them are there, how much actual investment have they made and how do they expect to lose out? Not saying the story's necessarily wrong or untruthful, just that as presented so far it can't be considered anything more than gossip.

The article talks about "many of the comments on Mr Muglia's blogpost", and links to that blog post a bit further up: http://team.silverlight.net/announcement/pdc-and-silverlight...

I would also like their phone number, home address and working place.

The gradual shift in strategy has actually been evident for a while. With Silverlight 3, they introduced the ability to run outside the browser as a desktop app, and with Silverlight 4, they placed a lot more emphasis on this scenario and less on in-browser scenarios.

The point of writing a Windows desktop app in Silverlight is that not only is your code more portable (relative to a native Win32 or full .NET/WPF app), but SL apps are like iOS apps in that they are isolated from each other, easy to update/service, etc. If Microsoft introduces a Windows app store soon as is strongly rumored, it will probably feature prominently.

I think what has most developers upset is not so much the substance of the message, which most of them already knew, but how it's being picked up in the media (for which Microsoft's delivery of the message is to blame). The media don't really care about development frameworks, so to them Silverlight-the-browser-plugin (the part they see) = Silverlight. Therefore if Silverlight is being de-emphasized as a browser plugin, it means Silverlight is dead, so that's going to be the headline, and CIOs etc. who only read headlines are going to see that and decide not to allow Silverlight development, even out-of-browser.

As a developer in the Microsoft stack, this is par for the course. Choose your tech wisely.

Sorry, I'm going to play the "greater good" card here. Yes, it's a shame if/when your project gets canned. The important thing here is that the masses will not be stuck with proprietary crap.

I'm going to play the progress card here.

The web just jumped back a decade to crappy old html. XAML is infinitely superior. You can't even extend HTML server-side without writing code :( It's got loads of odd rules. It still thinks everything is an overflowable document when I want to write an application.

The web just jumped back to a kooky little language that's not grown for decades. Javashit.

Silverlight had support for static, dynamic and even functional languages! C#, VB.Net, IronPython, F#. And it's actually fast and compilable unlike javaslow.

(p.s. I don't actually mind javascript that much, but I do think it's pretty poor compared to C#, Python, Ruby, etc. I do actually detest html though.)

A lot of us don't need cross device compatibility. In fact cross device compatibility is pure laziness in many instances as a iPhone optimized web app should look very different to a laptop/desktop browser optimized app.

I'm not really sure where I stand in all this, but I'm not dancing around with joy about HTML5, which I think is pretty basic and ridiculously behind the times. These are features they should have introduced a decade ago.

I also really wish they'd make some way I could use other programming languages in the browsers.

I'm not sure what you're complaining about. You can still use SL, it's not going away, and it will continue to get new releases. If it meets your need, I don't see any reason why you wouldn't continue to use it. It won't work on every platform, but you say that's not a problem for you either. Go write some XAML. Personally, I find that it really is awesome for Windows Phone development.

As far as HTML / javascript goes, I'm beginning to get the impression that you don't like it. I understand, I actually hated touching javascript until I fell into JQuery. Even with it, I think developing javascript feels half-assed. But that's what the platform uses and modern browsers are generally fast enough. I tend to think of HTML as basically a display protocol though. It's ugly, but it works.

Take heart though, you can write the server side in whatever language you want, and your server-side can write your javascript for you.

I'm tired of the 'woo html5' and 'hate flash/silverlight' camps.

It feels as if the same people who talk about beating 'proprietary' with an open web, open standards and open source code seem to miss we have _no_ choice when it comes to web programming, it's html/javascript or bust.

And javascript is not fast enough. Not even vaguely close to being fast enough. You must have seen the semi-regular 'look what I did with HTML5' demos, yeah great, until you put them next to a flash or silverlight demo. Then suddenly it looks pathetic. It's not up to the job.

This is a step back, not a step forward. It's not for the 'greater good', html needs decent competition to keep it modern. Competition is healthy!

While it would never happen and will never happen, Microsoft releasing Silverlight as an update to Windows could have given the HTML standards the kick up the ass they need to start taking application development seriously.

For most of us the browser is the new defacto OS.

We're a community of entrepreneurs and developers for pete's sake. And yet we're cheering on a virtual quango creating HTML5.

Relying on committees to drive forward the web is absurd. When did meetings and committees ever actually solve a problem or produce amazing innovation?

Developers complaining that a company with a history of treating people poorly betrays them? Cry me a river.

I see where you're coming from, but in regards to MS and developers I really can't disagree with you more. MS, more than most, bends over backwards to reach out to developers. You can still target COM and OLE, you can still write VB4. I have applications written for DOS and Windows 3.1 that still work on Windows 7.

In this case, I think it is more a failure in PR than anything else. Silverlight will continue to work on Windows and Mac, and it will be the primary way of writing apps for the Windows phone. MS has just publicly acknowledged that it's stupid to try to get the Silverlight VM to work on every stack of hardware / OS. They just did it in a very ham-fisted fashion.

That wasn't really news to anyone that's been paying attention anyway. We've been warning our developers away from writing any web apps with Flash or SL (except for video playback) for over a year now. You can write very capable, very pretty web apps using just javascript, CSS and HTML4. HTML5 just makes it that much easier.

Honestly I think Microsoft's extreme desire to maintain backwards compatibility has hurt developers in the long run. That's an awful lot of legacy to carry around and they haven't been proficient at phasing products and platforms out when they're clearly way beyond their prime. They should focus more on being innovative and having solid strategies for moving developers incrementally towards new platforms and frameworks.

One word: Enterprise.

I currently contract for a company with 50,000 + employees and some pretty specialized software that way too many people need to access. Getting that to change is like turning a convoy of sea liners with no radio communication.

The programmers who actually understand the weird ass domain of that software really wish they were still writing COBOL on a mainframe, the users who know how to use the software are understandably reluctant to have to learn another application when the existing works fine for THEIR needs (and job security). We have succeeded in getting a lot of stuff transferred to Citrix servers, but there are still a few hold outs / problem areas.

So, yeah, sometimes we really need the existing old stuff to still work.

> it will be the primary way of writing apps for the Windows phone.

This, right there. Silverlight isn't going away. It's still useful. They just aren't trying to compete in an area where it won't be as useful. However, Silverlight developers suddenly have this transferable skill that Flash developers don't have on the mobile side.

Yes, Silverlight will be used for WP7. But I doubt MS has any commitment to keep it for future releases of WP7.

Considering Silverlight and WPF also go together, I don't see Silverlight going away anytime soon. I see it changing direction, or at least, not being limited to just the web.

"You can write very capable, very pretty web apps using just javascript, CSS and HTML4. HTML5 just makes it that much easier."

Which part of HTML5 are you talking about here? I always assumed that in this context HTML5 meant 'Canvas + <video> element'. I hope that developers aren't going to use Canvas to replace CSS/javascript menus, rounded corners etc. with Canvas-based ones 8(

Canvas makes some of the shiny things easier to implement. And shiny animations (that don't impede usability) make the customers want more.

Sure, just like shiny Flash components that don't impede usability make the customers want more. I'm not against HTML5, and an open standard is better than Flash, I just hope people won't use 'it's the standard' as an excuse to throw together complete websites on a Canvas and break copy and paste, scrolling, bookmarkibility etc., all those things that suck so much about Flash websites.

Maybe I just don't know enough about Canvas and will it play nicer with the html/css separation model than Flash does, but from what I've seen so far I'm not getting my hopes up.

It seems like just yesterday that developers were saying Microsoft betrayed them by breaking compatibility between VB6 and VB.Net.

I think Silverlight as a WP7 app framework was the right thing to do. The "R&D" team at my company has done alot of silverlight web apps, and most of them have resulting in very pretty but completely unintuitive messes.

Reply on news.ycombinator.com