XAML and DX Interop

[John Gossman]

http://blogs.msdn.com/b/windowsappdev/archive/2012/03/15/combining-xaml-and-directx.aspx

[Jeremiah Morrill]

Thanks for the link John!

This feature is really going to make XAML (and DX) applications shine.

-Jer

On Thu, Mar 15, 2012 at 12:35 PM, John Gossman <gossm...@gmail.com> wrote:

http://blogs.msdn.com/b/windowsappdev/archive/2012/03/15/combining-xaml-and-directx.aspx

--
Microsoft MVP - Client Application Development

HJT, Inc Software Developer

[Jeremy Alles]

Yes this is a nice feature. 

It looks like the team working on WinRT has learned a lot from WPF ;-)

Jeremy

[Peter O'Hanlon]

Now if only this could make it into the desktop experience. I'm sorry Microsoft, but I can't help thinking that you're leaving those of us on the desktop side out as third class citizens here. You see, things like this excite me (not enough for me to go all scary stalker, but enough to give me shivers) and then we find out that it's going to be denied to those of us who have to write and support desktop applications.

How can something be such an epic success and such an epic fail simultaneously.

--
Peter O'Hanlon

[Colin Blair]

Have you seen the video of the guy trying to use Windows 8 for the first time? Not sure if the video proves that the desktop mode should be improved or just done away with.

 

http://www.youtube.com/watch?v=v4boTbv9_nU

[Peter O'Hanlon]

Rightly or wrongly, the message that's now out with the tech writers is that Metro is a disaster on the big screen, and that MS is chasing after the wrong hare here. Honestly, if contracts made it into the desktop, I would be happy. If desktop could be a first class citizen, I would be delighted. As it stands, several of my clients have flat out stated that they will not upgrade - they have too much investment in desktop based multi-screen systems, and Metro gets in the way.

 

Sadly, I can't argue with them. This is the first time I haven't been excited by the prospect of a Windows release. It's a very sad day for me.

--
Peter O'Hanlon

[Colin Eberhardt]

I totally understand your perspective, I can't see win 8 making a big impact in the financial domain either for those self same reasons. Which is a shame, because I do think win 8 will make a fantastic tablet OS, in the same way that WP7 is a great phone OS.

The problem is, how to create a completely new experience, whilst maintaining the old. I totally understand why they split win 8 down the middle.

So, Pete, are you just disappointed that win 8 desktop does not really add anything to win 7? Or do you think a unification could have been possible with more thought / work?

---

From: Peter O'Hanlon
Sent: 15/03/2012 22:15

[Peter O'Hanlon]

I see nothing that we can class as a desktop improvement. Metro actually gets in the way of the desktop, rather than enhances it.

If we were to have a unification of features such as WinRT, XAML as a first class citizen, the feature posted here, charms, etc, then MS would have a surefire hit in the corporate world. I have just had a client ring me to tell me not to offer them a Windows 8 update to our software because they will not upgrade. Now, these people throw money around on just about any software upgrade because they do this to protect their budget. For them to say no to this is unprecedented. Heck, they even upgraded to Vista when that first came out.

I don't know how clear we can make this, but there are lots of companies out there who have applications that cannot be offered in a Metro way, so not offering access to the same features and not listening to people who really do need to treat the desktop as their primary system is just alienating us. I have tried to love Metro, but I just can't. Embracing it would be financial suicide for us. The single screen app does not work for our clients, this being one of the reasons we never went the ipad route.

[Jeremiah Morrill]

There is some value add to Windows 8 for desktop UI applications (DirectComposition comes to mind, D2D additions/optimizations), but for me, the most compelling technologies are Metro-only.  WinJS and Win8 XAML seem to be the first Microsoft UI technologies that don't need a 12 core nuclear reactor and the first time ever MS had C++ as a first class citizen for rich UI.

I don't mind as much that these technologies are Metro-only, but the restrictions in API usage they bring.  This is a big plus for consumer, but I was hoping for an enterprise option that would allow free-reign over many of no-no API calls.  Things like access to custom hw/drivers, IPC and using a wealth of code that depends on non-safe metro APIs are pain points for custom enterprise software (software that is already "trusted" within a company) that want to move to a Metro API.  Until something like this happens, or these technologies get official presence on the desktop...It would be safe to call the traditional desktop moving towards obsolescence. 

Even though I have my complaints, I'm pretty excited for this new technology and starting to see a lot them learn from a lot of mistakes of the past.  I agree with Colin that it will make a fantastic tablet OS...not sure about comparing it to WP7, which I hold in pretty low regard. :)

I've seen the video of the old guy running windows 8...But I'm sure there was a lot of similar reactions to Windows 95 from Win 3.1 users.

-Jer

[Laurent Bugnion]

The error that many people do is to consider Metro applications as an *upgrade path* from classic desktop apps. This is not that at all. Just like a phone application is not an upgrade to a desktop app.

 

In fact Metro apps are companions to full blown desktop apps, with less features and touch oriented. The firms who don’t get that are doing it wrong. Metro apps should be developed *in addition* to standard applications.

 

There is no doubt that in some cases we will see Metro-only apps (just like we have some phone-only apps), but if the client has a running application on the desktop already, proposing an upgrade to Metro doesn’t make sense.

 

All this IMHO of course.

Laurent

[Peter O'Hanlon]

Which begs the question, why does metro get in the way of the desktop? What I keep getting asked is why can I not have the desktop as my primary interface?

---

From: Laurent Bugnion
Sent: 16/03/2012 12:29
To: wpf-di...@googlegroups.com; Colin Eberhardt

[Colin Eberhardt]

I totally agree, there are a great many business apps for which metro is not a viable upgrade path. Which is the reason why I strongly disagree with the statement Jeremiah made ...

"It would be safe to call the traditional desktop moving towards obsolescence."

I couldn't disagree more ;-)

Colin E.

[Peter O'Hanlon]

Indeed. The sad thing is, if the extra metro features made it onto the desktop, this would be an unstoppable OS. Why can't desktop apps take advantage of contracts? That would simplify some of my applications no end.

Oh well

---

From: Colin Eberhardt
Sent: 16/03/2012 12:47
To: Laurent Bugnion; wpf-di...@googlegroups.com

[Karl Shifflett]

Laurent,

Strongly agree with your comments.

To demonstrate I'm trying to finish the latest version of BBQ Shack. All of the backend code will be WPF.  The cash register will be WinRT Metro with the nice touch experience.

As far as running Win8 on my desktop, I'm not there yet, but I have not really given it a proper chance either. I happen to love Windows 7 (except for the Outlook and Visual Studio 10 crashes I get).

BTW:  I've got Mole running on VS 11.  Bummer, that WinRT aps do not support Visualizers.  We are efforting to get our licensing and activation changed to an in-house application.  Once this is done, I'll put out a beta for VS 11 of Mole.

Cheers,

Karl

[Sacha Barber]

Sadly I am in Peters camp.

I have nothing against Metro itself,  and like the idea of simpler touch apps, I just think metro should be like a fancy task manager is all which can be launched using windows key, I have not watched enough videos to even confirm if it may already do that.

WinRT everywhere should be the thing Microsoft worked on, that would have been the big win in my opinion.

--
Sacha Barber
sacha....@gmail.com

[Peter O'Hanlon]

I think the big question is, if you use Windows on a PC, what is the incentive to upgrade? What does Windows bring to the desktop user?

I know Microsoft has been seduced by the tablet boom, but the simple fact is that desktop PCs far outnumber tablet devices. So why offer a standard form factor for both platforms that disadvantages the majority of your users? What am I missing? That is what I want the answer to?

---

From: Sacha Barber
Sent: 16/03/2012 14:41

[Colin Eberhardt]

Pete,

Yes desktops far outnumber tablets, but what about in five years time? I don't see the desktop being eliminated, it is a more suitable form factor for daily office work, but I do see its overall share being eroded considerably. Microsoft can hardly ignore this.

Colin E.

---

From: Peter O'Hanlon

Sent: 16/03/2012 14:54
To: wpf-di...@googlegroups.com

[Peter O'Hanlon]

I didn't say ignore it. I said don't treat the desktop as a second class citizen. Microsoft had the chance to sew up both markets but they have dropped the ball here. There's no innovation for the desktop in there so there's no reason for companies to upgrade.

Had some of the Metro innovation made it in, they could have sold desktop as well as tablet upgrades. The big problem they have is that Windows 7 is so good that companies are perfectly happy to stick with it, those companies that have upgraded from XP. So I ask again. Why should companies invest in something that does not offer what they need on the desktop? I really want someone at MS to explain just what I am missing. What the killer feature is that will make me fork out money.

---

From: Colin Eberhardt
Sent: 16/03/2012 15:18
To: Peter O'Hanlon; wpf-di...@googlegroups.com

[Sacha Barber]

Yeah but they are kind of ignoring desktop devs by doing this too, you must all see that surely

--
Sacha Barber
sacha....@gmail.com

[Brennon Williams]

Not sure if my comments will conflict or agree – probably some of both. No offence intended. A lot of this are “my thoughts” – the are based on strategic decision making and how I logically see it playing out from a commercial point of view. Commercial scenarios control the technology choices we have in front of us, not technical features. That is the way it has always been.

 

 

 

Metro is for both desktop and tablet form factors. You must demonstrate that your WinRT based solution can run on both for marketplace submission. Thinking that it is for one over the other is naïve – I am not saying that it is the right way to go of course.

 

The start screen is just that – a start screen, an extension of the existing start menu. It is not the place where WinRT application run. It is a snapshot of applications that have access to the start screen (menu) properties such as tiles and common contracts.

 

Metro is a touch first UI technology, but beyond that (and no chrome) they are what we traditionally think of as desktop applications. WinRT is everywhere in Windows 8 and I think what we will see is that you will have the choice of building a chromeless Metro style application for “consumer” experiences, or a traditional Windowed application with no constraints on design language implementations for “enterprise” experiences – both in WinRT with the features and benefits of WinRT. We already know that in order to target ARM – you will need to provide two different install packages, and it is not inconceivable that you will also be able to specify “metro” or “desktop” either. The reason why is that companies like Banks need to be able to privately distribute and install solutions without public stores. I don’t think that will be there for GA – maybe, but I wouldn’t bet on it.

 

The downside to that with LOB is that the entire framework is not there yet – nor do I believe it will be for GA timeframes. We honestly don’t even have parity with Silverlight 4 at this point (both in tooling and framework), but that is not to say that it isn’t coming. When you throw the phone into the mix – getting framework consistency has to be a higher priority, because it has been pure crap up to this point being restricted in one platform to the other, building multiple code bases to support similar features of the same solution on a phone compared to a desktop. This is not about creating the same application on all platforms but creating companion solutions because the use cases are different from device to device.

 

In terms of users – yes there are many more desktop’s, but that simply won’t be the case moving forward. The stats and buying trends backs that up – but again we are talking mostly around consumer use.

 

There is nothing stopping you from creating your traditional Silverlight or WPF applications and ensuring that they will run under compatibility for Win8. So where is the issue?

 

Developers in general maybe angry at the direction of MSFT – but pause for thought – The existing frameworks were heavily limited to what could be achieved with them moving forward in terms of hardware requirements. The entire architecture of the under-lying OS needed to be modified to move towards the performance requirements of solutions and hardware moving forward. This is a reason why we see HTML based “Applications” performing faster in rendering terms as compared to C#-C++ solutions. The platform is optimised for rendering and performance – as opposed to execution speed of the code that is powering the solution within the framework.

 

 

 

In the end – I hope that they remove backward compatibility altogether. They have needed to do it for years and it keeps getting pulled. I am sick of the frameworks, performance and functionality of cross platform delivery being constrained by it.

 

If users of older solutions want to use those solutions and are happy and efficient doing that – they have 10 years to get used to another scenario. As a developer, if you want to keep targeting that market, again, there is nothing stopping you from continuing to develop for it in those technologies – WPF and SL.

 

There are a whole heap of other reasons why MSFT would head this way – commercially with revenue in content distribution, eco-system ownership (people owning all MSFT devices as opposed to Apple), improvements in management scenarios, and lastly the disastrous cloud. The cloud will get better and the terms under which we build for it are about to change as well – it has to in order to make the entire proposition viable. MSFT know this. Let’s see if they can actually execute it without all the BA’s and other people that have never written a line of code getting in the way.

 

 

Either embrace the change and try to understand how to make solutions for it that benefit users, or master the tech that you are currently happy with. In the end, the users (both consumer and enterprise) will decide what works and what doesn’t. It’s our job to be able to build for that expectation and that is why (love or hate Windows 8) you should all be learning how to build for it.

[Jeremiah Morrill]

I remember the first time I quoted someone out of context ;)

"Until something like this happens, or these technologies get official presence on the desktop...It would be safe to call the traditional desktop moving towards obsolescence."

Imagine 5, maybe 10 years from now...you are developing a new WPF or SL5 application, and MS has not brought metro to the desktop (or improved upon other existing techs).  You would be the equivalent of a DOS programmer today.

It's easy right now to say, "The desktop is still viable. I can't make certain apps that fit Metro restrictions." because Metro isn't even done, or will it be really "done" by GA of Win8.  I'm talking vectors Microsoft technology is headed with the info we have now.  Having no programmable desktop on ARM might be a great hint of how they will kill backwards compatibility...or maybe not if you think x86 will be as popular in the future as it is now.

-Jer

[Colin Eberhardt]

Hi Jer,

Sorry, see your point, have been following this on my phone whilst looking after the kids. Its a WP7, lovely device ;-) but the school run is not the best time for following discussion. The best I can do is skim read and pick up on the controversial points.

I really am not upset about Win8 or MS technology right now. Quite the contrary, I don't think things have been so exciting for a long time. As long as you are happy to broaden your horizons and learn new things, which is basically what I get out of bed for.

Colin E.

---

From: Jeremiah Morrill
Sent: 16/03/2012 21:31

[Jeremiah Morrill]

No apology needed!  This is the internet :)

I agree about being excited with Win8 tech!  The APIs available to devs, even in a sandbox are top notch.  There's a few areas I can complain about, but isn't there always :)

-Jer

[Michael Brown]

The XAML DX interop I saw a mile away...they need it if they're going to support running WP7 apps on Win8. With regard to the rest...it may be burnout talking but I'm pretty much feeling m'eh about the whole technology carousel. I'll let the next group figure out how all this stuff works. Not sure why they took the start button away from the consumer preview. If having it bring up the Start Screen was too confusing, not having it at all is probably worse. Like I said right now, I'm really not as hyped about Win8 as I should be. It might pass, it might not. It's just another cycle of learning followed by a few years of waiting for the market to follow. Judging from the adoption of WPF, we have about 4-6 years before we start seeing any true demand for it.

 

Then again, I could be totally off base and it takes off like gangbusters. There are a lot of issues holding it back from being the default new project template for the business world. But the consumer space may take a liking to it. We'll see.

 

--Mike

---

From: wpf-di...@googlegroups.com [wpf-di...@googlegroups.com] on behalf of Jeremiah Morrill [jeremiah...@gmail.com]
Sent: Friday, March 16, 2012 6:16 PM

To: Colin Eberhardt
Cc: wpf-di...@googlegroups.com

[Sacha Barber]

It may be burnout talking but I'm pretty much feeling m'eh about the whole technology carousel. I'll let the next group figure out how all this stuff works 

ROFL that seriously cracked me up, and I kind of feel like this too.

--
Sacha Barber
sacha....@gmail.com

[Brian Noyes]

+1 Keep trying to get myself psyched up for it, but always end in a "meh" kind of mindset

--
-----------------------------------------
Brian Noyes
Chief Architect, IDesign Inc
Microsoft Regional Director / MVP
http://www.idesign.net
+1 703-447-3712
-----------------------------------------

[Sachs Barber]

Shame really, just not that cool. Law dictates odd windows versions rule, win 9 os will be the one

Sent from my iPhone

[Colin Eberhardt]

I'm definitely in the 'meh' camp with you guys. Love the metro look, but not feeling inspired by the development environment.

---

From: Sachs Barber
Sent: 20/03/2012 19:08

[Josh Smith]

Me too. I *want* to be OCD-level obsessed with WinRT/Win8/etc but I just don't care much at all. It's another almost-beta from MSFT of something that I am not even sure will work out in the end. I almost feel like investing in WinRT anything more than casual interest is risky, considering all of the other hot technologies worth learning these days.

josh

[Peter O'Hanlon]

What's interesting is that opinions have varied from one week to the next from "I must learn this" to "I hate this" to "I have nothing but complete apathy for this". Maybe I'll be more interested when I see practical demos of Dynamics running inside Metro (a few of my clients use Dynamics, so that might interest them - although saying that, we just demoed a product using Dynamics running on an iPad).

--
Peter O'Hanlon

[Josh Smith]

Pete, 

I think most .NET client devs are feeling ambivalence towards WinRT. I know I am. It's not that I "hate" it or "love" it. Really, it's just annoying that there's yet another platform, with its own bugs and quirks, that MSFT is releasing that we all "should" learn. First there was WinForms, then WPF, then Silverlight, now WinRT. Sure, professional programmers should not opposed to learning new things, but at this point I feel that learning the latest Microsoft stack is an exercise in futility. They keep pulling the rug out from under us.

I can already hear the rebuttals of "But if you know WPF or SL, you practically already know WinRT!" To that I say, so what? The delta between WinRT and WPF/SL seem significant enough to require non-trivial time investments. And therein lies the rub...

Josh

[Peter O'Hanlon]

Well, I'm currently dividing my time between CodeStash and HTML 5 running in Chrome - as IE won't support WebGL, we are having to discount it for some of the rendering work we are doing. We are getting some pretty impressive results with LIDAR imagery in Chrome, and as we improve our backend server functionality with decent realtime filters, we are starting to get some real traction from there. The thing is, none of this needs to be on the Microsoft stack - we have spent the last six months rewriting the server core in C++ and it is portable - it's a pure number and data crunching server.

 

At some point, we'll try it out on Metro, but I don't feel the need to push it right now - we can roll out our application onto Android with not that much effort, and we have got some positive feedback about offering it to our clients via iPad, and I never thought I would hear myself say that.

--
Peter O'Hanlon

[Josh Smith]

I truly believe that the new iPad is the start of an era. That whole "post-PC" thing that Apple went on about seems pretty real to me. Sure, desktops and laptops aren't going anywhere any time soon, but they won't be the center of most people's computing lives.

No longer can a tablet be a serious contender, in my mind, unless it offers the amazing clarity of the Retina display. That all but obsoletes any first-gen tablets that are running Win8 (unless they decide to include Retina-like quality screens, which I haven't even heard any rumors about). I suppose there might be a lot of people out there who care more about having Windows on a tablet than having a beautiful, popular tablet device with a great track record.

That's another reason why I don't care much about learning WinRT. Apple is crushing right now. If that sentence bothers you, I suggest you reassess your priorities.

Josh