Joe Beda - Is Avalon a way to take over the Web?

"whether Avalon will be the technology that Microsoft uses to replace the Web"

What extent do those involved in Longhorn (or marketing) etc. believe uptake to be? I think questions like the above are totally dependent on who actually has the technology on their box.

Do you see business, the home, or both, craving to shift to Longhorn as quickly as possible (say, like the Windows 95 launch, where even our newspapers ran front page adverts for 95) or do you see a more skeptical and more cautious audience to slowly win around?

Interesting video btw. added some clarity to the future

For sure, update will be slower than something like a downloadable browser.  However, we think that the capabilities will be sufficient to merit start the upgrade cycle.  The recent discussions as to the longevity of WinForms attests to the long term strategy that we have with regards to longhorn.

ChrisAn has some comments wrt WinForms: http://www.simplegeek.com/PermaLink.aspx/75ae91f2-53d9-41e9-8ca8-845367c254e3

I'm hoping that Longhorn will merit the type of excitement that there was around Win95.  Only time will tell if we can really start a wave that big.  However, we don't want to overhype it now seeing as it is still such a ways out.  It would've been cool if we could have shiped it on August 25th, 2005 -- that would have been 10 years after the Win95 launch.

Joe

So Avalon will be a sort of replacement for the .NET ease of building webservices and consumers of those webservices all rolled up into the upcoming longhorn release?

I had not thought about it before like that but now that I have been made to see the potential I feel the excitement beginning to build for that day when it finally arrives. I have been involved with building rich clients communicating over thin wires for some time now. I can say that barring the deployment of the fairly fat .Net runtimes, the use of customized web services have given our development the ability to deploy 'RICH' with 'REACH' for our customers. Indeed this has given us some wins where we might not have had them before. If Avalon expands on that notion then I for one will be happier for it.

my 2c

Mscott - Love the Avatar Bro

mscott wrote:

I wonder how the proposed 3D graphical desktop will be effected by the interaction of the web through Avalon. Will we be able to write 3D websites to be viewed on the desktop?

Most of our effort is really concentrated on doing 2D graphics in new and interesting ways.  3D is cool, but 2D is the bread and butter.  The 3D support in Avalon will fall in to two categories: 3D in an app/page and 3D on the desktop.

The 3D in an app will be a way for normal humans (not DX gods) to do a little 3D and have it integrate well with their 2D UI content.  While you can load XAML in the browser (and have it downloaded from a web site) the main thrust will be to enable regular old 2D UI people to start getting their feet wet with 3D.

The 3D on the desktop will be used mostly to show windows in new and interesting ways.  This is the stuff that our Shell team (wait until Kam's interview is up!) is doing to make the window manager cool.

I should also note that it is the DX team who is still doing the "down to the metal" D3D work.  Avalon builds on top of DX and pushes it in new directions.  While you will be able to do cool things with 3D in Avalon directly, nothing is going to replace D3D for writing Quake.

Any impressions on Sun's Looking Glass project? I remember seeing it, thinking "wow! that's so cool!" then imagining trying to find my windows in 3D space when I seem to loose them in plain old 2D...nevermind worrying about which way around they are or at what angle

It'll be interesting to read (or see!) the practical applications of 3D on the desktop from your Shell team.

On a quick note about WinForms, its certainly great technology (reminded me of Delphi), but why didn't it ship with the controls to build modern GUIs which users of XP, Office 2003 etc. start to feel comfortable with? I've had to use 3rd party controls to achieve standard functionality like icons in menus. It was something that really frustrated me, so will XAML etc. make UI development more open to any leaps in UI functionality (I'm thinking of, say XP sidebars)?

One thing at a time

I haven't taken a close look at Sun's looking glass.  The team that I'm on is providing the base capabilities for other developers to build on.  Specifically, we are providing tools for control authors and the shell to take advantage of.  Kam (one of the guys driving the look and feel of Longhorn) also did an interview for Channel9.  Hopefully he can shed more light on these questions.

As to WinForms, I wasn't involved in that, but I do see what you are saying.  My understanding is that the WinForms team is still working on providing more  and more capable controls in their future versions.  As for Avalon and XAML, we are hoping that it will be much easier for the average user to get some of those advanced looks.

However, we are upping the bar to some degree also.  Advanced graphical capabilities will allow developers with artistic skill (or access to someone with artistic skill) to push the boundaries.  Just like you can do a lot with an artist helping you out with your website, you will be able to apply some of those same skills to writing a windows program.  Not only will you be able to easily put an icon on a button, you'll have to have design that icon.

Thanks for the info guys. Just the focus I wanted to hear!

-one happy Channel 9 viewer

Make that two

"...whether Avalon will be the technology that Microsoft uses to replace the Web..."

Do you people hear yourselves?  Do you have any idea of the sheer arrogance embedded in that statement?  Sheesh.

Arrogance? Vision? Innovation?

Agh, someone help me, I can't see the lines!

Sure, most people who 'change the world' didn't mean to, but there's nothing wrong with pushing a revolutionary technology set instead of evolutionary (yeah, I know, you could argue whether Longhorn et al are revolutionary, but I'd call them that since they are several evolutionary steps all at once this time. Ones I know I'll love working with).

To me, Longhorn isn't "really far away". Let's face it, large companies (like the one I work in now) start projects which dont' require purchasing for 12-18 months.

So, Longhorn is 30 months away. That means that in a year we'll be starting projects which will use Longhorn. There had damn well better be something for us to demo at that point

For many companies that are our size, the sensation is that we are nearly on the cusp of doing some very cool things with some really great new tools. Sure, we'll need to wait 2 years (or whatever). But we're already used to waiting a year and a half so it's not that bad. It's only made worse by the fact that Longhorn excites us.

Let me see if i got this right: Avalon (with Indigo) is planned to be a revolutionary rich content delivery system with the tools to make it work with applications across multiple servers.

The Rich vs Reach debate (or even demarcation) seems to me to be a bit academic really,since this kind of technology is in demand worldwide right now. If it's not it will be: that's what it's really about right, a great way to make interfaces that do very useful things and are easy to work with (surely?) - and surely more.

I just wonder if someone will put my mind at rest that the infrastructure is in place (or acheivable) to deploy these across national boundaries or in areas where networks arent very mature. If there's currently problems using (as I read today) tools like VSS across VPN connections while they interact with databases, will the vision of Indigo (as I'm, to be honest, guessing at here) come to fruition? Is the idea to make it just so compelling that it will have to?

That'd be nice

Anyway, i was watching something last night that blew me away. It's an NSV formatted video stream (i think that's something to do with AOL now, hope that doesnt put any of you off) and you'll need Winamp 5 probably. Here's a link.

I also just wanted to throw in here, in an off topic type of way that this site is marvellous and seems to work very well with Firefox.

It's clear to me now thought that no single technology or company will ever actually take over the web : don't worry about that!

Not that I think there's anything wrong with having a go...

Is Avalon a "super" java?  Delivering Rich content with reach?  Maybe I'm just being cynical.

btrey,

I'm the one who asked the question. Actually, it's a question a customer asked me.

So, am I arrogant, or is the customer who asked me afraid?

Channel9 is designed to answer the customer's questions.

So, what would you like me to ask in my next interview?

danhopkins wrote:

Is Avalon a "super" java?  Delivering Rich content with reach?  Maybe I'm just being cynical.

Avalon isn't about reach at all.  In fact, it is all about providing a better platform for people to write Windows software.  Java, in my opinion, tried to have its cake and eat it too.  By going for rich and reach they ended up compromising on both.  IE/DHTML/Trident fell in to the same trap, I think.

Microsoft can't kill the web -- it would be arrogant of us to think that we could.  Avalon will run just on Longhorn, so, by definition, it isn't a replacement for the web.  All we can do is make Windows a more compelling platform for people to write software on.  The strategy that Avalon is taking is to take some of the things that people love about writing for the web and "adopting" those in to windows.

scobleizer,

It's one thing if a customer, particularly one who's technical expertise is limited, asks you a question.  It's another thing if you present the question as a matter-of-fact lead-in to a story on an official MS blog.

I'm not sure how you intented the statement to read, but it comes accross as "MS is going to replace the web.  Is Avalon the technology we're going to use to do it?"

I'm not a MS basher but neither am I an MS fan.  Computers and operating systems are tools.  I don't get emotionally involved with tools.  Windows has strengths and weaknesses.  So do other operating systems.  But MS as a company all to often comes across as if they rule the world.  MS doesn't own the web, they don't control the web and they certainly don't have it in their power to replace it.  If the statement was meant tongue in cheek, then quotes or some other method of indicating that might have taken the eglomania out of it.

What would I like to ask you?  I understand that Longhorn is supposed to address the weaknesses of Window's command line. Will it add real links to the file system?

btrey wrote:

I'm not sure how you intented the statement to read, but it comes accross as "MS is going to replace the web.  Is Avalon the technology we're going to use to do it?"

I can't speak for Scoble, but I can see how you could take that away from the title.  I think that he meant it as (and I answered as) "A lot of people are scared that Avalon is a way to take over the web.  Is that the case?"

The answer (which I hope you can take away from the video) is obviously no.

From a graphics enthusiast perspective, Avalon sounds like a dream.  I think many people have misconceptions about just what exactly it is however.  As I understand it, Avalon is much more like a retained mode, highly hardware accelerated, and fully scalable version of GDI+ than it is like a 3D API.  This is obviously a good thing because the average GUI developer doesn't want to worry about the details of a 3D system, they just want the blending, transforming, and smooth animation that 3D graphic accelerators can provide when used for 2D graphics.

Also, I haven't heard many people considering a new drawing model a way to take over the web, but I have heard people considering XAML in this way.  You can't blame them because as I see it XAML + Indigo + ClickOnce seems to solve all the problems I commonly run into working with DHTML + server side scripting; so I wish it WOULD take over the Web.  Although that will obviously not happen because Avalon only runs on Longhorn. 

I think it would be good if Channel 9 did an interview with someone from the XAML team so they can clear up the fact that XAML is not just for doing graphics.  I think there are a lot of advantages to moving much of your code into a declarative form such as XAML for maintainability reasons.

Good idea, we'll get over to the XAML team.

Pseudo wrote:

As I understand it, Avalon is much more like a retained mode, highly hardware accelerated, and fully scalable version of GDI+ than it is like a 3D API.  This is obviously a good thing because the average GUI developer doesn't want to worry about the details of a 3D system, they just want the blending, transforming, and smooth animation that 3D graphic accelerators can provide when used for 2D graphics.

Where does that leave games, and DirectX? Or are we just talking about a purely UI design (equivalent of Swing on steroids for the purposes of web-content) toolkit/API?

And I can think of atleast one application that could use a scaleable graphics viz. combined with a good content delivery system.

Pseudo wrote:

Also, I haven't heard many people considering a new drawing model a way to take over the web, but I have heard people considering XAML in this way.  You can't blame them because as I see it XAML + Indigo + ClickOnce seems to solve all the problems I commonly run into working with DHTML + server side scripting; so I wish it WOULD take over the Web.  Although that will obviously not happen because Avalon only runs on Longhorn. 

Avalon is more than just graphics!  It is also a framework for building UI.  About half of our team is concentrating on the low level details (like graphics or threading or input) and the other half is concentrating on controls -- either building them or building the world that they live in.

Yes, it will be network and data connected (with Indigo and rich data binding) but that is more about bringing the Windows platform into the future than taking over the web.  Again, this stuff is Longhorn only, so there is no way that it can take over the web.  The web is more than just Windows.

Karthik wrote:

Where does that leave games, and DirectX? Or are we just talking about a purely UI design (equivalent of Swing on steroids for the purposes of web-content) toolkit/API?

DX will still exist and that is where most games will be written.  It is possible (and some people already have!) to write games using Avalon.  But we are talking more toward the Minesweeper end of the spectrum than the Quake end.

Web type content is also only a corner of what Avalon can do.  The main thrust is rich UI for client applications that have easy integrated access to online data and services.  In fact, if you want to get content out to the widest set of people with the cheapest cost, Avalon isn't going to be the way to go for quite a while.  Even when Avalon is on 100% of all windows installations (after everyone has upgraded to Longhorn -- we can dare to dream, right?) you still won't be able to reach as many people as with simple HTTP/HTML.

However, if you want to do something a more interactive that takes advantage of the client in fundamental ways, and you are willing to narrow your pool of potential users, Avalon will be a very attractive option.

Joe,

I don't know if you're at liberty to shed any light on this, but I'm interested in the 3D on the Desktop plans.  I know that until now, 3D desktops have been nigh unusable, but I can't help but think that if anyone could get it right, Microsoft could. 

What I've envisioned is a fairly simple interface for manipulating windows on the desktop in 3D, using the Microsoft Mice featuring the Tilt Wheel as the centerpiece.  Standard mouse movement works in 2D, with the Tilt Wheel manipulating the Z-Axis.  An applet like Apple's Expose could temporarily bring all windows back into reach in case you "lose" them in the 3D space.

It all seems fairly straight-forward to me, and I'm surprised that we haven't seen anything like this.  Any chance this is where Microsoft is looking to go with Avalon/Aero?

SuperRob, this is in response to your comments --


There was an analysis of 3d Desktop Environments a long while back, that I'd seen on some usability forum.

It will not work out - simply because you are trying to convert a 2D windowing environment into 3D, and by doing so, the effort required to access any object rises exponentially for the third dimension (think Fitts law).

I'm not saying its not possible - similar things have been attempted before (ofcourse, not using your specific idea, but quite similar ones) - its just very unusable.

There is a reason Apple came up with just one button - there are users out there who do not even know what right and left mouse buttons are.

3D Desktop environments would be ubercool and fun to play around with, but they would largely be unusable for most people. And even more, it would be really resource intensive.

On the other hand - if only Microsoft would bring out multiple desktop environments for the same user, where I can switch between various desktops, it would be really cool.

Microsoft should go back to simple, elegant and usable designs - rather than complex hey-look-shiny-thinga-magic designs. I think Aero is headed in that direction.

Just because you can does not mean you should - especially when it comes to usability. The simplest ones are often the best.

i'm web developer and i'm totally frustrated by the poor client side programming environment.(DHTML/Javascript). I think microsoft has learned from the past(investing a lot of money on developing a free browser to kill the netscape, and now using .net to compete the java.) the competition is a great thing because it benefits the consumers.(hopefully the programmers can benefit from it as well, getting better and more powerful development API/Tools). It acutally comes down to a simple issue here. How can longhorn take over the web? Looking at the server side, it's not very difficult to implement the functionality to process the XAML data sent from the client(from longhorn supposedly). XAML is a just a xml stream, the server just need to understand its schema. But the client side, it's a little bit more complicated. It needs the fat .net API to work on the XAML data sent from the server. So the worst thing is the micorosoft controls the desktop which it's the real situation at the moment). HTTP, TCP/IP, no matter how the indigo unifies it, the data arrives on the server is still the XAML foratted data. So the server technology will not be controled by the microsoft. but the client side, it's going to be the longhorn dominated world no matter you like or not. I'm looking forward to seeing being able to develop the rich client UI(XAML) by the web(i mean the browser).

Karthik wrote:

SuperRob, this is in response to your comments --


There was an analysis of 3d Desktop Environments a long while back, that I'd seen on some usability forum.

It will not work out - simply because you are trying to convert a 2D windowing environment into 3D, and by doing so, the effort required to access any object rises exponentially for the third dimension (think Fitts law).

I'm not saying its not possible - similar things have been attempted before (ofcourse, not using your specific idea, but quite similar ones) - its just very unusable.

Oh, I understand that concept very well.  I know quite a bit about usability.  My main concern is that while simplifying the OS for the mainstream users is fantastic, people that think differently (no, not Max Users) need a different way to interact with the OS.

That's one of the reasons why I like the idea of using the Tilt Wheel to manipulate the Z-Axis ... if you don't want to use the 3D aspects of the OS, you wouldn't even know it's there.

Microsoft has historically done this very well, but I'm concerned that the reliance on tasks and wizards (while innovative in its own right) is eventually going to catch up with them.  The task panes on all these windows overlap and are obscured, and the idea of these tasks will literally get lost in the window glut.

Objects weren't a bad idea way back when, but I think the problem was that we didn't have a 3D space to interact with them.  I mean, if you show someone a printer, they know they can do with it.  Show someone a multi-fucntion device, and they know that they can print, scan, copy, fax, etc.

The problem with these "task panes" is that you have to assume that you know what the user wants to do with it, and present them with those options.  While the CD burning task is effective, I find that simply putting my files into the CD burner and telling it to burn is MUCH simpler, and more intuitive.

I think that there are ways to do both, I'm just wondering how much more research Microsoft is doing with it. 

SuperRob wrote:

Joe,
I don't know if you're at liberty to shed any light on this, but I'm interested in the 3D on the Desktop plans.  I know that until now, 3D desktops have been nigh unusable, but I can't help but think that if anyone could get it right, Microsoft could. 

Hi Rob,

I'm really not a liberty to talk about that stuff, but I admit it would be cool.  You should ask Kam once his interview comes up.  As to stability, I can talk to that.  We are building Avalon on top of D3D.  However, traditionally, 3D drivers have been stable enough to run the top 10 games.  Other games just worked around driver bugs.  We are working hard on driving new standards of quality to the video card OEMs in an effort to make it acceptable to rely on DX for everyday use.  We are going to be talking about this more at WinHEC, but things like the new longhorn driver model is one of the ways that we are doing that.

SuperRob wrote:

I think that there are ways to do both, I'm just wondering how much more research Microsoft is doing with it.

I do know that MSResearch is looking in to this quite a bit.  I don't know how much of that will make it in to the product in the longhorn timeframe.  However, with the desktop compositor, we will have the capability to explore some of these ideas better than we have ever been able to before.

Check out the Task Gallery example of something done ~4 years ago from MS Research.  Some of the same people have done some interesting things since, but I can't find it on the MSR site right now.

"Joe also has a weblog where you can keep in touch with what he and the Avalon team are doing."

Uh, right.  Like his latest barbeque and speeding ticket case.  Yup, I sure am interested in knowing about these very important activities of the Avalon team.

NOT!

Marc

He has written about Avalon from time-to-time there. I'm sure that when he's able to you'll see more there.

In the meantime you'll have to put up with reading about my BBQ.

cliftonm wrote:

Uh, right.  Like his latest barbeque and speeding ticket case.  Yup, I sure am interested in knowing about these very important activities of the Avalon team.

NOT!

Ha -- I guess the Avalon stuff has scrolled off the bottom. 

Here are some links to get you started:
http://www.eightypercent.net/Archive/2004/06/30.html
http://www.eightypercent.net/Archive/2004/06/18.html
http://www.eightypercent.net/Archive/2004/06/21.html
http://www.eightypercent.net/Archive/2004/05/18.html
http://www.eightypercent.net/Archive/2004/05/24.html
http://www.eightypercent.net/Archive/2004/05/12.html
http://www.eightypercent.net/Archive/2004/05/06.html
http://www.eightypercent.net/Archive/2004/04/12.html


It is hard to keep coming up with new topics.  There is a ton of info but I'm not sure what people are interested without repeating myself ad nauseum.  If there is something that interests you then send me an email at "eightypercent AT Bedafamily DOT com" and I'll do my best to answer.

My cousin did a lot of work on Avalon

Pete
http://www.businessservicesuk.com