Hmmm... just as a side note, I'm not sure I would trust the cross-browser, cross-platform implementation of WPF...
What ever happened to the Unix and Mac versions of IE?
Again from wikipedia (http://en.wikipedia.org/wiki/Internet_Explorer):
"After the first release for Windows 95, additional versions of Internet Explorer were developed for other operating systems: Internet Explorer for Mac and Internet Explorer for UNIX (the latter for use through the X Window System on Solaris and HP-UX). Only the Windows version remains in active development; the Mac OS X version is no longer supported."

Perhaps Kurt was a little too pessimistic about SVG in Mozilla. Maybe it's a dirty little secret.

@Jeremy,

I can see your point, but we're also talking about two different moments in time, two different Microsoft's, and two different agenda's. That not to suggest that the MSFT of 1996 had evil intentions, and therefore chose to kill the Mac and UNIX versions of IE such that they could maintain a stronger monopoly share -- you would hard pressed to prove that if MSFT had kept these products alive we would be staring at a much different desktop landscape. In fact, from the Mac standpoint, one could easily argue that an agressive MSFT could have easily integrated features into IE for Mac in an attempt to lure Mac users away from the Mac OS, though it would be a bit of stretch to suggest that in doing so they would have been successful.

None-the-less, when it comes right down to it, there are several reasons why any given software company would kill any given product, and at the top of that list sits financial justification. It's hard to justify pouring money into a product that has, in essence, zero return on investment. In fact, that's basically what happened to IE before Mozilla introduced Firefox and subsequently reignited the browser wars by making the browser matter again. Before this event took place there wasn't even going to be another IE for Windows, and I think we can safely suggest that if there is any one platform MSFT is going to take interest in providing application support, first and foremost it's going to be Windows.

If you take a snapshot of the state of the computing landscape today and then compare this to a snapshot from ten years ago, the primary difference in terms of market direction would be that of software as a desktop application and software as a service. In the software as an application world of '96++, the money is made on the applications, where as in the software as a service world of '06++ the money is (obviously) in the services provided.

The key difference between software as an application and software as a service (in terms of revenue generation) is pretty straight forward -- applications, generally speaking, are platform dependent where as services are not, and therefore your revenue generation will come from platform-specific applications and service specific applications, respectively. Obviously the software as a service model still requires applications, but the applications themselves are much more modular in that they are pieced together using smaller components that provide a particular service (or grouped services), whatever these might be.

If you look at XAML and you look at XUL, it becomes immediattelly obvious how this plays out from a platform agnostic standpoint. The XML-syntax is inherently more accustomed to the declarative-style programming model in which one doesn't desribe how to do perform a task and instead the task that needs to be completed, allowing the underlying system the ability to determine how to go about accomplishing this task. With this in mind, it becomes obvious that declarative programming models are inherently agnostic, and therefore much more portable than their procedural-styled counterparts.

The result? In a world that pays for software as a service, the underlying platform becomes less important to the bottom line. And the bottom line is pretty straight forward: How can we best profit from the current state of any given market? If that market is application-centric, your best bet is to focus on the platform-specific approach. If software as a service-centric, your best bet is to focus on the services themselves. In the end, thats really what it all comes down to: Where is the profit?

Thanks for taking the time to comment, Jeremy!

@Dan,

The problem with this is that I simply can not foresee SVG being able to gain the mindset of developers without a considerable wow factor. Firefox 3.0 is a solid year away from its arrival, and during that time developers will have been given access to an EXTREMELY powerful document, graphic, and application wiring framework in XAML/WPF/WPF/e that just so happens to support every major browser on every major platform. In this regard, it seems to me that Adobe/Macromedia:Flash is the only true competitor who has even the most remote chance of competing against MSFT. It's one thing when A/M:F has the mindshare of several million web developers, quite another when an entire legion (10mill+) of VB.NET/C# developers are suddenly empowered to move their skills directly into the web development space, leapfrogging the existing web development space by a considerable margin in terms of capability and pure and simple wow factor.

I may be completely off-base on this one, but I simply can not foresee SVG becoming a contendor, as to do so would mean delivering what is planned for Firefox 3.0 (scheduled for a release at the end of next year) within the next three months to have any chance of providing an attractive alternative to developers who will have a cross-browser/platform XAML in the here and now.

One other important point: When all is said and done, its the DevTools that really determine whether or not any given technology will be truly successful. Even with all of the potential that SVG both had and has, the development tools that integrate directly into a development language/platform with any level of marketshare simply do not exist. That's why I personally believe than in and of itself, SVG is simply not enough. It needs that platform tie-in to become a truly powerful solution, and as of yet, nobody has stepped forward to provide that solution. In fact, Adobe had the opportunity to do just that, binding their SVG engine with Flash, and they chose not to. Why this is I can't say for sure, but my guess would be something similar to the reason MSFT is spending as much time and effort as they are building support for VB.NET and C# into the WPF/e offering: When you have a solid base of developers who already understand a particular language, platform, and toolset, its kind of hard to risk losing these developers by moving forward with a completely different technology, regardless of whether it is the better overall technology.

I have to agree about SVG never gaining mindshare. I had just assumed from Kurt's SVG post that SVG development at Mozilla was effectively dead. Maybe it should be.

@Dan,

> Maybe it should be.

I think you're probably spot on! Mozilla is really going to have to think long and hard about how to best attack this space. The simple fact that they have such a powerful weapon contained inside of the Adobe/Macromedia VM I think suggests that they really should reconsider their priorities. If they want to compete, they need to provide development tools, and to provide development tools they need to find a partner with existing developer mindshare, so again, I think your spot on, without a doubt!

Dead or deader?

SVG didn't have a compelling market outside of mapping applications and a) it's not that big a market and b) it is a market dominated by ESRI and they don't have to care.

The SVG people did themselves in with arrogance about being 'THE W3C GRAPHICS FORMAT: NO OTHERS NEED APPLY' just as the W3C cache began to wane.

That said, wait a while. Old things become new again when there is some requirement from the environment either technical or political that can't be done with the new dominant platform. Most of the time this requirement will come from the edges of the environment. 3D on the Web was pronounced dead and is now a hot topic in VC circles but the battle is to kill off pre-existing communities and build shiny new ones (Indians? What Indians?) because the old ones chose royalty-free values for their standards. On the other hand, there is a lot more awareness out there of the high longcycle costs of storing high cost shortcycle data in proprietary formats. There are applications for vector graphics that simply can't be done with raster without considerably more and more complex code. So 2D and 3D vector graphics communities can work together or starve separately and wait for the next wave of rebranding.

It will be interesting to see if forgetfulness trumps learning. One thing one learns from history is culture doesn't learn from history.

Any next-generation solution here is going to have to be truly open: an open spec, developed in an open forum, freely implementable by anyone, with open (and free) code libraries and such. Not really following SVG that closely, this is one thing that is clearly has in its favor, which may well be why proprietary-minded companies (including MS) are uneasy with it. The issue of the support for non-MS platforms goes back to this question of openness.

With WWW you mean Windows Wide Web?
With decreasing you mean decreasing like x-(2sin(x)) ?
Of course Microsoft will pull of an impressive trick, but the world learns about "vendor lock-in"...

@len,

>> Dead or deader?

Ahh... Good point. Which is really a sad thing given than SVG is such a well designed and thought through specification.

>> SVG didn't have a compelling market outside of mapping applications and a) it's not that big a market and b) it is a market dominated by ESRI and they don't have to care.

Very true.

>> The SVG people did themselves in with arrogance about being 'THE W3C GRAPHICS FORMAT: NO OTHERS NEED APPLY' just as the W3C cache began to wane.

There has certainly been a sense of "What do you mean? What other formats even matter???" in regards to SVG, thats without a doubt. Fair enough: Given the fact it is such a well designed spec that serves its primary purpose of a simple, straight forward vector graphics language, its understandable why. None-the-less, it wouldn't be the first time that the assumption made that it would win regardless was the dagger through its heart.

>> That said, wait a while. Old things become new again when there is some requirement from the environment either technical or political that can't be done with the new dominant platform.

That's actually a really good point. Again, it wouldn't be the first time something like this took place. In fact, one could argue that XMLHttpRequest is a perfect example of an old and, for all intents and purposes, buried technology that suddenly found life because of a simple cliche marketing terms bringing realization to fact that it existed in the first place.

>> Most of the time this requirement will come from the edges of the environment. 3D on the Web was pronounced dead and is now a hot topic in VC circles but the battle is to kill off pre-existing communities and build shiny new ones (Indians? What Indians?)

Ooohhh... VERY GOOD ANALOGY!

>> because the old ones chose royalty-free values for their standards. On the other hand, there is a lot more awareness out there of the high longcycle costs of storing high cost shortcycle data in proprietary formats. There are applications for vector graphics that simply can't be done with raster without considerably more and more complex code. So 2D and 3D vector graphics communities can work together or starve separately and wait for the next wave of rebranding.

It will be interesting to see if forgetfulness trumps learning. One thing one learns from history is culture doesn't learn from history.

@Bruce,

>> Not really following SVG that closely, this is one thing that is clearly has in its favor,

Ummm... Not so sure that can be said about *ANY* W3C-based specification. In fact, one of the problems with the W3C that has made the IETF such an attractive alternative (maybe alternative is the wrong word here) is the simple fact that the W3C does things behind closed doors. Every so often they open things up for public comment, but its yet to be seen how much influence the public commenting process even has on the final outcome. SVG is open, yes. But it certainly wasn't designed in open forum, and I think you would find nothing but brick walls if you were to attempt to find support for developing community extensions to the spec, which is how you would gain the "open spec" attribute value of true.

@stelt,

>> With WWW you mean Windows Wide Web?

Okay. You got me on that one ;)

>> With decreasing you mean decreasing like x-(2sin(x)) ?

Not sure I understand the comparison.

>> Of course Microsoft will pull of an impressive trick, but the world learns about "vendor lock-in"...

Fair enough. But isn't vendor lock-in exactly what Flash represents (or at very least represented)? Flash is as proprietary as proprietary gets (or at very least, it was) and it seems to be doing just fine, REGARDLESS of the negative attitude towards its usage via a significant portion of the WebDev community.

But if perhaps the process behind SVG is not fully open, surely the MS alternative is MUCH less so. I can't believe why anyone would buy into any single-vendor solution these days.

@Bruce,

>> But if perhaps the process behind SVG is not fully open, surely the MS alternative is MUCH less so.

About the only difference I can think of is that @ the W3C you can buy your way into a working group. For that matter, you could probably buy your way into the development of a Microsoft technology, so the difference between the two is probably not all that much. Well, actually there is a *BIG* difference -- for better or for worse, there is a single person who will have final say as to the final design of any particular specification. At the W3C, or any other standards body, you gain the [advantage||disadvantage] of any given party being able to stall the development of a specification for one reason or another.

If you look at the example of XQuery and LINQ, two similar (yet not-so-similar ;) technologies, the former took five years to develop, the latter ~2 years, with a year prior to this as part of the COmega project @ Microsoft Research. And the reality is that it's really been less than a year that they really began to tighten things down and push forward with a somewhat definitive focus of what the language was going to include and what it was not.

And in that same time frame there has emerged the possibility of plink (see: http://www.bluebytesoftware.com/blog/PermaLink,guid,81ca9c00-b43e-4860-b96b-4fd2bd735c9f.aspx ) in which will enable the ability to take full advantage of parallel programming capabilities, without the need to write any code that implements concurrency directly.

This is one of the clear advantages when you let private industry develop technology: Generally speaking, you get REALLY SMART PEOPLE like Anders Hejlsberg, Erik Meijer, Joe Duffy and Ralf Lammel working on things without having to answer to, in essence, the competition.

Please don't take this as a statement that I believe design-by-committee is inherently evil, and proprietary design inherently not-so evil... there are distint advantages to both approaches. But when it comes down to it, the market tends to select the technology which works the best for the particular task at hand, whatever that might be, and regardless of whether it was developed by a multi-corp committee or a single-corp individual (or group(s) of individual(s)), and regardless of whether it can be seen as open or proprietary. For the most part (speaking in regards to the masses) people could care less how it was developed and instead: "Does it work for what I need it to work for?"

>> I can't believe why anyone would buy into any single-vendor solution these days.

Fair enough. But they do. Again, its more about what works and less about how it was developed. The market tends to make what seems to be strange decisions on a fairly regular basis, so I can understand why you feel this way. But none-the-less, that's the way the system works.

Ah, see what happens when I put my head down for coding for a few days ... sheesh!

I've been playing recently with Adobe Mars (more info to come shortly on that) and have to admit that it opens up some very interesting possibilities. SVG IS supported under Mars, btw, so I think its probably a little premature to say the SVG has gone the way of the dodo - especially given that Adobe knows a thing or two about the specification.

One of the realisations that I've made over the years on SVG and related W3C technologies is the fact that these are ultimately just "pieces" of the puzzle, rather than the whole enchilada. Working in Mozilla, I use pretty much the whole W3C "stack" of technologies - XHTML, SVG, XForms, XSD (though moving to RelaxNG), XBL (okay, a late-comer, but its still up there) + a JavaScript that's considerably more up-to-date than where Microsoft is with JScript (why MS didn't take the time to revamp JScript significantly while they were working on IE7 is still beyond me). No - there's no fancy brand names at work here, no Vista or XAML or WPF acting as a centralizing imbrella, but I rather think that's the point. While it has its warts, the W3C framework overall is still surprisingly robust - and the applications I'm developing by using this solution are translating surprisingly well to other browsers (including IE with an Orbeon XForms engine).

It's easy to criticize the W3C (and the SVG working group in particular) ... the gods know that I've certainly done it often enough myself. However, consider their objective - to create a platform that doesn't favor any one particular vendor, that can be used without a license or obligation, that can be implemented in the same manner, then consider how extraordinary this mission really is.

I agree with Len here - things cycle. You could do AJAX in a limited fashion even six years ago, but now JavaScript is one of the hottest languages on the planet. I distrust Microsoft not because of their "evil empire" mentality - they are a company like any other - but because they are a software company that makes new and better products by breaking compatibility with their old and (by implication) inferior products, then require that the whole industry move in lockstep to the new version. Standards only benefit the industry leader when THEY are promulgating them ... otherwise they reduce the barriers to entry for competitors.

I'm not saying that Microsoft can't produce a "better" version of SVG or XForms or any other technology - sure they can - but raising the issue that we've discussed before of whether in fact the whole point of SVG or related standards is not to produce the most efficient solution for the leader but to produce the most achievable solution for everyone else. I think you can look at the success stories coming from the Dojo camp to see the implications to that.

@Kurt,

>> Ah, see what happens when I put my head down for coding for a few days ... sheesh!

When the cats away -- the mice will play, Kurt ;) :D

>> I've been playing recently with Adobe Mars (more info to come shortly on that) and have to admit that it opens up some very interesting possibilities.

Adobe Mars? first time I've heard of it (searching)

>> SVG IS supported under Mars, btw, so I think its probably a little premature to say the SVG has gone the way of the dodo - especially given that Adobe knows a thing or two about the specification.

So does this then mean that Adobe is dropping support for their existing SVG engine and replacing it with something brigher and shinier? Hey, I'm all for it... If Adobe plans to put some muscle behind a bigger, better, faster, more shiny SVG++ engine, I can only imagine what the result will be. As mentioned in the original post, I have no idea what pushed them to decide to drop support for their existing SVG engine. If its because they have bigger plans, SWEET!

>> I'm not saying that Microsoft can't produce a "better" version of SVG or XForms or any other technology - sure they can - but raising the issue that we've discussed before of whether in fact the whole point of SVG or related standards is not to produce the most efficient solution for the leader but to produce the most achievable solution for everyone else. I think you can look at the success stories coming from the Dojo camp to see the implications to that.

>> With decreasing you mean decreasing like x-(2sin(x)) ?

> Not sure I understand the comparison.

It's a function that alternates decreasing and increasing, with an overall increasing trend. Where and how you measure makes a lot of difference too: In mobiles it's big, in marketing of the tech itself it's small for example

>> Of course Microsoft will pull of an impressive trick, but the world learns about "vendor lock-in"...

Fair enough. But isn't vendor lock-in exactly what Flash represents (or at very least represented)? Flash is as proprietary as proprietary gets (or at very least, it was) and it seems to be doing just fine, REGARDLESS of the negative attitude towards its usage via a significant portion of the WebDev community.

Flash has great tools and marketing to let people know about these tools, plus was there serious competition before? The only somewhat famous tool that is known as an SVGtool is Inkscape and though that is really nice, it only does static pictures.
Tools that are both easy for 'non'-coders and are known by 'everybody' is something that SVG misses at this moment. The number of tools doing sómething with SVG, grows very fast however: somewhat silently, but steady.

And again it depends on what you look at. Flash can be somewhat accessible, but takes extra work. re-purposing, indexing, meshups, semantic web and accessibility are natural with SVG and clumsy with Flash. Where i see Flash going strong is (packaging up of?) video. A few big rich companies really offering powerful interfaces to their data. And of course still (but decreasing): the annoying websites that after a annoying splashscreen look like a pinball machine (SVG will also be abused, but the people paying for the website being built understand more that visitors usually just want to read, find or do something really fast)

@stelt,

Thanks for the understanding of the formula presented. I had a general idea of what you were implying, but wasn't completely on target with how it related. Now I am. :)

re: the rest of your comment. All very good points. If nothing else, it seems that I am learning that the SVG landscape, while the final outcome still undetermined, has greater potential at success than I had given credit. Thanks for your help in bringing some of these indiscrepancies in my argument to the surface!

In the 'dead language' camp, I've been sitting here for 12 hours a day for five months building with VRML97 not X3D. Why? I have the books and the only tools I need beyond some old VRML carving editors are PFE and the Blaxxun plugin. Old tech in old hands still has life. I don't recommend anyone start making Mayan rock carvings with obsidian knives, but if one can and wants to, it still works. Surprisingly to those with more modern blades, one can get quite far.

Then the purpose of the content hits a wall in the tools. For example, VRML97 has rock-wall Inline namespaces and while that sounds good to a dataHead, to a real-time 3D author, it's a pain in the patoot because routing events (think ints, timestamps, strings, etc.) is how one determines contingent behaviors as in any GUI, but let's say one wants to keep up with viewpoints (literally cameras) and when a script is completed or a sensor fires, unbind that viewpoint. If the current viewpoint unbinds inside an animated viewpoint that is also inside the inline, one has to accept that it pops the stack and the user finds themself somewhere else. The predictability of where that will be depends on accepting the last bound node, or having some kind of floating viewpoint that other sensors are updating. Not bad for a set of local stories but if one is trying to present linearly, not good.

In X3D, inline opaqueness goes away. So at that point, one moves on to the new spec.

In my opinion, that is what should drive the choices of tools: not planned obsolescence but planned enhancements that actually do improve the lot of the builder.

As for SVG, worrying about the spec org polities takes too much of our time and is really a 'arms too short to box with deities' game. In my opinion, vector graphics are a family of application types where authors and content owners have a natural familial relationship that they should exploit to keep the business relationships of their tool suppliers from screwing them. SVG has more in common with X3D than it does with HTML. That should be obvious but too often when careers and egos and company logos are up front, it isn't. Believe it or not, familial relationships among applications are waaaaay more powerful in the medium cycles that drive the long cycles than say Moz or Fox or IE wrappers and their supporters. Content has a natural affinity that code doesn't have at the same strength although some think the reverse to be the case.

Content is the strongest reason for open systems, not objects. Objects tend toward closure by their very nature as instances of classes. Classing itself as an activity is an act of closure with some measure of relief given feedback therefore evolution.

@len,

>> As for SVG, worrying about the spec org polities takes too much of our time and is really a 'arms too short to box with deities' game.

Point well taken.

>> In my opinion, vector graphics are a family of application types

Another good point

>> where authors and content owners have a natural familial relationship that they should exploit to keep the business relationships of their tool suppliers from screwing them.

I agree. But how?

Don't chase trends to be trendy. Master a tool and use it to build good content. If a new tool offers a feature you need, get it and master that. If it is just a replacement for the last tool, don't.

Why do you think HTML hasn't been supplanted? Why do you think that Flash, despite the fact of being a closed system, succeeded?

Why do you think SVG has such a hard time getting a foothold outside of Adobe products?

Why is porn still the largest swapped content type?

The buyer has power. The content author has it to the degree they can connect to the buyer. Some technologists have become middle men who try to disrupt and control that connection to skim off the top. The WWW has shown time and time again that it is unable to route around them. The same kinds of content are being recycled through ever more byzantine pieces of technology and with the single exception of the 3D worlds, very little new is being seen. YouTube? Gimme a break. It is the ultimate expression of ADD for the unsophisticated web surfer.

You want innovation? Take risks in the content. The technology is just stuff but it gets the press and the money follows that. We are told the web is a social system; so is the telephone. It didn't become a hot market again until the content on it became more than talk. There is a limit to online sociability and the kids are already discovering they need to un-jack to get jacked; so they are moving back into meatspace where the jack is.

So the smart money and developers pick a community that needs tools to keep their meatspace connections in working order. That is what you want to do with your new project. The problem for the technologists there is that they have to find ways to access and sell to that community in languages and with terms that the community (not the Web; it is a polyglot, a real community such as plumbers) understand and accept.

We're back to old fashioned business: sell services, not ideals.

You may well be right, and I'm not going to put any money on it. :) But for another possible outcome, can I point you to Why Microsoft might not embrace XHTML (and then again they might).

In it I say:

I certainly think that if the W3C was to get its act together and turn its collection of disparate languages into a coherent whole, then XHTML could be a serious challenger to XAML, and that might make it less attractive for Microsoft to support.

Of course, it might make it more attractive to support, too. Standards are a double-edged sword for the big guys like Microsoft, since they would all ideally like us to only use their proprietary formats, but they generally realise that this isn't the way of the new internet. It's interesting that we haven't heard much about XAML in the recent period, which either means that MS are taking advantage of the 'current obsession with Ajax' to quietly get XAML ready, or they realise that the 'current obsession with Ajax' means that people actually like standards.

With the conclusion being:

But then the last year or so has been quite a surprise for everyone, and one of the key messages is the groundswell of support for open standards and consistent browsers. In today's climate it would take an incredible marketing effort to 'sell' XAML as the 'new language' of the web, and you do wonder what exactly it would gain them anyway.

@len,

Firstly, my apologies for the late response! I started to respond, got distracted, and then forgot to finish.

Secondly,

>> Don't chase trends to be trendy. Master a tool and use it to build good content. If a new tool offers a feature you need, get it and master that. If it is just a replacement for the last tool, don't. > Why do you think HTML hasn't been supplanted? Why do you think that Flash, despite the fact of being a closed system, succeeded? > There was a need, that need was filled. No need to learn something new if what you know, works.

>> Why do you think SVG has such a hard time getting a foothold outside of Adobe products? > I have though about SVG plenty of times, and I have thought about XForms plenty of times, and both times I have chosen a different route due to the fact that neither of them has enough widespread support to achieve the necessary critical mass of "99.x% of the world has default support in their primary browser, and the remaining .x have access to a secondary browser if they really and truly want to access the site."

With SVG, while the Adobe plug-in works, it doesn't really provide a seamless compound document solution (which in my view is mandatory!), Mox/Fx still has limited support, Safari seems to have placed their focus on Canvas, and Opera, while the support most definitely exists, their market share is simply too small to suggest that in and of itself, it's enough to push SVG into the "enough people have support" category.

Of course, I do believe that when it comes to the mobile web, Opera most certainly has a chance of getting a strong enough foothold to make SVG a significant success in the mobile web space. Time will tell, but it seems to me that this is at very least a realistic possibilty.

>> Why is porn still the largest swapped content type?

Cuz' Google makes it easy to find for free? ;)

>> YouTube? Gimme a break. It is the ultimate expression of ADD for the unsophisticated web surfer. > You want innovation?

Yes please. :)

>> Take risks in the content.

Oh, I think we can safely state that I most certainly agree! That, a secret, it is not. :D

>> The technology is just stuff but it gets the press and the money follows that. We are told the web is a social system; so is the telephone. It didn't become a hot market again until the content on it became more than talk. There is a limit to online sociability and the kids are already discovering they need to un-jack to get jacked; so they are moving back into meatspace where the jack is.

Very good point!

>> So the smart money and developers pick a community that needs tools to keep their meatspace connections in working order. That is what you want to do with your new project.

Why yes, it is :D

>> The problem for the technologists there is that they have to find ways to access and sell to that community in languages and with terms that the community (not the Web; it is a polyglot, a real community such as plumbers) understand and accept.

Agreed!

>> We're back to old fashioned business: sell services, not ideals.

AMEN BROTHER!!!

@Mark,

Thanks for your comments! Please see my follow-up update at the top of this post.

Of course, MS may also be working on a next-generation XML browser similar to the concept of the XSmiles browser (www.x-smiles.org). If so I would expect SVG-based functionality within that.

@Ralph,

You may be right, and I believe that starting over from scratch with a next-generation web browser (as opposed to retrofitting IE -- IE7 is an obvious compromise between doing as little damage possible in regards to breaking the existing IE-specific web while at the same time attempting to integrate a more standards compliant rendering engine, making security the primary driving point of the release) makes a lot of sense. Given that IE is (more than likely) already on the system, no web-breaking would even need to take place, while at the same time providing an interface for the next generation of web-technologies to make their way to the Windows desktop.

That said, I can't think of any reason why just such a technology couldn't and wouldn't be developed utilizing the existing integrated support for MSFT's next generation web technology: XAML. Why go to all of the trouble of building Yet Another Desktop Application when, for all intents and purposes, the *traditional* desktop application development lifecycle has long since expired, having moved well into the end of lifecycle support phase?

Take, for example, Windows Live Writer [ http://windowslivewriter.spaces.live.com/ ] -- While yes, it is still a desktop application, it certainly isn't anything close to what could be termed traditional, providing direct integration into both Internet Explorer and Firefox. Of course, not providing direct integration into IE and Fx wouldn't make a whole lot of sense for a *web*logging tool, but in and of itself, I believe that is the entire point -- The web is so completely integrated into each and every one of our lives that it has literally taken over as the centralized focus of our daily habits. And at the center of it all is our web browser(s) of choice. So why fight it by providing Yet Another Way to access and interact with the web when, again, for all intents and purposes, that same capability could easily be developed as an add-on extension to what already exists?

Add to this,

- the fact the with WPF/e, this same modularized/componentized utility could effectively run on every major browser on every major platform
- the fact that we are moving into a services-based tech economy.
- the fact that these same mentioned services (e.g. instant-on, instant-off, pay for only that in which you use applications and/or pay a subscription fee each month to gain access to a core base of Click-Once-styled applications) could easily be built using XAML+C#/VB.NET/Any-other-CLR-enabled language

... and what you have are a whole bunch of reasons *NOT* to build Yet Another Traditional Web Browser and instead, a whole bunch of reasons to integrate new capabilities into what already exists. This same model would fit well with the point Mark Birbeck made regarding MSFT not attempting to replace, and instead enhance what is already provided by (X)HTML, as if you integrate with what already exists, then what already exists doesn't need to be replaced, and what is new can become a cross-breed of available technologies, using the language/tool that best fits the need of the application being developed.

One final thought: I think history has proven that if MSFT wants to go after a particular market, they will purchase before they build. Given the fact that Opera already provides support for all of these technologies, coupled with the fact that Opera runs on every major platform and, in particular, has a primary focus on the mobile-web, if MSFT had any desire what-so-ever to expand into the standards-based web browser world, my guess would be that they would make a bid to buy Opera before starting over completely from scratch. In fact, it was about this time last year that I made a plea to MSFT to do just that >> http://www.xsltblog.com/archives/2006/01/hey_vsnet_team.html

the svg thing akes no sense to me. I use it all the time, it's very well supported, uses bugger all cpu, gives stunning results in a tiny file and, above all has for years benefited from an open xml standard that allows developers everywhere to incorporate svg support in about 10 min of coding.

i work in a print shop. i take scalability without loss of quality seriously. my peers haven't invested the minimal time to master good use of svg tools and here's why.
poor compositional skills, poor artistic sense, poor grasp of essential and elemental design.

my response to this post in a single phrase:
'Only a poor workman blames his tools'