Mobile version (Display Regular Site)

Skip to: Navigation | Content | Sidebar | Footer


Weblog Entry

Web Apps are Hot

June 23, 2004

An ongoing conversation about web applications is highlighting key points about the future of computing, the web, and the industry.

While Joel Spolsky’s seminal article on Microsoft losing the API war appears to have been the first drop in this torrent, it echoes conversations that go back least as far as the W3C’s Workshop on Web Applications. (which is the same workshop that people like Opera employee Ian Hickson came away from, disillusioned, just prior to the announcement of the WHAT WG [which was covered on this site earlier this month]).

A Technorati search reveals the interest in web applications in general (and Joel’s article in particular), but what has been particularly interesting to me is how the conversation has morphed as it continues.

The beginning, or at least where I came in, saw discussion of which technologies future web applications would be built on. HTML and CSS are great for documents, but not so good for applications. Microsoft has XAML on the horizon, the W3C offers XForms and SVG, and the WHAT WG aims to offer an extension of basic HTML that’s backwards compatible with IE6. Two of these technologies require radical change in end-user technology; one does not.

Whether Joel’s article was written in response to, or prior to these discussions is largely irrelevant — whatever the case, the publication shifted the entire tone of the conversation. Instead of discussing the technologies themselves, it moved web applications into the spotlight as being an important way forward and a threat to traditional software business models.

John Gruber picked up at this point and wrote yet another tone-setting piece, The Location Field Is the New Command Line. Arguing that web applications provide different, less comprehensive, but easier methods of interaction, Gruber rightly points out that Google’s mega-popular GMail is light years ahead of the competition in the webmail space but almost non-competition when compared against desktop mail clients. Gruber concludes that “… web apps don’t need to beat desktop apps on the same terms. What’s happened is that they’re beating them on an entirely different set of terms.”

Steven Garrity goes on to compare Gruber’s reasoning with yet another article buzzing around the web, Cory Doctorow’s presentation on Digital Rights Management given to Microsoft last week. Ian Bicking throws two cents in the ring with his list of features web applications offer that traditional desktop apps do not.

Of course, nobody seems to be talking about Macromedia’s Flex and the continuing application development capabilities within Flash itself. Hindered by stigma that Flash is evil and purely about animation amongst developers that should be buying into Flex, it’s hard to see this combination taking off; however, as a dark horse, it’s a hard one to ignore on sheer numbers alone. Ninety-odd percent of web users have Flash installed, and Flash renders consistently cross-platform. Remember what Joel reminds us though: it’s all about developers.

An incredibly thought-provoking article in April speculated Google is building a massive, global platform for distributed computing. Microsoft has renewed an interest in developing Internet Explorer. And don’t forget that mobile and wireless devices are reaching incredible saturation levels — the global mobile phone subscription base is expected to hit 2 billion in 3 years. All these things are related.

It’s a lot to digest and that’s only scratching the surface. The obvious trend is the key to understanding the future of computing: the web is it. Servers are becoming more important than clients. While raw processor power will remain useful for applications that need it, simple and general purpose data management — including email, scheduling and time management, office applications, and all other text and information manipulation tools — will increasingly move to a globally shared environment that makes it easier to collaborate and easier to access. The recession is over, the slump is ended. Web development is in demand, and the demand is only going to increase.


Reader Comments

1
Jim says:
June 23, 02h

Your last two sentances are what every web developer wants to hear. I hope it’s true.

2
Keith says:
June 23, 02h

I think you’re last two sentences are going to be proven to be true, and it’s good to note that even though Web applications are a big (huge) part of the Web’s future (along with content publishing and the informational Web) they have huge room for improvement in many areas. Designers and developers rejoice. Heh.

It is quite a bit to take in, and thanks for the post. I really enjoyed Gruber’s piece. I think in many ways the location bar IS the new command line.

3
June 23, 05h

It’s kind of scary now that i think of it, the server based world Sun predicted a few years back really is the way thing seem to be headed. Though it doesn’t seem it will be the Java only solution they propsed.

The real advantage the web really has is that it’s there. Wherever you are, if the nearest computer is connected you can jump on and use that App you need. No installing it on every computer you need it on, no inability to work outside the office.

*Loves the WEB*

4
Vlad says:
June 23, 06h

I strongly agree with Spolsky’s position, but I think it doesn’t exactly apply completely to all Web Applications.

Look at the most successful Web Applications of today - Amazon and EBay. They work very well in pretty much any browser on any OS. I’m sure there are things the developers of those applications wish they had access to, but the fact remains that they were able to get their job done and make a lot of money on an amazing application. This type of development is so successful, as Joel and everyone else has noted, that I can’t see it changing for a number of years even if WHAT develops some nice extensions to HTML that are implemented in 2% of the world’s browsers. The developers already have enough trouble fighting their management for the right to make their code work at all in Mozilla and Opera (e.g Gmail which doesn’t yet support Opera), let alone code to some specific extensions that those two companies have invented.

So, my position is that today’s Web Applications are doing mostly ok, and there isn’t a need to cry for help yet. This is how I read Spolsky’s article.

However, there are a whole set of Web Applications that can’t be done with HTML/CSS/JS. For example, anything with interactive graphics, multimedia or complex network interactions. This is the market that Flash is addressing, and this is the market that XAML is attacking. I can’t see anything in WHAT proposes that will lessen this threat.

If you look at the Spolsky’s followup article, you see he asks for a bunch of features that are incompatible with HTML, CSS and IE and *will* require a radical change in end-user technology.

I guess I just get the impression that the W3C is fighting with XAML, while the WHAT group are a bunch of casualities from the browser wars trying to relive their glory days. As Joel commented, the Microsoft/W3C approaches won’t replace today’s Web Applications. This is a good thing. I’m not sure they’ll even be used. The same applies to WHAT to some extent. But let’s not confuse who is fighting who. Today’s browser wars are over, and as Gruber says, the browser was the wrong war for Microsoft anyway. Microsoft realise the war is about technology now, not about a particular implementation. Let’s hope that enough people keep their heads out of the sand to be able to fight the new war if it happens.

Maybe it is just this week’s blogging trend?
Sorry for ranting :)

5
Mike D. says:
June 23, 06h

People are right in saying that web applications are already possible with the tools we have: HTML, javascript, and the DOM. That combination has been able to produce web applications (like Gmail for instance) which are just good enough for their convenience factor to outweigh the advantages of a truly rich client (like a client-side e-mail app).

There is only so far you can go with HTML though and companies like Microsoft and Macromedia know this. Not only is there only so far you can go, but HTML was not even meant to go further. HTML is a language with which to mark up documents, not to create applications. The fact that it has come this far is a testament to work workers’ creativity.

If you’re wondering why Microsoft stopped updating IE years ago (yes, the decision was made years ago – not when it came out in the press), it’s not because they won the browser war. It’s because they lost interest in HTML. I can’t say how but I have first-hand knowledge of this… just trust me. Their recent renewed interest in IE does not signify a renewed interest in HTML. It’s a good thing, yes, but it’s only an admission that Longhorn may take a little longer than they originally planned and that Longhorn won’t be able to wean people completely off of HTML.

Microsoft has moved on to true web applications in Longhorn and that is what Macromedia is trying to beat them to the punch with. Rich clients which display XML data in very rich ways. There are clearly standards-related things to worry about such as who’s XML-dialect will win out (Macromedia’s, Microsoft’s, W3C’s, etc), but what standards junkies will love is that the XML data going into these applications will look A LOT prettier than XHTML. A pure separation of style from content.

6
June 23, 09h

In light of comment 4, it’s useful to distinguish between Web *Services* and Web *Apps*. While the line is somewhat blurry, Web Services are things like Technorati and eBay (and every blogger’s RSS feed). Web Apps are things like GMail and Intuit’s “TurboTax for the Web”.

IMHO, the direction in Web Services is moving *away* from the browser and toward specialized clients which are more lightweight and whose user-interfaces are more finely-tuned to the Service at hand. My poster child for this is Watson
http://www.karelia.com/watson/
but yours might be as close as your favourite RSS Aggregator.

When it come to Web Apps, the story is more mixed.

Some things are very naturally-suited to being packaged as Web Apps. Webmail is a prime example. Much of the infrastructure associated to the desktop email client (POP3, IMAP, … ) is an unhappy kludge associated to the fact that you want your email to collect in one central location with good 24/7 internet connectivity, whereas you want to be able to access your email from many different location, some of which may have only sporadic internet connectivity (like whenever your laptop joins a WiFi hotspot). A really good Webmail App *solves* that problem once and for all.

Other things are less obviously suited to being packaged as Web Apps. Intuit’s Web App version of their tax preparation software has only one advantage (price) and a host of disadvantages, relative to its desktop cousins.

And that, despite the fact that Intuit’s App is, in a certain sense, also a natural candidate for the new paradigm.

The current hype about Web Apps is just another aspect of the general push, from some sectors of the IT industry, towards a subscription model for software (Sun’s “free hardware” pitch is another). Rather than buying a shrink-wrapped desktop application, which is then yours to use until the vendor can coax you into upgrading, you’ll now pay an annual subscription fee to use the software. And if you stop paying, … there goes your data.

I suggest a little thought-experiment: go through your most-frequently used desktop apps and decide how many you would willingly trade for an (even significantly cheaper) Web-based version.

That many, huh?

7
June 23, 10h

I really hope Flex doesn’t take off. From what I’ve seen, its very similar to XAML in its ability to take us right back where we started;

<?xml version=”1.0” encoding=”utf-8”?>

<mx:Application xmlns:mx=”http://www.macromedia.com/2003/mxml”>

<mx:Label text=”Hello World!”/>>
<mx:Label text=”Hello World!” fontSize=”40”/>

</mx:Application>

That code above is from the Flex example repository on Macromedia (http://www.macromedia.com/devnet/flex/example_apps.html#), notice the way they apply fontsize. Oh dear. I actually believed Eric had managed to enlighten them too.

8
Matthijs Aandewiel says:
June 24, 01h

Hey dan,

i disagree with you on the ‘lash is not evil’ comment.
Flash disables anyone wothout a flash player to view your website.. it adds those splash pages, because people dont understand that i dont need distracting flashes and annoying sounds. I honestly believe, that flash has never contributed to the “spreading” of information. I agree, it looks nice.. sometimes it is beautiful.. But i have never thought: “Wow! That’s so great! It’s good that the flashing light in the top left corner flashes, else i would have never seen … “

9
Kas says:
June 24, 01h

This is an amazingly well-timed piece for me - I’ve been involved in several “discussions” on this very topic this week. Despite the fact that for several years we’ve been successfully creating complex, open and compliant web apps, some people have been driving hard to return to Windows-locked approaches. I’m not sure where the desire is coming from; atho’ I could make a few guesses. ;-) All I do know is everything they showed us that “just has to be win-locked” doesn’t. Not even a little bit. Thanks for sharing this info - I’ll add it to the conversation and see what happens!

10
Isofarro says:
June 24, 01h

Certainly the interest in web applications is there - from the vendors and the developers. There’s a handful of different projects assembling the right building blocks for web applications - so at some stage developers need to start plumping for which toolset they will use. The groups are: WHAT WG, Mozilla, RICHIE, W3C web applications, XAML.

I feel that WHAT WG and Mozilla fit in well with the Web, Mozilla is better right now because web apps are deployable right now.

11
June 24, 02h

Nice roundup of the current thinking on web apps Dave. However you mention, XAML, XForms and the WHATWG HTML forms extensions, without mentioning Mozilla’s XUL.

XUL offers fully rich client applications which can be deployed over the web. Plus unlike XAML, XUL is cross platform, open-source, and is available and working today (if you’re using a Mozilla based browser, you can run this rich web application right now http://www.infodraft.com/~faser/mab/content/mab.xul ).

More? Web reference have a good intro: http://www.webreference.com/programming/xul/

12
Adrian says:
June 24, 05h

Jan, thats a very good point, one which I had forgotten about. I think that definitely gave web apps the ‘in’ the needed to get going.

The problem was then corporates caught on that employees could access all this cool ‘non work’ stuff from ebay to personal emails to blogs.

Now I the majority of my friends are banned from at least one website, blog or web mail page, and the majority from most of them. So when you can’t blog, or read a blog, or read webmail because they are all web apps, and it doesn’t take much to block them on a firewall, what then?

The very thing that makes web apps great (the easy to get to bit) is being blocked by immature corporates. That could kill or slow down web apps enough to mean it never gets a decent foothold outside the home. Which means the critical mass needed to make something pervasive, and too have enough of an effect to get browser developers to include the functionality in them is reduced.

I’m not saying this is the way things will go, but it is a concern

13
web says:
June 24, 05h

One thing I don’t understand is why Microsoft decided to go with a web interface for their windows update utility rather than a standard API?

Its clunky at best and I believe that to be able to use the web interface they needed to poke open a lot of security hole in Internet Explorer.

Would it not have been easier to develop an “updater” as a Windows App rather a web page? Maybe I’m just missing the point here, but 98% of the software that I have used that has “automatic updates” uses a standard API to achieve this.

I guess you could write a whole book on questionable Microsoft business practices.

14
JPS says:
June 24, 06h

web: the updater really is a windows app - it dld’s an ActiveX control from the page which does the heavy lifting. The web pg is simply a convenient location to see the available updates and trigger the control.

It uses the “holes” that already existed to good effect IMO.

15
ForgetFoo says:
June 24, 06h

man, what a great read…! lots a good links and you drive home the overall point of it all…. and i loved the last paragraph.

web-apps will start to rattle the cages once again, and the slump is over. heh.

16
Micha Schopman says:
June 24, 06h

Some browsers need to improve support for Javascript and XML. The latest Opera isn’t even capable of handling XML document, it does not support xmlHTTP or even createDocument. Same story for Safari.

I heavily use the browser as a platform for applications, and it is just plain stupid only Mozilla and IE (or Konqueror CVS) are able to handle XML.

17
June 24, 08h

It’s a shame that the stigma of bad animation is going to be around Flash for a while to come. I’ve used it, along with Central (http://www.macromedia.com/go/central), to create several in-house apps for maintaining our company’s website and customer data.

The designer in me loves it because you can actually place objects *anywhere* you want by just drag and drop rather than try and figure out the CSS for the exact placement. You can customize any element to look exactly like you want, and you never have to sacrifice your typographic sense since you can use any font you want. You can be as creative as you want and not have to sacrifice because of the limitations of the medium you’re designing for.

The developer in me loves it because you don’t have to worry about statelessness - you can take user interaction, feed it to a server, return something different and the user never has to know (with page reloads, redirects, etc.) It’s a beautifully seamless experiece, much like a regular ol’ desktop app.

The idea of Flex is basically to provide the best of both worlds - the rapid development of webpages with the interactivity of Flash. For me, a big drawback is the cost. The last estimate I heard was $12,000 for the server software! And it’s not even standalone - you need to run it ontop of a Java server! Yikes!

18
Scott says:
June 24, 08h

(Perhaps off-topic, but related on implementation terms..) Safari 1.2 in fact does support the XMLHTTP object. To quote, it’s “a welcome addition” even though it hasn’t yet been sanctioned by the W3C.

http://developer.apple.com/internet/webcontent/xmlhttpreq.html

I think Web Apps will be that much easier if all browsers can implement XMLHTTP (or a future W3C equivalent), allowing dynamic TX/RX of XML data/forms, received data being transformed via XSL (perhaps also client-side.)

Gmail for example makes good use of XMLHTTP on their registration page - you can check for the availability of an e-mail address without refreshing the page.

Once this “dynamic transaction” restriction is reliably overcome, I think the biggest remaining issues are mostly behavior-related (layout, CSS etc.)

19
Yannick says:
June 24, 08h

I must say Dave, this was a very interesting article. I found it most interesting. I currently work at my university in the web development department and I am seeing that web apps are being used and are definitely becoming an integral part in the way things are done on campus. And I agree with Jim that those last two sentences are what every web developer wants to hear. My eye’s have been opened even wider after reading this article. Keep up the good work.

Have a nice day.

20
Yannick says:
June 24, 08h

I must say Dave, this was a very interesting article. I found it most interesting. I currently work at my university in the web development department and I am seeing that web apps are being used and are definitely becoming an integral part in the way things are done on campus. And I agree with Jim that those last two sentences are what every web developer wants to hear. My eye’s have been opened even wider after reading this article. Keep up the good work.

Have a nice day.

21
dan says:
June 24, 09h

Great read yes!

But the entire I couldn’t get it out of my head….

“Flash is evil”

It is not :(. Everything has its place, and I think although not so accessible(for the time being), it is amazing for creating apps. Especially when you mix it with cold fusion components. Now that MX 2004 is full blown OOP, it’s just a matter of time before people realise it’s power… and all under 10kb!

I think the biggest problem with flash is as much as we are trying to seperate content and functionality, we need to remember that most of the time we need to keep the designer separate from the developer.

22
June 24, 10h

Isn’t a web application pretty much the ultimate expression of seperation of content and presentation? Content is usually stored in a database and the app itself becomes the container for presenation. For some apps even behavior is seperated in that it’s being handled by web services or other server software …

23
Adrian says:
June 24, 12h

Interesting article, and there defiantly are changes to what the web signifies afoot.

Personally I can’t wait till the web can change content on a page without changing the page itself. That’s the true power of Flex really, without that it’s really just flash and is something you can do on WebTV which runs the equivalent of version 3 browsers.

There is however a problem with the location bar being the new command line. You can’t block a command line. You can block a address bar. Webmail (the most common example of web apps) is a great example. My client (I’m a consultant, work at the client) is a major company and for ‘security’ reasons have blocked hotmail, gmail, yahoomail etc, as they fear access to non work mail means virus.

I can bring my laptop in because it has a virus checker and run any application I like. However the clients I work for can block with their firewall any application I want. until companies mature about their thinking of the web, major and big companies are still fairly old school. I’m sure the banking sector is ever worse, and blocks a lot more.

The web as an application medium is great if you have unrestricted access. But I cannot shift to web apps for the most part until I can guarantee that they wont be blocked when I arrive at a client

24
Derek says:
June 24, 12h

I like what Tim Bray had to say about a year ago:

http://www.tbray.org/ongoing/When/200x/2003/07/12/WebsThePlace

http://www.penmachine.com/journal/2003_07_01_news_archive.html#105846244509554738

“…all application interfaces (VB, win32, X Windows, Mac OS) used to be ‘richer environments,’ and the users abandoned them by the millions, in favor of the browser, the moment they got a chance (granted, the browser was an advance on previous online-service interfaces).

“I said millions and I meant millions: tens of millions, hundreds of millions of browser downloads from the Netscape that was, and the software vendors fighting the rearguard actions to defend their ‘richer,’ ‘more responsive,’ ‘higher-performance’ client software; and losing, losing.

“Hey, I cashed in on it. Open Text got to be a successful vendor of content management software largely because we were the first to do it all through the browser, with no client software. Our stuff didn’t do all that much more, but given a choice between client and browser, the people wanted the browser.

“That’s why […] flexibility and usability […] is so completely 100% wrong. Browsers are more usable because they’re _less_ flexible.

“I think I’ve told this story before here, but it’s a good one; at a content management conference, a woman from the Tandem-that-was saying ‘It was so wonderful when the browser interfaces came on; the vendors had to discard all those stupid sliders and cascaded menus and eight-way toggles, and only leave the stuff that mattered.’”

25
June 24, 12h

—– start digression ——

“Jan” wrote “I doubt XAML or Flex will take off very soon. As the articles say, being able to run anywhere, anywhen is the core feature of the web. These ‘rich webapps’ wont run everywhere until well into the next decade, if ever”

I think you may be confused on terms here… Macromedia Flex is an XML-based way to create SWF files, which the majority of consumers can already view today, on their current machines, without downloading anything new. It doesn’t require a new round of plugin downloading, much less OS installation, much purchase of new hardware to support the new OS.

Flex is just another way to create SWF, not a new type of deliverable format itself.

Regards,
John Dowdell
Macromedia Support

—– end digression ——

26
June 24, 12h

Continuing on Derek’s reiteration of what Tim Bray had to say about the web browser. The fact that the web browser has such a limited set of controls is (I feel) the major reason why it has been such a success to deliver ‘applications’.

You only have to watch an average user, and by an average user this is someone who is just using the computer as a tool to get things done (and is not interested in the computer/software itself), to them even a tree control can be an odd concept. This may sound strange to developers but go and watch average users working with computers. These are the people who will use multiple spaces in word to align columns of text as they don’t know about tabs.

This in a round about way is the crux of the problem (bear with me on this) to understand about tabs you have to be taught/learn and build a mental model of tabs and aligning text. With more ‘advanced’ controls available in traditional applications you have to learn how the control operates/build the mental model. With the limited set of controls provided by the web browser which are all very clear (limited?) the learning curve is very low and once those few controls are learnt they are applicable to all web forms on every site everywhere.

These limited controls have also forced web developers to rethink how to display information. Whereas before you might have used component 58 from your Super Developer Pack of 256 components, now you have to rethink how the data is going to be displayed and how you are going to ask the end user for their response.

So remember when you look at the controls available in a web browser you may only see limitation - your users will see instant familiarity and understanding.

27
Jan says:
June 24, 12h

Adrian: On the other hand, one of the reasons webapps have done so well is that they let you bypass (for good or ill) firewalls which block anything but port 80.

I doubt XAML or Flex will take off very soon. As the articles say, being able to run anywhere, anywhen is the core feature of the web. These “rich webapps” wont run everywhere until well into the next decade, if ever.

(Dave: when you forget to fill in an email address, you get sent to page which still uses the old design. Is this intentional? And if you hit preview on _that_ page, you get an unstyled page!)

28
david gee says:
June 24, 12h

I’ve been involved in designing and developing behind-the-firewall financial webapps for the past few years. Given the conservative culture surrounding these things, I doubt XAML is going to take off anytime within 2-3 years of being released, as it is tied to Longhorn. I doubt Flex will ever become anything but a “boutique” platform, either - it has the stigma of being associated with Flash (which makes corporate developers wet their pants in fear) as well as a VERY hefty pricetag for a platform with very little markut saturation. I’m really hoping that XUL takes off, it seems to have a lot of promise, the major thing that is lacking at this point is a “plugin” style browser install of the required libs. The little XUL documentation that does exist is 100% geared towards creating Mozilla extensions, very little attention has been given to enabling developers to create standalone XUL webapps.

As far as creating web-apps goes, this is where IE5.5+ still definitely beats Mozilla - it has basic but functional (and very easy to code) drag ‘n drop events, the .htc extensions are really cool, and the MSXML4 DOM tools are fucking AWESOME, and blow Mozilla’s XML handling out of the water. For chrissakes, it pretty much has full XPath support from Javascript. That’s powerful stuff, and it’s very very useful for rapid, iterative prototyping.

29
June 27, 05h

There seems to be three technologies emerging in this discussion:
1)Classic DHTML web applications using JavaScript/HTML /DOM/CSS, etc centered on the client; 2)Flash based web apps like Macromedia’s Flex and Central plus Laszlo Systems centerd on a rich-client based Flash Player with ActionScripting 3)Microsoft’s Longhorn XML based rich client entry modeled after Mozilla’s XUL. Whats interesting is that these are ALL client side technologies.

One other major client side technology, Java, in the form of J2ME and other trimmed down JVM players is emerging from players like Altio, Droplets and Nexaweb and they are very powerful client side alternates as they support offline AND online operations as well as being cross-platform and cross-browser.

The other cross platform technology that seems to be ignored is the huge, server-side JSP/PHP/PerlCGI community of applications that are all getting major upgrades this year. JSF-Java Server Faces, despite Tim Bray’s arguments to KISS it - does allow for much improved Web UI sophistication as appropriate. Java Studio Creator and Oracle 10g JDeveloper will probably be first to market but lots more to come. Meanwhile PHP 5 is at release candidate 3 and has added some deceptively nifty touches including much easier and faster XML parsing, client side CLI-Command Line Interetation (just needs trusted app security model) and much smoother OO implementation. Finally Perl 6 should also emerge with a much better and cross platform runtime engine - Parrot which can purportedly execute JVM and CLR code.

So if Dave is right about the up tick in Web and Application development - there will be an absolute wealth of newly empowered tools to choose from - and all of them incorporate Web Services as just another component.

30
tlack says:
June 29, 10h

Everyone I know is switching to Mozilla (or some variation of it), accelerated by the terrible IE viruses that have come out this week. This is probably a good time for the progressive developer to wrap their brain around XUL. Does anyone have a simple example of XUL in action to improve a Web app? (That Amazon example is way too weird and all the other documentation I’ve read starts too small; this is a window, this is a button..) Maybe if we rally behind XUL it will have a chance while Microsoft is asleep at the wheel w/ Internet Explorer.

31
Vincent Scordo says:
July 02, 06h

Couple of points here:

- It’s not clear the recession is over. Look at the Feds comments after each major meeting; they are very cautious about the ‘recovery’ and so should everyone else.

- Web applications built via CSS/HTML are crude and almost always very difficult to use, even with the best intentions (here I’m talking about B2E and B2B business applications and not consumer sites like Amazon or eBay). I don’t think the software developers or UI designers are at fault, it’s just a byproduct of the GUI elements and metaphors used when building web applications with this technology.

- Flex is very promising and I think the future leader in the web app UI space in terms of interaction design.

- Bottom line is that browser interoperability is still a big issue and in turn why web standards (or using CSS/XHTML) are still important. When the browser become ubiquitous, I think true interactive GUIs will emerge - as Usability and UI specialists our hands are still, unfortunately, locked.

32
Ben Nolan says:
July 02, 07h

I am surprised there are not more rich-client applications being developing in dhtml. Mozilla and IE are now similair enough - and powerful enough - that developing a desktop-alike application that works in both browsers isn’t technically difficult.

See oddpost / webfx for examples.

My contribution to the rich client world is LivePhp:

http://ripcord.co.nz/weblog/?id=314

Which is a php library trying to wrap some of the complexities of developing rich client apps.

Ben

33
Jimmy C. says:
July 12, 11h

I have a few fairly trivial questions:

First, isn’t Joel Spolsky’s article not really “seminal” if it is based on previous ideas? The definition [1] seems to imply that a seminal article generates the idea, and your description implies the contrary (i.e. echoing other conversations).

Secondly, Microsoft doesn’t really seem to be trying to position XAML for the world wide web - at least that’s the idea I get from reading the developers’ blogs. XAML seems to be just a fancy way of instantiating .NET classes. While the implementation often generates a GUI, I get the feeling that XAML is useful in other situations too. In fact, I think XAML’s non-W3C-language syntax is mainly due to the layout of objects and classes in Microsoft’s .NET frameworks like Windows Forms. Is this right? If so, then I don’t see how XAML could displace XHTML and company. Different strokes for different folks, after all.

Again, I’m sorry about the tangent. These issues just bother me a little.

[1] http://www.google.com/search?q=define%3Aseminal