Digital Web Magazine

The web professional's online magazine of choice.

Completely Rethinking the Web : Comments

By Dirk Knemeyer

May 18, 2005


Edwin van de Bospoort

May 19, 2005 1:01 AM

Thanks for your insights Dirk. Quite interesting. You handle a lot of topics in this article, I want to comment on two:
E-commerce should begin shifting away from Web applications and onto Web-powered desktop applications.
I’ve been a huge fan of the smaller ones we’ve known for long, e.g. icq, etc. These web-powered dt applications (WPDTAPP) are the most challenging ones from a design-perspective (for the record: I’m a user interface designer, educated in (physical) product design), yet, there is one thing that bothers me most from a business-point-of-view. Take E-commerce, why on earth would a user want a web-powered dt app from Amazon, Barnes and Company-X on his desktop. I think this is where the current “mental-model of webbrowsers”, the window-shopping-factor, gives users anonimity and a lower barrier to ‘shop around’. (For other aplications like online banking, I totally agree. Create a secure and fast WPDTAPP and it meets user needs much more (my opinion).)
I’m curious what your thoughts on this one are.

There could be a setting on each browser or computer to set how a user wants to experience the Web

Jarvis Addison

May 19, 2005 1:04 AM

This was a great read – very forward-thinking. Ironically, I was discussing with a colleague of mine why Apple has been so successful with their iTunes/iPod combination. Apple truly understands the user experience and it is definitely advantageous that they design the hardware AND software. Actually, this is a very distinct advantage. Companies like Real, Napster, Microsoft all think that the key to unseating Apple is by charging lower prices or by trying to out-market Apple. That approach is very short-sighted and ineffective. Since Apple’s competition only controls the software end of their solution, they cannot emulate and surpass Apple’s model. Those companies think of the user last, when in fact, they should be thinking of the user first. On top of that, the hardware that works with their software all have different interpretations of how a user should be able to listen to music. As a result, the experience is very disjointed and lacks continuity from device to device. This is something that is not currently present in Apple’s model.

I also love how the author implies that web applications are dead and I agree. Going back to Apple’s iTunes application – what Apple created is near perfect. The interface is much more rich, easy to navigate, very organized and very hard to replicate on the web. Remember That failure is a perfect example of why web applications cannot work. No matter how fast a connection a user may have, the current web browsing model is not sufficient to tackle the complexities of organizing and recreating an experience of a person buying music at a brick-and-mortar. With the web browsing model, you have to surrender much more control to network traffic, poor Javascript, horrible Java applets, a weak interface and poor navigation. That all equates to a poor experience. The client side application for accessing all the back-end content is the way to go.

I do hope that as time moves on, hardware and software integration will provide richer and more streamlined user experiences. The author is right – the way the web works now is broken. It’s time to fix it.


May 19, 2005 2:58 AM

Hi Dirk,

Thanks for sharing these thoughts with us. It was interesting to read. What captured me especially was the phrase “By installing an application on the client, companies can..”. It sounds very passive to me in regard to the user. After all, it is the user who decides what to install on his machine and not the companies.
And agreeing with Edwin above, I simply do not see people installing dt app after dt app. This is especially true in environments where people are not allowed to install applications.

True, desktop applications are much more powerful than a web browser is right now, but hey, a web browser is a desktop application as well. So instead of having companies produce their own dt apps, we should see that our existing browsers gain more functionality (based on standards).

Of course, there are certain fields where specialized dt apps make total sense, but on a general level, I do not think that turning away from the browser would benefit the user or the companies, as it simply shifts the same challenges we have in browser based design to the desktop, with the difference that desktop based design offers many more options to screw up, resulting in applications that are much more demanding to the user.

Thanks again for sharing.


May 19, 2005 5:33 AM

An interesting read, definitely, and it’s right to challenge the notion that web apps are simply better than desktop apps.

I agree with Gordon though, that we need to think of browsers as flexible desktop applications in their own right, which we’re population with functionality when the user works with a web application.

The fact that browsers offer that kind of standard framework is a huge advantage over desktop apps. Instead of worrying about applying updates and software conflicts at the user end, all that hassle is dealt with and resolved by the application designer and browser vendor – and I’m sure that’s a big plus for enterprise IT departments and end users alike.

But there are usability and accessibility issues too of course. Writing an accessible desktop app across platforms which uses interface elements people find intuitive and engaging surely requires massively more effort than using the standard framework of the browser market and writing web apps which use standard elements and behaviours, and for which screen readers and the like are available and already-distributed?

Matt Livingstone

May 19, 2005 5:35 AM

Quote: They shouldn


May 19, 2005 7:06 AM


Good article. Your enthusiasm is easily seen. Truthfully, what you’ve said about moving the web, to desktop applications has (in a crude form) happened. Everyone experiences the web through a desktop app (browser). The only difference at this point is that every website is seen through the same browser. What you’ve called for, essentially, is a better browser. Something that handles truckloads of information easily, and more customizable.

I think your user persona idea is terrific. With the advent of good CSS support, all it would truly require is seperate stylesheets.

However, there is one point you made, that I totally don’t understand/agree with: “Once devices like flash memory (portable plug-and-play memory that fits in a wallet) have the capacity to hold software applications, the notion of Web applications will reach the point of imminent obsolescence.”

You imagine that the “universal access” of web-apps can be solved by the above. While it might be, I don’t think users will take to it. Because they now need another physical device. Any web-connected desktop app could (most likely) fit on a 512 flash memory stick. I think the software world is already coming at this solution (for all applications) – the ubiquitous login, where a user can login anywhere and have his/her applications available to them (granted the homogenous environment). But carrying around the program also requires the same homogenous environment (or carrying around duplicates for each environment).

To get back on track, I must agree with the first commentor: each entity on the web has the right to choose how to display themselves. To remove that and have them provide just their information, you’ve seriously reduced the internet to a phone book. Albeit with much more information. There is no room for “experience”, the “feel” that people use the web for. Only in recent times has the internet truly given people a face that the world can interact with.

Furthermore, the very nature of the web, where everyone can publish, runs against the grain of your idea of consolidation of services (almost “specialization”). While I agree specialization is a good thing, and will provide the best in each service a healthy business, the market is free. You can’t force a defacto-standard. Ultimately this defacto-standard will come about when the industry matures. The industry of PC industrial design is fairly solidified. Looking at the history you can see some pretty wacky looking computers. Today most everything is standard. You know where to look for the CD drive, on the front of the case, you know where to look for the printer port, on the back of the case. It has (by and large) matured. (Granted the ‘alternative’ form factor cases and some of apple’s designs [the older iMac]). The same thing will happen in the web when the market decides on a standard, it will take time.

Ryan Cannon

May 19, 2005 7:19 AM

I’m sorry, I have to disagree; I find your idea quite frightening. While I agree that the iTunes music store is a novel and innovative application, it is also only available for those people with Macintosh computers running a certain version of their operating system, and much software is like this. In order to synchronize my Motorola v180, I need to use software created by Motorola (Windows 95/98/XP). My College’s networked projectors also have a desktop front-end with the same requirements. As does the desktop search programs spawning about the web.

The “one advantage” of web applications is a huge one: it makes (next to) no assumption about the user’s hardware and software configuration, requiring at most a modern browser or plug-in that is available for free download. Let me repeat that: free download.

When corporations make desktop client applications, they are forced to balance development cost for multiple platforms over the potential loss of revenue due to thumbing their nose at those 5-6% of users on other operating systems. The ability for users to have a free market for operating systems in the future is directly tied to the future of web applications, because they are not tied to a licensed OS. Competition is a good thing, and this competition will not exist if the Desktop-Web Application model that has persisted in the past continues.

Joshua Porter

May 19, 2005 7:47 AM

Interesting, wide-ranging article. I’m still digesting it all…

I have one question about your graphic, Dirk. I like the way that you made a distinction between what companies spend money on and what experiences users have, but how can you not include “content” in the “where people actually experience web sites”? Aren’t you throwing the baby out with the bathwater?

Hugo Tremblay

May 19, 2005 9:09 AM

While I might agree that, in theory, controlling the client-side environment by forcing the use of desktop apps is a great way to ensure a richer, better integrated user experience, I don

Ivan Icin

May 19, 2005 9:13 AM

I think that people don’t like to install and use so many different applications.

Probably real applications executed in browser are the solution. Currently only Firefox can do it (see Mozilla Amazon Browser), but IE7 will be able to do similar things.


May 19, 2005 9:40 AM

Great article, Dirk, and one clarification for Ryan Cannon. Ryan, iTunes is a free download for Mac and Windows.

One aspect of your article that needs further thinking is, in the case of iTunes, what is it’s primary purpose. Is it a conduit to the music store or is it a conduit to the iPod? Does it have to be both?

The problem here is that we can’t expect every company to offer a desktop client to their particular service. For another example (a bad one), I use Audible Manager on a PC to download their content to a Palm PDA. Audible is good in that they are not tied to a device, but they are horrible in that their desktop client is not integrated into their store. I have to open a browser to buy content. I’ll be cancelling once my year is up and will just buy Audible content from iTunes.

An ideal desktop client would be developed by a third party. Let’s say someone develops a desktop client to buy and manage text (ebooks, magazines, etc.). It should be able to connect to any store that had an open API and should download the content to any device with the reader installed. Revenue would probably have to generated from ads within the client application.


May 19, 2005 10:06 AM

Thanks for all the good comments so far!

Edwin asked (and was seconded by others) why people would want an application for something like Amazon on their desktop, with so many different shopping experiences in the market. There would need to be general/generic web storefronts that provide information and capture the people as customers, but once they are customers, in many cases they would benefit and invite a desktop app that worked really well. Sticking with the amazon example, I know a lot of people who do a significant amount of their shopping through amazon. While they may not want lots of different apps on their client for different web stores, I feel confident saying they’d enjoy a much more robust and usable version of amazon, ideally that integrated with other logical and related apps and products to create an integrated information experience. It is not a question of having dozens of desktop apps, it is a question of being able to make the experiences you want to have demonstrably richer, and this can best happen on the client, not through RIAs.

To Edwin’s question about people not taking advantage of customizable options on products – such as changing the side of the refrigerator door – this would be a simple and built-in part of the setup of a new computer/operating system/web browser. It would be integrated into the very basics of getting set up, so the barrier to entry would be almost non-existent.

Regarding the question of companies wanting to present their information in a different manner, and our objections to not wanting to experience a generic, cookie cutter web, the idea is that this would be a synthetic and interative dynamic. Currently, far too much of the load for making the web usable is on individual companies that do not have the resources, experience or institutional investment to truly create good experiences. Much of that load needs to be shifted to the operating system, or the browser, or desktop app(s) or…? But the model of millions of companies trying to solve the same problem separately is just so broken. The handful of broswer or OS companies should come up with a solution, not every company around the world. Its just illogical. Then, once that load begins to balance – that some degree of the presentation is shifted away from the companies and to the platforms, there can certainly be allowances to keep experiences from being too similar, which would be to the detriment of companies and people. But it is about balance.

As to the question of who would do this and why, Mozilla would be a great candidate. Talk about taking over the web: if they could launch a product that required content providers to provide far, far less programming and specification, and their content reached people in a much more customized and usable way, it would completely revolutionize what and how the web is. Every browser would need to adapt and come over to a similar model. It would completely change the paradigm. The opportunity is there for an innovative, nimble company or start-up to see this an move hard into the space. I guarantee they’d be bought out quickly, and for lots of money. In fact, if anyone wants to develop the idea, call me: we’d love to be involved.

Busy morning; more later…

Christopher Town

May 19, 2005 10:08 AM

I really couldn’t disagree more with you about the move from web to desktop clients. While there will definitely be a gradual move to a heavier and heavier client – I tend to believe it will be loaded dynamically on the users visit. I cannot imagine somebody wanting to load up separate applications for each company they are involved with (banking app, amazon, ebay, tigerdirect, apple, etc, etc). The rich internet application (RIA) paradigm will most likely persevere – but only after considerable maturation. Give it time.

Jeff Croft

May 19, 2005 11:29 AM


Great writing, great read, as per usual. Very insightful and I agree with most of what you have to say. Let me apologize in advance if some of my comments have already been said — I haven’t been able to read the whole thread yet.

Much as I agree with most of your article, I do, question the practicality of web applications becoming desktop applications. You state that there is only one advantage that web applications have over their desktop counterparts, but I think you’re forgetting about another big one: development time/resources. Creating a web application in today’s standards-friendly environment really is a “write-once, run-anywhere” ordeal. On the other hand, if, say, Amazon, was to move their store to the desktop, they’d have to write an application for nearly every platform out there — Mac, Windows, Linux, other Unix, Palm, Windows Mobile — and they’d still not reach as wide an audience.

Also, many people sit in environments where they are not allowed to install desktop applications (offices, etc.). For example, I’ve been known to shop Amazon from work once in a while — but I wouldn’t be able to if I had to install a desktop application.

In short, I think the web-powered desktop app is a great platform for building these types of things, but I’m not sure it’s the most practical. It may provide the best user experience — but only if I’ve written it for their platform and they happens to be allowed to install it on their system.

What Apple has done this way is brilliant, but most of Apple’s work is Mac-specific. They don’t really have to deal with these issues because they’re targeting one platform, and it happens to be a platform that tends to sit in less-controlled environments (homes, creative offices, etc.).

William Bardel

May 19, 2005 12:15 PM

Dirk, baby, where’s the love? Glass half full? iPod won’t play your latest U2 album? We designers and technologists can’t rest on our laurels, but the web “broken?” No way! Every day I log on to the internet it is a wonderful adventure. I buy books, book trips, read the news. I have great time, a great experience. Sure there are problems and constraints, but that’s natural with all living, breathing things. And speaking of natural and living, the web is evolving every day. Evolution is a gradual process that moves forward in fits and starts. And with every company around the world working on the same problems, more threads of possibility are being generated, and such genetic diversification yields a greater potential for innovation and good change (you want another sole provider like Microsoft designing your web experience?). I know, I know, having a few heavies like Apple and Microsoft provide stability in the OS and web market. Still you gotta love a chaotic system where possibility is great and the small guys can still make the big dogs jump. So in the mean time, sit back, relax. Still do your part to make the web better, but don’t stress with gloom and doom. Even with its glitches, the web still works for and is fun to use according to millions of people everyday (if it didn’t or wasn’t it wouldn’t be such a popular place). So get some perspective, turn that frown upside down! Life is still good! Hey, did you know our gas prices are still roughly half of those in western europe?


May 19, 2005 12:29 PM

I am wondering, Dirk, if you are thinking through the problems you write about using more information than you’re providing in your article. Given what you’ve actually written, it appears to me that what you are suggesting is not so much a solution as the relocation of a set of problems — i.e., let’s fix what’s wrong with the web by moving the problems elsewhere.

I am also concerned with the reasoning displayed in your comment, above:

Currently, far too much of the load for making the web usable is on individual companies that do not have the resources, experience or institutional investment to truly create good experiences. [...] The handful of broswer or OS companies should come up with a solution, not every company around the world.

Who is defining and controlling good experiences here? How will companies that are rarely in total agreement on even trivial issues find a solution to anything so complex? And even if there were “a solution,” in theory, what would its implementation cost and who would pay for it?


May 19, 2005 12:40 PM

It seems to be primarily designers and developers clamoring for this desktop-application future. In the meantime, tens of thousands of people have poured their photos into Flickr, or switched completely to using web-based email clients, or dumped bloated CRM-ware for Sure iTunes and some others are well-liked examples of hybrid apps, but if there’s a groundswell of demand for more desktop apps, I don’t see it.

Ryan’s point above is crucial: from the business side of things, there’s an awful lot to be gained by not having to support multiple legacy versions of an application.


May 19, 2005 2:02 PM

A number of commentors have focused on, as Steph put it, “we need to think about browsers as flexible desktop applications in their own right.” While I think this is literally true, browsers as they are currently consitituted just don’t get the job done. They are simply a platform for displaying what individuals and companies are designing. Browsers can only be part of a better solution if they dramatically change what they do and how they do it; this gets back to the ideas of set personas for different user types, or a customized browser/interface for specific human actions, such as shopping. Browsers today are trying to accommodate everything, and at an incredibly thin level of integration. There are a continuum of reasonable solutions and ways to do this – it doesn’t have to be exactly what I wrote about – but we really do need to rethink the basic conceptual model of how all this comes together.

Matt Livingstone’s point about Google as a success is very well taken. My basic thesis is that, while theirs might be a model for how to do things well in this paradigm, things would be resonantly better in a new paradigm. More, given that more powerful environment, just imagine what Google could do then!

My example of flash memory carrying applications was an intentional simplification to communicate the concept, but not literally how I see the solution occuring (a lot was happening in this article, and I tried to just show he tips of different icebergs instead of icebergs themselves). The point of flash memory was to illustrate that a lot of memory will soon be so tiny and portable that carrying it around is negligible; however, more likely the actual device will be some sort of convergence device. BlackBerry with a huge hard drive? iPod with a mega hard drive and optimized plug-and-play integration of local apps onto a remote portal? They are a few different models, all with different strengths and weaknesses, but the overarching point is that instead of our seeing portable memory as something that holds content and files, portable memory can ultimately house even larger applications. And that really begins to change the rules of engagement.

Returning briefly to the question of people not wanting to install apps onto their client, we already do install a lot of apps that we simply take for granted. Instant messaging software is a great example. Also, we think of “desktop applications” in a very defined way, typically to mean something like Photoshop that is a total hog and is cumbersome in a lot of ways. The sort of desktop applications that I’m talking about would be more along the lines of instant messaging software (lean, effective, focused). It is somewhat of a new category, but very grounded in taking advantage of the many benefits to being on the desktop. It could be a free download. The development costs could be a fraction of traditional desktop apps. Think instant messaging clients. Think Apple’s new widgets. There are lots of ways to skin this cat.

Joshua, to your question about content, the gap is in how I phrased/communicated the point (read: perhaps badly). That is to say, it is the browser and the hardware that ultimately dictate how people are experiencing. The content is part of that, certainly, but is contingent on the browser and hardware. For example, if the content/site design specifies a certain resolution or layout, and that is not supported by the actual front end, the content either cannot be seen at all or, at least, cannot be seen in the way it was intended. That is the critical point: we need to begin controlling the environments that our work is being experienced in.

Hugo, you raise some good points, but that is where I would go back to desktop solutions: amazon needs to develop a desktop client that allows their customers to have an ideal amazon experience. Web applications have SO many constraints. Things like AJAX, ultimately, are putting perfume on a pig. The additional power of the desktop, and the many years of desktop application design that we have to draw on, can create a much more robust paradigm.

Britt, why can’t most companies produce desktop apps for their products? If we think of desktop apps in a different and leaner way, the barrier to entry should not be that much greater than doing a really good web application.

Jeff Croft, your point is well taken, but if the “desktop application” is actually living on portable media that is never “installed” on a work computer (for example), wouldn’t that eliminate the objection? To your point about quicker development of web apps than desktop apps, I think that is a problem of the moment we are in, here/now/today, and could change quickly if our mindset to desktop apps changed and the same amount of time to creating and innovating in the web application space instead shifted to the desktop. (BTW, you never got back to me about that contract gig I offered you a month or two ago! )

More later.

Jeff Croft

May 19, 2005 2:25 PM


Jeff Croft, your point is well taken, but if the “desktop application” is actually living on portable media that is never “installed” on a work computer (for example), wouldn’t that eliminate the objection?

Perhaps for some people, but in many office environments people aren’t allowed to use portable media with their office computers (for fear of data theft, etc.).

To your point about quicker development of web apps than desktop apps, I think that is a problem of the moment we are in, here/now/today, and could change quickly if our mindset to desktop apps changed and the same amount of time to creating and innovating in the web application space instead shifted to the desktop.

Good point.

(BTW, you never got back to me about that contract gig I offered you a month or two ago! )

Really? Wow, my apologies. I must’ve missed an e-mail or something. Certainly let me know if there is something you still need done!


May 19, 2005 4:23 PM

William Bardel, your post made me laugh. And, certainly, I am actually a glass half full kind of guy. But it is important to help people sit up and take notice, and doing that with a softer message may not get it done. And, even though (as you point out) the world of the web is much more advanced than the world was before it, we can be doing so much better if we re-examine the track we are on and consider changing at the next station. (BTW, I love the way you visualize your resume/personal information on your website, really good stuff).

Paul, you are right implementing these solutions will be challenging, and it would be naive for me to suggest that these things can just “be done” without much complexity. That said, a small and nimble start-up with a clearly articulated vision, passionate employees, and key leaders with a lot of experience shipping products and innovating in the technology space, these things can happen. Then, once something ships and the ball starts rolling, that’s when things can really get interesting.

Andrew, our job as designers and developers is to create better solutions and show a better way. The market and businesses are not clamoring for this because they do not see it, or understand it. We need to set the vision, we need to start the coversation.

Regarding business reasons for not having to support multiple legacy versions of an application, it depends. If we are conceiving of the desktop in a more innovative way, this will be less cumbersome. My other answer is, if you are an ebusiness (amazon, ebay, orbitz, etc.) you have every business reason in the world to create the best possible experience, and integrate as tightly as possible into the lifestyle of your user base. A desktop approach provides new and immersive channels to do that through.

Jeff Croft, yes, there might be obstacles to localizing applications to work machines. That can be solved through some innovative thinking, but that won’t be by me this afternoon. :-) Regarding the contract, its passed now. Next time.

And I’m officially caught up responding to comments, yay!

Joshua Porter

May 19, 2005 6:51 PM

Dirk, I think I get what you’re saying about my earlier question: The interface is the point where experience happens and (if done poorly) dictates much of the quality of the experience. Perhaps if the interface is done right then the content defines the experience as much as the interface?

In your reply you said, “we need to begin controlling the environments that our work is being experienced in.”. If you happened to read the article that Richard MacManus and I wrote (published last week), you’ll know that our observations are that the web is heading in the opposite direction: to where designers may have little control over the interface in which their content is experienced.

Whatever happens, it will be fun to watch. :)

Dave P

May 19, 2005 8:27 PM

Wow, where to start?

First off Dirk, I would say that it takes a lot of courage to think beyond the status quo, and even more to write those thoughts down, so for that I commend you.

I think there’s a lot to be said for treading in familiar waters, and Dirk it’s obvious to this trained software developer who’s migrated into web development that a lone designer with an powerbook has strayed much too far from the shore. :-)

Let me be specific in addressing some of the points you make:

You missed/ignored two major advantages of web apps, including:

1. The same people access the same data from different locations and computers. We do not all have laptops attached to our hips. At different times, I use 4 windows machines (2 XP, 1 2000, and occasionally a 98 box), a linux box, a cell phone, a pda, and a mac. I would like to access my bank from all of these locations and more, but there’s no way in heaven or hell that I am going to take the time to install and configure desktop clients on each of these systems everytime I a) use a new datasource (eg: new store/bank) or b) change hardware.

Ocationally, I may need to use a public or friends computer – complicating matters somewhat.

2. Web Applications are cheaper to make! Outside of Appleland in the real world, makers of desktop apps need to ensure their codebases can be used on differing platforms. Many desktop apps that work on Windows 2000 wouldn’t work on XP upon its release – and both OS’s share the same core platform! I could only imagine the bank fee jump on may statement now that they need to build and test 6 versions of client banking interfaces. (or simply offer only one – ensuring I’m SOL.)

The internet was created in part to specifically solve these problems, and you want to go back to them like a love-struck fool.

As for your praise of Apple’s hardware/software synergy, let me say this:

That 5% margin isn’t going anywhere soon. Apple has success inspite of this amazingly foolish decision, not because of it. I love macs, but I also love paying $48 CDN for 512MB of ram.

As for Apple’s mysterious venture into business computing: I’ll take that bet any hour of the day. I’d love to see where apple gets without MS Office for mac. Thankfully, the brains at Apple have realized that their current model is working just fine, and it’s better not to go screwing with a good thing.

I’m also not sure what apps you’re running, but 2GB SD flash ram is available now. You could load more that a few complete versions of Linux on that. No need to wait.

Your vision of the web Dirk, with hardware, software and application conformity is close to Microsoft’s circa 2000. Keep that in mind the next time you boot up OSX.

Be careful what you wish for. Sure user experience could be improved, but is the user willing to pay such a steep price for it? I doubt it.

Adam Burmister

May 19, 2005 8:39 PM

It was an interesting read.

I think your insights into the future of web-applications and their disadvantages/advantages over client-side apps is somewhat off-track.

If you look at the recent trends in software development you’ll see that it’s the small, compact, cleanly defined, and managable products that are popular. E.g. Firefox (less bloated that competitors), OSX Tiger Widgets…
This is the power of the web application – it doesn’t require installation, but also means that it’s always up to date (no hefty downloaded update and less security problems), persistent state across browsers/platforms/location, less costs, and transferable knowledge on how to interact with the sytem. That last one might be quite a big one — web browsers behave the same on most systems (the look and feel is quite consistent), so it’s not as big a mess as applications using Java (with its varying LAF).

Ron A

May 20, 2005 6:35 AM

You make many valid points, but the fact remains that web applications are different, and at times inferior, to desktop applications. This is only going to change when one of the following happens:
1. Bandwidth and server cycles become more abundant to the point of being a non-issue. At present we are parsimonious with interactivity because too many server requests slow things down for ALL users.
2. Clients become thicker and we take advantage of the fast processors our desktops have. Use the web for serving up content only.

Right now we’re limited- and no amount of css or flash or is going to change that. We’re dealing with http requests over limited bandwidth using html and scripts and we’re trying to take flight with two feathers in each hand. It’s just not going to work much longer.

We need to learn to take advantage of local (desktop) computing power and need to create tools that harness that for large-scale, distributed applications.

William Bardel

May 20, 2005 8:51 AM

Always love getting people to chuckle. Thanks. I agree with your view that the web can be much better. As an information designer and former architect I’d love to see the boundaries pushed. For example, in form; why are websites always flat and not 3D? What happened to visions of the future ephysical web espoused by books like Snow Crash?. There’s a way to go, but it just seemed like “broken” was a drastic, gloomy characterization to something that does bring some joy to many. Even to me, as fixing problems is fun and challenging. Thanks for the kudos on my site work. Same to you too, your work is recognized.

Dick Richard

May 20, 2005 9:21 AM

Reading your article took me back to the days when I first really started leaning about computers, in the pre-DOS, pre-IBM days of CP/M and Digital and Sinclair and one of the most important computers of all times, the Radio Shack Color Computer. (Ok, be honest. How many of you folks out there cut your teeth on a Trash-80, learning to shoot down rockets in Color Basic?). For those under 30, I am talking about your parents and older siblings. You got infected from them.)

At that time, the conmputer market was a glorious anarchy, dominated by people who dreamed, not by people who had to get a job done.

The along came Big Blue, and that kid Bill, who rented them a program to try out. No one gave IBM & Microsoft permissions to set standards — except the people who bought the hardware and software. Businesses found a computer company they were used to and one they trusted to make computers which would reflect business needs.

Was it the best choice — who knows? And really, what difference does it make. IBM-compatable became the gold standard for computers, and still is. When IBM came out with OS/2, it failed for one simple reason — it wasn’t IBM compatible and the ordinary user saw no reason to change. That attempt to reset the standards almost killed the company. Was OS/2 better, then DOS or the early versions of Windows? It doesn’t matter, the average user doesn’t get care about benchmarks, technical prowess, or even security. They care about getting the job done with the least effort, with the least learning, at the lowest cost.

Today, the web reflects some of that glorious early anarchy. It is “broken”, in the sense that you can never be quite sure what a site will look at, where the controls will be, or what they will do. Some place, in some basement or garage somewhere, some young person will find the right confluence of knowledge, inspiration, timing and plain old fashioned good luck to develop a set of ideas and techniques to tame the web, just as Bill Gates and company tamed the personal computer market. (And if you think that Microsoft hasn’t tamed the world, then I ask you — when was the last time you pushed F3 for help?)


May 20, 2005 9:41 AM

1) To take away individuality and make each site the same is ridiculous. I see this never happening! What makes some sites more successful then others is careful planning, a good design and easy to use interface. If you let the user decide how to present information, you will remove the positioning and image many companies strive for. Why do you think phonebooks are at the lowest usage ever? Boring and uninspiring.

2) People will not download and install desktop applications. Think about hard drive space, plugins, operating systems. Great example – how many people actually use of the array of applications on an average cell phone? People don’t want to read help files, installation guides or for that matter figure out where the file they just downloaded went.


May 20, 2005 10:12 AM

Um, if I read your article correctly, I believe that you’ve completely ignored the realm of marketing as it relates to user experience. Standardization of experience is anathema to the entire idea of branding, marketing, and the like. I’m afraid you’re going to have to explain that part through the rose-colored glasses of your thesis before you get any real buy in from the world.


May 21, 2005 8:53 AM

A good read. Ideas for the future of some applications? Yes. Massive evolutionary adjustments toward personalized websites for the rest? Doubtful. Public image and the process of designing that image is far too big a business and far to enjoyable to just push to the side for a standardized environment.

My top two differences of opinion:

Consider the massive financial power of just some of the companies that Apple competes with:
   - Microsoft (operating system, productivity applications)
   - Dell (personal computers, servers)
   - Sony (personal electronics)
   - Google (desktop search)
   - Yahoo! (digital lifestyle management)
   - IBM (microprocessors)

Doesn’t Apple use PowerPC processors? Developed by IBM? Not really competitors now are they?

Apple understands that hardware and software are not completely disparate, but part of the same continuum and benefit from a fusion in conceptualization and design. Rather than simply make hardware bigger and better and faster and more, Apple thoughtfully creates or evolves hardware to support the lifestyle and behaviors that software creates in the digital world.

They aren’t the only ones. How about the Microsoft’s XBOX 360 or Sony’s PS3 and PSP? These devices are integrating the web with the lifestyles and behaviors of their owners and the digital world.

Ahmad Alhashemi

May 21, 2005 12:02 PM

The web as it currently exists is much easier to use than the real world. Retails shops for example use different layouts, designs and methods of handling operations. And they still hire professionals to do the same thing everyone else is doing: make good designs, layouts and methods of handling.

Allowing variety is good for the following reasons:

1. Creativity. If it was all up the makers of those unified clients (Imagine: Microsoft Web Forum Explorer), then there will be no room for creativity in such fields. All they have to do is manage the content and send it in a proper format just like everyone else.

2. Fun. Otherwise the web will be really boring. Imagine looking at something like an excel spreadsheet listing all books available in all stores in the world, instead of just browsing around Amazon.

3. Survival. If all of the web was exacly the same, a single flaw could knock the whole thing off. But when there is variety, a flaw is much more likly to affect only part of the web and the rest will survive by the law of natural selection.

4. Usability. A websites unique look can reflect many things about its goals and how things run. When you visit Google or Yahoo your immediate impression will be fast and practical. But that might not be what you are looking for. Not that these qualities are a bad thing, but think of it like entering a retail shop that looks nice but cheap, you know you will probably get the best prices, but sometimes you are looking for quality, class, or service. Design, it utilized correctly, has an important role in convieying that impression, and if all online shops looked the same it will be almost impossible to get the right impression.

I do, however, see where you are coming from. I actually felt exactly the same thing for a long time, until I realised what my dream meant, and started thinking again about it.

We do need:
1. Better markup languages that will make the content better semantically.

2. Standardized user preferences that are sent to the server as part of the HTTP request. (This is actually implemented for the users prefered language now, but we need similar headers to know the users prefered font size, bandwidth, etc).

3. We need better styling languages and browser compatability that emphasize usability. It shouldn’t be this harder to design a decent website that can correctly respond to the user changing the browsers font size setting than one that breaks this feature.

Michael Almond

May 22, 2005 6:11 PM

Nice article, Thanks. I am writing an article that shares similar themes, but approaches the “problem” quite differently (It would be nice to finish the article of course).

To summarize the main idea, technology is not serving technology; nor is design. They exist to serve human and social needs. The tools that we create to assist and help communicate the need to change. And this is a major requirement now.

This belief is just starting to catch on in the Web community, but many other disciplines are looking at and discussing this phenomena. I am glad to see that we are beginning to as well, because it leads to the creation and use of tools that better help-are needed, not just desired-by human civilization.

We are evolving, technology is reflecting that; not the other way around.

I better finish the damn article. It’s just that this concept of “working” for money in order to eat and thus “stay alive” keep interfering. I hate that.

Thanks again for you P.O.V. as well as the other comments. I learn so much from these forums.

Joshua Porter

May 23, 2005 8:53 AM

Dirk, I recently listened to a podcast on ITconversations that directly addresses the whole web app/desktop app issue you address. Here’s a link: The Platform Revolution

mark rush

May 23, 2005 8:54 AM

dont you think that by completeley unifying experience you’re killing identity… Look at blog design for example – based on the same experience, most are indistinguishable from each other and in the commercial world where the brand is king, this is is simply never going to happen. Yes there are some things that users expect to be able to do (why do design companies still create sites in popup windows?) but on some level brand experience also needs to play a massive part in a design…


May 23, 2005 10:05 AM

Joshua, yes, we are understanding each other. I did read your article, and thought it made a number of very strong observations. I think the difference is that you and Richard were showing where things are organically heading, whereas – in most of my points – I’m suggesting that things go in a different direction than they are currently going. Thanks for the link to the podcast, BTW.

Dave P., most people primarily use one or two machines. So, while in the most literal sense the fact that people do use various machines at different times seems like a barrier to entry for the desktop application direction, I think in practical terms it just means that you have a primary thick experience most of the time, and access to a thin experience other times. Your (and a variety of others’) points about the user experience are well founded, but my thesis is that there would be strong and variable user experiences. Only, the way we get to them, is by first creating optimal task environments and then creating identity and branding opportunities therein. Consider other contexts: grocery stores have shelves, aisles, displays, shopping carts. Restaurants have menus, tables, chairs, servers. Newspapers have pages, folds, are made from similar material, frequently share the same fonts. Different contexts have matured and refined to the point where – to some large degree – they are very much the same. Yet, within those standard similarities they also find room to play. At the same time, the needs of myself as a shopper compared to the needs of an 80 year old who cannot stand up out of there motorized scooter are dramatically different. The idea is that we do not currently have or understand optimal online task experiences, much less do we successfully cater to very different types of people, whose requirements and expectations for use are remarkably different. We need a better way. And while that might (short term) mean sacrificing some business requirements, identity or branding, ultimately it will provide more business success through moving people’s behaviours increasingly online and into these very low overhead task environments.

Adam Burmister, I agree with your basic analysis but think that the same things can be accomplished on the desktop side, if we take a little different approach to what software is, and how it can be optimized for the client.

Ron A, I think we see things the same way.

William Bardel, love the 3D thinking (and thanks for the lovely optimistic mindset!) Anyone seen Plumb’s Visual Thesaurus ( Great app, even experienced web people I know light up when they see it. Also, note they have both a web and desktop version, and the desktop is more robust. Additionally note that the desktop version is very “traditional” in its conceptualization and delivery, so I’m not advocating how they do it in the sense of what I’m talking about, only pointing out that they saw the value in not just producing a web app.

Dick Richard, excellent introduction and contextualization of historical elements, very insightful.

ControlZ, I think phonebooks are at the lowest usage ever b/c they are become obsolete thanks to the web: it has nothing to do with their being “boring and uninspiring”. Also, people are downloading and installing more and more desktop applications. And, if we continue to innovate what an application is, they will continue to download more. That is certainly where the future is. Web apps are just too weak.

Andy (and mark rush), haven’t ignored branding/identity at all (if you’ve read me in the past, you know that I came up through the advertising/marketing side), you just need to put the proverbial cart before the horse. A better and more usable web will brand better. It is why IA and usability have been so hot over the past five or so years, to the detriment of hotshot graphics and identity-forward solutions. People figured out if it does not work, it will not be sustainable. So we need to make things work better and better, which will in turn make our identity and branding increasingly more effective.

Chuck, you’re probably right on the microprocessors, thanks for the correction! Regarding your Microsoft and Sony examples, neither company comes to market as a true synthesis of hardware and software the way that Apple does: Microsoft is a software company trying to move into the synthesized space; Sony is a software company trying to move into the synthesized space. But of the three, only Apple both provides the hardware at top levels as well as the platform and software in the same way. Microsoft’s hardware entries are still very early, and Sony’s software – in gaming, for example – has proven notably inferior to what third party developers produce in the same genres. But your point about all moving into the same lifestyle space is well taken. Apple just happens to be a lot farther ahead, and represents something of a best practice model at the moment.

Ahmad Alhashemi, thanks for the good insights.

Michael Almond, agree about serving human and social needs, but equal to that is serving business needs. So long as we are in the capitalist system, that will be as much a driver as the “user-centered” part of what we do.

And as a bit of an aside, while I am a (generally satisfied) Apple customer, by no means do I think they are perfect. In fact, I’ve had more issues with them than I care to share. But, right now today, they really provide a lovely picture for how the synthesis of software, hardware, and really good design can provide a meaningful marketplace advantage. And if we understand the hows and whys of that dynamic, we can innovate the work that we and our companies/clients do.

Michael Almond

May 23, 2005 3:57 PM

Dk, thank you for the clarification. However, I did not mean to imply the business needs are not human and social needs. They are for sure-if you care about such things as food, shelter, and the basic requirements that keep a human alive. I am all in favor of these.

My comment simply implies the hope the all human needs will be better served, regardless of the domain; that this becomes more of a priority. While I know very little about the business sector (this is translated by Google to mean: “Nothing”), I can’t help but observe that there is room for improvement in economics matters as well (Google: we are in a shit hole).

Human needs is a broad term; I simply hope we become more of the focus in all areas pertaining to our existence on this planent._________________

Michael Almond

May 23, 2005 4:17 PM

One more comment, I promise. Read an interesting article of why the internet is stupid and should remain that way.
What the Internet Is and How to Stop Mistaking It for Something Else.

The Nutshell:

1. The Internet isn’t complicated
2. The Internet isn’t a thing. It’s an agreement.
3. The Internet is stupid.
4. Adding value to the Internet lowers its value.
5. All the Internet’s value grows on its edges.
6. Money moves to the suburbs.
7. The end of the world? Nah, the world of ends.
8. The Internet


May 25, 2005 4:31 AM

Dirk, I thought your article was interesting and thought provoking although I have to disagree with you on a number of points.

Advocating a switch back to desktop applications is not the way forward, even “web-based”. The web browser revolutionised the way we interact by providing a one-stop-shop piece of software for interacting with a world of services. We need to improve on that. The future is web services. Having one client application that has service feeds from a multitude of web based providers to give the consumer the flexibility they need. While internet connectivity may be ubiquitous (in certain parts of the world) bandwidth availablity isn’t. we need to develop a smarter model that provides the best use of available bandwidth and is able to grow as capacity increases. Most pda and cellphones or only capable of receiving text based services without severe lag. While the situation is improving in places it is not the same for everyone. Certainly not enough people to make this solution viable.

Apple are good at what they do and that is usability and interfaces. There is nothing genious about their hardward or their OS. Functionally anyone can create a very similar system on PC hardware using linux and a customised desktop. The world of Mac is small and self contained and is not the same as the larger world of the internet and the web. People critisize Microsoft but they have to design a system that interacts with a huge array of hardware and most problems arise from driver incompatabilites. Maybe if they opened up they’re system manufactures could make better drivers but it’s unlikely to happen so why bother thinking about it. Or they could go the Mac route and create Windows for only a certain architectural configuration. Then we’d really hear people complaining about monopolies.

The fact of the matter is that we do need to blur the lines between the desktop and the web while realising the technological limitations that exist. A user wants to be able to access they’re services from anywhere and using a simple system. I don’t think these systems all have to be the same however and where available should take advantage of local resources to give the user a rich experience. We need to have web-service based applications that can be scaled with regard to the limitations of the hardware platform, perhaps with an emphasis of novel ways of interacting with devices. i.e. Tilt and turn scrolling on mobile phones. No messing around with silly little buttons.

Diversity in shopping and interaction experience gives us a rich experience on the web as it does on the high street. We need that diversity to evolve standards and best practice and provide the foundation for future inovation.


May 26, 2005 10:38 AM

Michael Almond – thanks for the insight and pointers.

Rafe, it sounds like we agree to an extent, although I would suggest your mentioning bandwidth limitations as a reason advocating web applications is really justification for more desktop applications: there just isn’t enough power coming through the pipe to enable a really rich, effective, robust experience. To take it to the next level, we’ve got to come back to the client.

Mike Robinson

May 26, 2005 2:03 PM

Hi Dirk,

Frankly, I’d like to have the option to change everything about my refrigerator – as long as it continues to meet certain requirements, like keeping my food cold and not refrigerating the whole residence. I


June 3, 2005 12:21 PM

I still don’t buy the basic premise of the argument, which as I understand it is simply: browsers don’t get the whole job done.

In the absense of really compelling, buildable alternatives, it remains merely a provocation, not an idea. I’ll allow that significant user bases could be convinced to use thicker, single-purpose applications that offer a “better” experience—the eternal 1990’s dream of an “eWallet” application retains some appeal. But I still think the overwhelming preference of people is to keep their Back buttons.

Dirk, you mention specifically eBay as a company that could “integrate” better with their customers behaviors and lives with non-browser applications. In fact, they have: there’s a huge ecology of seller-centric applications, both eBay’s own and many third-party ones. There aren’t really any for buyers because eBay doesn’t make its profits there. Besides, bazillions of incoming web links to eBay items for sale probably outweighs the potential value of some more comfortable buying tool.

Indeed, you’re free to build your own contributions to eBay’s—and other companies’—application ecologies now more than ever before through the magic of open APIs. If anything, it’s possible to imagine a future in which there is no product but the service on which partners build small applications, loosely joined. Heck, I’d like to see a physical ebay device, perhaps a bar-code scanner/seller’s tool gizmo. But I still bet one of those will be a set of HTML pages for a long time to come.


June 7, 2005 12:14 PM

Nice article, but I might disagree quite strongly. There are new technologies just coming into play, such as CF MX 7 which makes building sites a whole new ballgame.

What you fail to consider is the power of what a web application can be. A centralized area in which groups can work together. Yes browsers are somewhat clumsy, but with 25% of the web browsers using Firefox, I think that we have hope there too.

Web based applications arent broken, they dont need to be fixed and they will continue to run the greater portion of business in the world for some time to come.

Ten years ago, people claimed that we wouldn’t even need PC’s. I have heard such silly talk die down.

Derek Organ

June 18, 2005 12:20 PM

I hadn’t time to read all the comments. But it seems to me regarding the point about people installing multiple applications for different suppliers could easily be solved by Google. If Google were to release a new application based loosely on what Froogle currently does then I think you


June 20, 2005 10:19 AM

Wow. I’m sorry, but that’s just about the most ridiculous idea I’ve read. Going back to desktop apps? As others have mentioned, you’ve completely misrepresented the most important advantage web apps have over desktop apps. That advantage is huge: that it no longer matters what your desktop environment is, which happens to be the root of the major problems of desktop applications, and the reason for the move to the web in the first place: development of multiple versions for multiple platforms – the challenge of getting your users to upgrade, and then maintaining backward compatibility for those that don’t.

The browser is the new platform. Developers no longer have to build and maintain more than one version of their software. Why would anyone go back to that? On purpose? That’s just crazy.

Yes, desktop apps provide a much richer environment. But for much of what people need to accomplish online, that richer environment is wasted. The fact of the matter is, for whatever its shortcomings, the browser environment is “good enough” and getting better.

Yep, you’re right – people do already install a lot of local apps… what makes you think they want to install more? It seems to me that the actual trends show quite the opposite. How many people (non techies) do you know that know the difference between Email and hotmail (or yahoo or what-have-you)? How many people, other than in their office environments, actually use a desktop app to access their email? Those numbers are dropping fast.

Browser apps are good enough to handle what most people want to do with their email, and they’re good enough to handle most other stuff too.

iTunes does rock. And it is a good demonstration of a hybrid app. But that hardly seems a good enough argument for a move back to the desktop. iTunes works for its purpose, and I think that we will see other hybrid apps, probably many, but to say that web apps will be obsolete because of them? It’s just so… ridiculous.

steven streight aka vaspers the grate

June 24, 2005 8:17 PM

Way to go, Dirk. I love articles like this.

But as the nemesis of your buddy Andrei, I have some questions.

Are you sure there are only four things people want to do on the web?

I can think of about 12 or 16 things.

What about self-expression, as in blogs?

What about social activism, as in the actions against the FEC’s attempt to politically censor blogs, or in the Technorati/Live 8/ONE campaign?

What about porn?

What about file sharing?

What about conducting business via the internet at work?

What about project collaborations, especially with behind-the-firewall blogs and wikis?

There are many items that don’t seem to fit into your “Only Four Things” scheme, though I applaud your effort to organize these events.

You also seem, like Andrei, to dismiss usability and user observation testing as an unnecessary layer, words to that effect.

We cannot assume to know what users want to do on the web, nor how they interact with our sites, unless we observe them.

Matthew Clapp

July 7, 2005 7:46 PM

I agree with your basic premise as it applies to consumers. However, one of the major reasons companies moved away from the client/server model was because of the costs involved in rolling out and maintaining client-side software. As a consumer, it’s a no brainer – you just install the new version. In many regulated corporate environments (financial services, life sciences, etc.) desktops are locked down because of laws and federal guidance that regulate the drive images that are installed on each laptop or desktop. This makes software versioning and upgrades more challenging. I like the consumer angle, but I would be more interested in how to make your proposal work inside corporate walls. Especially considering the billions just sunk into the software industry who told us all that thin client was the wave of the future.

Sorry, comments are closed.

Media Temple

via Ad Packs