Web 2.0: Is Converging Towards the Desktop Good?

Commentary

I recently had the pleasure of talking at the Ajax Experience in Boston. You can view my power-point-free slides here.

This conference was particularly exciting for Humanized because it was the first time we’ve let Enso outside the office. I think it’s fair to say that the floor required some jaw-scraping after I demonstrated Enso.

But while I met some amazing people at the Ajax Experience and had a exhilarating time, I also discovered a worrying trend: interface design on the web is slowly migrating back towards inhumane desktop paradigms. More and more, people are reimplementing windows, dialog boxes, and tree-list controls instead of brainstorming more humane solutions.

I’m going to make an odd claim: interface toolkits on the web are starting to discourage innovation. In harnessing underlying web technologies in (admittedly) ingenious ways, these toolkits make it too convenient for us to fall back onto the desktop paradigms we know, simply because we aren’t prompted by technical constraints to think of something different.

Marissa Mayer, the Google VP for User Experience, said it well:

“When people think about creativity, they think about artistic work — unbridled, unguided effort that leads to beautiful effect. But if you look deeper, you’ll find that some of the most inspiring art forms, such as haikus, sonatas, and religious paintings, are fraught with constraints. They are beautiful because creativity triumphed over the ‘rules.’ Constraints shape and focus problems and provide clear challenges to overcome. Creativity thrives best when constrained.”

In 2004, Google chose to use one nascent technology, Ajax, to create an e-mail service: since there didn’t exist any Ajax toolkits that allowed them to reduplicate the desktop on the web, they were constrained to think simply, “how can we work with Ajax and the web to make email humane?”

Their answer was something that was actually more humane than any desktop e-mail client already in existence. What’s even more interesting is that traditional desktop developers had long been able to create an email client as humane as Gmail–but they never did, because UI toolkits made it so easy to create something that was familiar, that was the same, that was inhumane.

You cannot be better without being different.

The desktop-like web toolkits being developed today endanger innovation by entrenching us in familiarity of the past. We need to remember that there is something better than the desktop in many of today’s web applications, and we need to carry this innovation with us as we move forward to create new tools and new interfaces.

Aza, I hope your presentation was well-received. It talks about important things.

So now we know that Enso is a “launcher” (providing universal access), by way of services (whatever that means), and even does semantics! I’m trembling with anticipation to learn more!

By the way, the slide technology is cool, but how is it better than a PS or PDF document, or an HTML document where the beginning of each slide is marked with a string that can be searched for?

I had some issues with the slides:
1. They took about 20 seconds to load, making my computer unresponsive during that time and offering no explanation of what’s going on
2. Switching between slides takes several seconds, eating all my processor
3. Slides can only be accessed sequentially
4. I could not get to slide 55 of 55, the presentation stays at 54.

I’m using Firefox 2.0 on Linux.

While I agree with you, there is always some risk involved in putting innovative and therefore unfamiliar interface controls in front of users. For developers who don’t have the resources to do extensive user testing or the time to provide a familar interface control alongside the innovative one, falling back on traditional controls is often the safest route. The decision can also depend on one’s perception of the user base: will your users brave something new, or will they just stare at it like a deer in headlights, afraid of doing something wrong?

Brian,

Take a look at the religion that is Quicksilver on OS X and you’ll see just how successful nontraditional interfaces can be.

Aza,

Thanks for the slides. :)

I too like Web2.0 ajaxy applications but do not understand how anyone can seriously use them in a tabbed browser. Why consider something that runs inside a browser “window” a desktop, I have on eof those for dropping files I haven’t decided what to do with yet. Web Applications should be just that, applications, and as simple as. they should run like a windowed gui application.

Good presentation and nice slide software (though it doesn’t update the current slide index for the last slide).

Interesting how sometimes everything old is new again - the idea of applications simply manipulating data is very UNIXy sounding (and even Emacs has certain functions that are very familiar…).

PS aza I believe I gave my presentation on the same weekend (different conference), but mine was much more boring ;)

Matt,

I think Aza is arguing precisely against creating web applications that mimic the desktop. And I think he is also arguing we should wean away from GUI windowed applications. See the book “The Humane Interface”.

And why should you use web applications in a tabbed browser??

An application embedded in web browser has important advantages over desktop application:
- you don’t have to deploy it.
- it upgrades automatically

There are certainly lessons to be learned from desktop apps. This includes things to avoid (indiscriminant use of hierarchies is certainly one), but there are also many things worth emulating. For example, tagging and searching are fine for a sufficing navigation of vast quantities of content, but can be highly frustrating for optimal or exhaustive navigation especially of smaller quantities of content. The visual/spatial representation and organization characteristic of the desktop is perhaps unbeatable for the latter task. A tremendous amount of research and experience has gone into the desktop UI and it would be a tragic mistake to ignore it just because we?re all so Web 2.0 now. Surely we can evolve beyond the classic desktop, but I would only favor divergences that have demonstrated human performance advantages.

I completely agree. For years I’ve been against the “make it like the desktop” mentality of web development that seems to have taken over in so many communities. Desktop applications are not the gold standard for usability; they are the gold standard for power. The traditional web interface (cursor changes when you mouse over a link, clicking takes you some place else) is a very simple, elegant paradigm that is easy to pick up. Gmail is a shiny example of evolving the classic web interface using Ajax without abandoning the paradigm that users understand. While context menus and popups are “cool”, I don’t think they add to usability and invariably they end up slowing the entire application down.

Hello All,

First, I cannot take credit for the presentation software. I simply used a great tool called S5 which I templated and slightly modified. It’s what I actually use for giving presentations so the long load time is because it was original meant only for local viewing. There is no slide 55/55, that seems to be a bug in S5.

Brian, I think you are exactly right: not everyone has the time to dedicate to making truly humane interfaces (or at least, there is a false perception that the time required isn’t worth the investment). Thus, because of deadlines and economics companies often fall back on what’s easy: prebuilt UI toolkits and old paradigms. This is why the desktop interfaces have been stagnant for 20 years and why we must be so careful in constructing our current round of toolkits. What gets built now will become the du jour for the next 20 years.

Michael, my point is that the toolkits we have built on the desktop have kept us from revisiting inhumane design patterns because they have become entrenched. The Web has it’s share of usability issues (the lack of undo almost everyone, for one) but as Nicholas says, the desktop is not the gold standard to which we should be converging blindly.

Nicholas, I completely agree with your context menu comments: they simply aren’t scalable as a means of performing operations on content. That problem, in part, what Enso solves.

Pgan-

regarding the slides and searching. I use the Opera browser. It shows all the slides as a single page (long) page at first. If you hit the fullscreen button it switches to “OperaShow” with is a presentation engine that works similar to what is being seen in IE (I didn’t check FF).

Although there were some quirks (some of the bullet fonts were grey instead of black, and therefore too hard to read), it seemed to work quite well.

…and you can use your regular “inline find” features to search however you would a regular page.

Replicating the windows (WIMP) desktop on the web is truly a massive step backwards for Web2.0.

The “desktop” is a lousy interface that was better than the command line but really matched the way very conceptual people think (read “engineers” and “product managers”). It is very poor for tasks.

It is incredibly hard to create relevance in an interface that is made up of applications. Instead, we need to strive to maintain the great part of the web that is the page, and use widgets to increase the relevance of content and actions within the frame.

I wrote on this at Rich Internet Apps - Radical Simplicity or Oppressive Richness? and Rich Internet Apps are a Bad Idea

Does the presentation display for everybody. I have tried IE 6 and Firefox 1.5 and 2.0 on different machines and network connections but just get the title graphic. All the controls and content appear and then vanish under the title. I get scripting errors. How usable is a presentation that won’t display on some configurations?

OK. Tried a third computer and stumbled into the solution. It makes me feel stupid because it’s so simple!

On the other hand, there is no visible control which surely violates a usability principle? All it needs is a note that the arrow keys are the navigation controls.

My mistake was that I expect a web browser to have controls in the page or scroll bars. I rarely, if ever, use the arrow keys for web browsing.

On the article itself, I agree.

Eddie and Francis -
the slides show differently now than when I first viewed them. Probably someone at Humanized changed them. Before, they rendered one per page. Now they render all in one page, and my browser’s status bar is the status indicator.

“Death of the Desktop”? How can it die when it is already dead, and has been dead since the 70s. THE DESKTOP NEEDS TO COME TO LIFE! :)

Oops, my bad. It does not work any differently. The slides first display as one HTML page while they are loading, but after they are loaded they display one per page.

On slide #22 of 55 on “The Death of the Desktop” presentation, there is a diagram showing three circles overlapping, one is Photoshop, one is Word, and one is Mathematica, the overlapping area is where all the applications have overlapping features repeated. While I understand the point, it is not very dramatic, SO WHAT, the applications repeat some things, I as a user am willing to put up with it, I’m getting what I need done.

A more dramatic example would be to show different websites, and how each is duplicating functionality.

Thousands and millions of dollars are spent developing websites, as if they were applications. Yet each has to spend tons of money to get the features they need. And thus we use all these social websites, like Digg.com, Myspace.com, Del.icio.us, etc. because we don’t have the ability to create such things on our own, and it gives us a community and a voice.

Lets zoom the f**k out and see everyone’s desktop. Think earth as an operating system, then these overlapping areas of repeated functionality will be a much more pressing matter.

Nobody gives a s**t about their desktop, or their operating system for that matter, because it doesn’t do anything, they care about their ipods and music, their myspace page, their email, and s**t like that, ya know?

It’s frustrating that the internet and websites are so fragmented, not our operating system and applications.

I have never used Gmail as I am worried (possibly paranoid) about Googles lack of respect for their users privacy. Please could some one briefly explain why Gmail is so good?

I hope ENSO will respect users privacy. Remember just because your paranoid doesn?t mean your wrong.

I absolutely agree with the comments here. I’ve been setting up an email system based on Scalix (an Exchange replacement) which connects to MS Outlook, but also has a web interface. The web interface is quite good for what they are trying to do, emulate the desktop experience of Outlook, but that is its let-down too.

For example, the calendars. Outlook 2000 only lets you view one at a time. This makes public calendars almost useless, as you are unlikely to want to look at your own and another. In Outlook 2003 they let you view more than one calendar side-by-side. A little better but still a little “stupid”. And so this is what the web version of the email interface does too! When I look at Google calendar now I am amazed by how intuitive and useful the multi-calendar view is. Just click on the selection next to which calendars you want to view and all appointments show in the same place, colour-coded so you know which one belongs to which calendar.

Easy, simple, and perfect for the web. It should also make those non-web “Uncles” better, I hope.

I have made a recent writing that has been triggered from the discussion, and writing on this topic, and among many others. It fits well in this context.

___________________________________

Imagine Firefox as an Operating System

When we start our computers, forget about hearing the Macintosh chime, or the Windows startup sound, the Windows startup screen or the Macintosh OSX fast loading screen. Imagine you start your computer and you see a Firefox startup screen, and imagine a new Firefox startup chime, and a loading screen. Now imagine what your Firefox desktop will look like, what your Firefox Start Menu will have, or what your Firefox task bar, or dock will have in it. Web locations will be in your task bar, or in your dock. Your operating system’s root will not be your local computer but it will be the entire internet. Your computer will be one-part of a global operating system. You will be able to choose to keep things private, and then easy choose to share them.

Imagine this: Your designing a book including the collaboration between many people. You dedicate one computer to be the host for the book to live on. Each collaborator then will in their Firefox operating system will create a connection with the book, via the host’s IP address and location, for example: http://doubleday.com/book0045/. In the Firefox operating system, they will have all of the functionality that they would have in Indesign, Illustrator, Photoshop, Word, & Excell, and be able to work on designing this book. The writer will have their own area to just focus on writing, and while they are writing, the design can watch them and to start taking some of the text and creating design variations. The writer can also see and watch the designer. The photographer can start reading the writing, and design, and then start taking photos, sharing them. In all this while they can leave notes for each other, like post-it notes, or audio messages. And even have elaborate debates to solve problems. Each collaborator can be in any area of the world, and participate in the design as if they were side by side.

Imagine Firefox as an Operating System.

Braydon, my point with that slide is that there is a strong force pushing every application into becoming a portmanteau of every feature that a user could possibly want. The same is true for web applications. We need to start thinking about how the web can sidestep the very problems you point out: How can we break down the walls of applications (web or desktop based) and allow functionality to be be shared.

One way of breaking walls is a big ass hammer. Can open-source software be this hammer? Is Microsoft, Adobe, Apple scared of open-source already?

Open-source seems to be the way to achieve this because of their core philosophy, they don’t have walls. The problem seems to be that Linux seems to be trying to duplicate the Windows XP / Vista, Mac OS X experience, rather than creating a new kind of experience.

Could the open-source community, like Ubuntu(desktop) or Firefox (web), tackle this programming challenge of creating a system that isn’t based on a Desktop, Icons, Menus, Programs paradigm, but is based on how the universe works; we have stuff, we can combine stuff to create new stuff, we can modify stuff, destroy stuff, and transfer stuff, thus there wouldn’t be applications where there could be functionality duplicated, because their simple wouldn’t be a need for it.

If that problem you mentioned of functionality being repeated in both desktop and web applications is solved it will have been a huge step forward. I’d love to see it happen. The web seems to be a great playground.

The sideshow link is down; “404 Not Found”.

Its very interesting to see where its headed, I think command line interfaces are making a comeback http://digg.com/software/The_death_of_traditional_interfaces_on_your_desktoph_and_the_Web

This girl has married.
But we have also others!
http://wikewillywoll.tripod.com/

Hello!
Your site is just amazing!
Good luck!

http://buytramadol.ning.com/