The end of WIMP and the rise of Touch


I recently bought a Lenovo IdeaPad S10-3t multi-touch tablet/netbook. It was on sale for $200 off and I just couldn't resist. I figured it'd be a great way to test out the variety of new tablet OS's coming out and to compare them with my much beloved iPad. Even though I already have two other netbooks (my original HP mini, and a Nokia Booklet 3G) the ability to convert from a netpad into a tablet is key, and having a capacitive screen and an Atom N470 CPU just made it a really great device for the price. Here's an Engadget review with some video.

I should say that after using an iPad, the IdeaPad definitely feels clunky and heavy as hell. Oh, and the screen is awful at side viewing angles. You'll go blind if you tried to use thing thing in portrait mode (though it is supported). But beyond that, it's actually a slick little machine. If you're not willing to wait a few months for the next generation of cheap "netpads", I'd say the IdeaPad is a pretty good buy.

So let me tell you my experiences with using it as a tablet.

Windows 7 Tablet


First thing I did was play with the pre-installed version of Windows 7. I won't say it was fast, but it was definitely less pokey than my Nokia Booklet 3G due to the newer Atom processor. It wasn't long before I realized I wasn't really getting the full "tablet" experience though, because it came with Windows 7 Starter Edition, which doesn't have the tablet services built in!! So using Microsoft's convenient (albeit extortionate) Instant Upgrade service, I upgraded to Windows 7 Home Premium, and suddenly a bunch of settings and functionality appeared that wasn't there before - like mouse icon hiding, scrolling, flick gestures, right-clicking after hold, screen orientation, on-screen keyboard, etc.

Yet even though adding tablet functionality into the OS helped considerably, it was readily apparent that Windows Tablet functionality is really just a superficial overlay at best, more focused on pen-based computing than real touch. They've been doing pen-based stuff since the late 80s, so that's not surprising, really. ;-) Basically, despite the IdeaPad having multi-touch and a sensitive capacitive screen, using Windows with just your fingers is just painful.

This really solidified for me the idea that trying to go from a WIMP interface to a Touch based one isn't easy or straight forward. Before I write more about that, let me mention the other OSes I tried as well.

Netbook OSs: MeeGo, Ubuntu Netbook and JoliCloud


MeeGo's phone and tablet version hasn't been released yet, so I wasn't able to try that interface, but I did install the first public build of MeeGo that came out a week or so ago and it worked well! This made me very happy, as the hardware support for this first version is limited to Intel-only hardware - no NVidia, ATI or GMA500 support for video for example - which means it wouldn't run on either of my other netbooks. Also, believe it or not, MeeGo doesn't work well in VirtualBox either (unless they've done something in the past week to fix that). This means until I got it running on the IdeaPad, I hadn't actually seen MeeGo in action! Short review: It's *fast*, easy to use, and since I use Linux on a daily basis, felt very comfortable to use - like a mini version of my day-to-day computer. You can't really compare the speed and utility of MeeGo on a netbook vs. the monolith that is Windows 7 Home Premium. Though some of your favorite apps might not be there (iTunes, etc.), the basics of web, email, IM, etc. work great.

You can check out the MeeGo site for screen-shots of the OS, but what's not readily apparent is there's no traditional "Desktop" anymore and "windows" are de-emphasized (though you can still sorta see the window controls for apps that haven't been re-written, like GIMP). Instead the apps are all maximized by default, with no window frames or buttons for minimizing/resizing. These are definitely steps away from the traditional WIMP interface, and like I said, even though this version of MeeGo isn't focused on tablets, you can see the seeds of that new paradigm for working with computers there. I can't wait to see the tablet version when it's launched - I really hope they do a good job.


I also grabbed the latest versions of Ubuntu's Netbook version, and tried out JoliCloud as well. I'm not sure if there's a standard Gnome netbook UI or what, but they both looked and worked pretty much the same way. Like MeeGo, neither starts with a traditional Desktop, and they de-emphasize windowing. The rest of their GUIs works more-or-less like standard Gnome, and in fact, the new netbook dashboard can be reverted pretty easily to the standard Gnome desktop with an option switch.

I should note that none of the Linux OSs I played with supported the IdeaPad's touchscreen. I tried to install various drivers and kernel tweaks, but just couldn't get it to work, which is too bad. The reason I tried JoliCloud, actually, was because they are now supporting touch-screens out of the box, but sadly the IdeaPad wasn't one of the initial batch. Still, it was great to play with the newest netbook OSes out there just to confirm what I was originally thinking.

Goodbye WIMP!

So after playing with Windows using the touchscreen and messing with the various netbook OSes, it became pretty clear to me that using regular Windows-Icon-Mouse-Pointer based GUIs with your fingers just isn't going to happen. And if other manufacturers are going to catch up to Apple, they're going to really have to start from scratch and *re-write* their apps, because most of WIMP just doesn't translate to a tablet form factor. Lately we've seen some noise about future Windows tablets which will compete with the iPad and Ubuntu coming out with a tablet (which they've since clarified), but there's no way it can happen.

This may be pretty obvious, but let me go into detail about the various parts of the traditional WIMP interface and examine them from a touch-GUI perspective.

Pointer Icon - First is pretty easy, you can't have any sort of pointer indicator when touching a screen. Windows will let you hide it permanently - showing concentric circles where you tapped instead - but I've also seen the "autohide" simply mean to make the mouse pointer disappear when it stops moving, but appear under your finger when you touch the screen. No way - it has to be completely gone, because it's just disconcerting to have this pointer following your fingers around.

Windows - What are windows for, anyways? Well, back when managing files was the most important thing you did on your computer, the Desktop and Windows were a way of seeing various parts of your computers storage at once, and moving files between them. Once in a while it's nice to have one window up showing text, while you refer to it in another screen. When Windows launched, each window became the application itself, so managing windows was also managing multi-tasking. Since then, however, we've moved towards a more web-centric view of the world. Apps moved away from having 30 windows open, and started organizing them into tabs. Managing individual files became much less central to our use of a computer, and managing vast amounts of data took its place (this is why many of us use iTunes to manage our vast collections of music, not by manually creating hundreds of folders and copying thousands of MP3 files around by hand). So windows, in general, are losing their relevance to how we use computers, which is why we see netbooks moving away from them, and no portable devices using them, saving on both complexity and screen real-estate as well.

But most importantly, from a touch perspective, windows are almost *completely* useless. Moving, resizing, minimizing, maximizing, closing are all things that are just plain too hard to do with our big old digits in the way of the screen. On the IdeaPad specifically, the bezel at the edges of the screen made using the corner buttons (or any interface near the edges, actually) almost impossible because I literally couldn't fit my finger in the space required. But even with a perfectly flush screen, dealing with windows would be more effort than their worth. We just don't use them nowadays.

Mousing and Shortcuts - In addition to managing windows, "dragging and dropping" with a touch interface is painful, and individually selecting multiple files is impossible. Really, any sort of thing where you had to hold down a button or a key while moving the mouse just doesn't translate directly to a touch screen. (The IdeaPad actually has a left-mouse key on the side of of the screen for exactly this reason, I assume). Selecting text, scrolling, context-menus, etc. This means you need to re-think how that function works completely. Apple, for example, has added various multi-touch gestures (two fingers for scrolling certain areas, pinching for zoom, etc.) as well as touch-timeouts for functionality like selecting and copying text. An application like Photoshop, which has a million key-combos for changing the current tool, etc. is going to have to be re-implemented all together.

Menus - Traditional window menus are super useful things to have in computers, I have to say: When Microsoft got on their kick a few years ago to hide them, it drove me nuts as there's just too much functionality in modern apps to include with just buttons. Menus have names and are organized hierarchically which means you can talk someone through a problem over the phone using them (rather than "the last button on the left") which is also incredibly useful. But that said, using the Windows 7 tablet really showed me how antiquated menus are. Even with the display magnified 125%, they're tiny and hard to manage with your fingers, and if you bump up the size of the fonts more, you might as well just throw away a third of your screen real-estate (and even that doesn't really solve the aiming problem). The only real solution is to include an app's core functionality in the main interface itself (buttons, icons, drop-downs, etc.) and have a dedicated "settings" page for the rest of the functionality.

Desktops vs Dashboards - What about icons? Well, as long as they're big enough, these are obviously great for touch screens as they work just like buttons, which we're used to using in the real world anyways. But the traditional desktop is definitely going away to be replaced by the dashboard instead, filled with relevant icons, info and options. From what I've seen, most people use their desktop like this anyways - it's filled with shortcuts and important files and is the first place they look when wanting to start working on something, rather than the Start menu or Dock. Organizing the desktop is just a great idea - how many times have you seen someone's desktop that is just filled to the brim with crap - downloaded install files, icons from Adobe or Symantec or corporate image crap like VoIP icons, links to MSN or AOL dialup accounts, etc. Moving to a Dashboard like many mobile devices use nowadays is the right direction. The iOS (iPhone/iPad/iPod Touch) dashboard is most prominent example of this, as it displays the "springboard" before even the *phone* app itself, much to the users (and app developers) delight.

Moving to Touch


So I'm sure there's other elements of a normal WIMP GUI that don't work as well (scrollbars for example), but these are the major ones in my mind, and succinctly show that slapping touch capabilities on top of a desktop OS just isn't going to work. This is why the iPad has had such a huge success, yet Tablet PCs pushed by Microsoft for years have languished. In short, there's just too many fundamental issues with the Windows interface to just "tweak". It's not a matter of size, weight, power or portability of the device that matters - the core underpinnings of a WIMP-based interface is just incompatible with touch usability. Everything is going to need to be re-written from ground up - OS and apps alike. That is not just true for Microsoft, by the way, it also applies to Linux guys as well (Gnome, KDE, Xfe), and the distributions that use their code like Ubuntu, Fedora, JoliCloud, etc. Just like Microsoft, they can't just add a few changes here and there to support touch and think their old codebase will work just fine, because it won't.

Apple figured this out first - and I think others are now stumbling upon this truth as well: Google with their Android OS, HP with their Palm-developed WebOS and Nokia with MeeGo. Happily for these guys, focusing on touchscreen mobiles (rather than mostly QWERTY ones like RIM) has enabled them to get a jump ahead in what I see as the future of *all* computing - not just mobility, tablets are just the start. These guys are in a much better position to "scale up" their OS's from their small-screen roots, rather than having to pare back from full-blown desktop platforms. The rest of the platform creators out there just don't seem to have gotten it yet.

A great example: Despite pushing MIDs for the past few years (portable, Intel-based touchscreen computers), Intel's first efforts with Moblin (which became MeeGo) were all focused around a desktop UI! Combining efforts with Nokia was a perfect way for MeeGo to get some "touch DNA" added quickly, as Maemo had been focused on touch screens for years now. I really can't wait to see how it's turning out. (Note, as always, though I work for Nokia, I have no direct or inside knowledge of the MeeGo project. If I did, I wouldn't write about it.)

What's interesting is how Microsoft seems to get it in some ways - the Surface table-top touch platforms and the Kinect gaming platform - yet they've done so little when it comes to really supporting non-WIMP interfaces in general or mobile computing. They desperately need to realize that Windows' 20 year run is at its end and that a whole new way of working with computers is here. Not that Windows and WIMP is going to disappear overnight, by any stretch, but I bet you the shift is going to be huge, happen quickly and surprise a lot of people. The floodgates, I think, are opening.

What's also interesting is Google and it's two competing OSes - Chrome and Android. It seems that Android has won this internal battle and could easily do most of what Chrome is supposed to *and* work perfectly on tablets out of the box, since unlike Chrome, it's been tweaked relentlessly to mimic the iPhone and its touch interface. (Remember the first implementation of Android looked like copies of the Nokia e61, rather than the un-released iPhone.) I wonder how long they're going to pursue this two OS strategy?

Finally, the fate of WebOS still seems up in the air. HP had few choices if it wanted to launch a truly compelling and competitive tablet computer. Putting Windows on its Slate device they showed at CES wouldn't please anyone and none of the other Linux systems are really any better sans Android, which comes with its own set of issues, I'm sure. Buying WebOS was a great idea and *seems* like it'd be a great fit to scale up to a tablet from the Palm Pre size devices. But nothing has been announced, and weeks are slipping by. The only thing Mark Hurd has mentioned is putting WebOS on their printers(!?!?) and sorta/kinda/maybe saying they weren't going to kill the Palm smartphone line... yet.

It seems like a great opportunity for the Open Source folk to really get out in front of this, no? MeeGo's license is truly free, so that'd be a great start for the Gnome and KDE heads. (Especially the KDE guys, as the underpinnings of the GUI will come from Nokia's Qt division). But if not, then someone should be starting a dedicated "Tablet OS" project, as there's obviously a huge vacuum for this sort of thing.

Maybe it's because of the hardware? Despite having shown tablet PCs for the past year or so at the big gadget shows, there really hasn't been a whole bunch of touchscreen enabled PCs out there at a reasonable price. That's why I was so excited about the IdeaPad when I saw it on sale, actually. At one point, I was even considering buying a DIY touchscreen kit for my HP Mini. Since I'm *not* a hardware guy, that tells you how few options there have been lately. Maybe once some decent tablets get into the hands of developers, there'll be more understanding of what needs to be done. I hope it happens soon! Because the quicker I can use a slick non-Apple tablet device, the better!

Ok, that's enough brain dumping for this post. At some point I need to write up the qualities which I think make tablets so amazing - I just didn't feel like adding more noise to the "iPad is awesome" posts out there, so I haven't yet. Also, I feel the need to write about how I think touch based UIs are going to change the web as well - in many ways, web pages inherit many of the bad traits of their WIMP based containers - scrollbars, too-small-to-click links and other elements, hidden areas, etc. Besides just magnifying the page, there's got to be some more standard changes to be made, I think. Ok, enough...

As always, email or tweet to me if you have any thoughts or comments.



< Previous         Next >