Submitted by The BumpTop structures files and programs as 3D cards with weight that you can shuffle around and move about the desktop. You can throw a bunch of similar cards on a pile to represent stuff you’ll get to later, or stack them neatly to represent stuff you’ve already seen.
The folks who developed this are -I think- demonstrating some creative genius here. The problem is, the extension of the desktop metaphor to such a level results in massive reduciton of intuiveness, almost to the point of absurdity. And man, I am trying to be nice here.
I can’t help but be amazed by the innovative ideas presented. I just think that trying to make the computer match the destop metaphor might need some fundamental rethinking. I always lose things because when I make “tidy piles” on my “real desktop” or file things away in my “real file cabinet”, the shit becomes out of site and out of mind. Many times, I will go looking for something, only to go on a diversion as I find previously filed things that I forgot about.
I am not being negative I just think the power of the idea needs to rethink fitting in the desktop and file cabinet, stacking, etc. metaphor. If the information could be culled using verbal ques such that it appeared on the desktop in the manners illustrated, that might be one approach to using this excellent idea. And then each document could “mind map” itself to related information that is also on the drive but not necessarily part of he query, but a more indirect location or association.
Folks, I am really amazed at this. Excellent work!
It all looks pretty slick, it also looks like a nightmare to navigate if you’re disabled. Think of people who are partially sighted, there’s a lot to be said for the current system of text and clear abstracted icons vs the moving jumbled mess of bumptop. The same goes for people whith less than perfect motor control.
Oh dear… I’m already having problems with hitting the right icons… oh dear oh dear
Oh dear… I’m already having problems with hitting the right icons… oh dear oh dear
Haha, I was trying to make a serious point though – throwing away the current desktop metafor is also throwing away the last 10 years work on accessability.
If you’re having trouble hitting the right icons now, imagine someone with parkinsons. Of course research into how to actually improve usability and accessability is less sexy, not to mention less profitable, than designing a “throw icons around” playground interface.
Although it looks fancy and intuitive, this project is useless crap (I’m probably be moded down for saying this).
Documents get covered one by another? Bad
Stacking piles of documents. Yeah, like hell. Even now it was bad when two icons were covering one another.
For now desktops were cluttered, and this is just maximizing clutterability
Yep, having all on desktop, on reach like on real desk. Would be nice. But hands in real life are much more controllable than mouse pointer. Though luck on this one.
Better option would be sanely tagged file manager, which would support different views.
btw. Another project of this kind
http://macslow.thepimp.net/?page_id=18
Edited 2006-06-21 23:00
Documents get covered one by another? Bad
Stacking piles of documents. Yeah, like hell. Even now it was bad when two icons were covering one another.
You know on my real desktop I have piles and piles of documents because, you know, I’m messy and lazy. Some of my more useful stuff I’ve sorted and put into folders and drawers, it’s much easier to find a specific document that way. Maybe someone should base a computer environment around that
Edit: spell czech
Edited 2006-06-21 23:07
OSAlert does not allow users to mod opinions down just because they are expressed in a rude manner. Otherwise, your suppositions regarding such would likely be validated.
OSAlert does not allow users to mod opinions down just because they are expressed in a rude manner.
It happens regardless …
You mean a sanely tagged filemanager like gnome with leaftag for fully integrated tagging features?
http://www.chipx86.com/wiki/Leaftag
http://code.google.com/soc/gnome/appinfo.html?csaid=FDDA9F7744B44BB…
Upcoming versions of gnome will have tagging integrated directly into the filemanager and integrated with everything thanks to presence software like galago.
I love the arcade style physics added to the desktop. Some of the ideas in it are absolutely fabulous and should be integrated into the KDE 4 or GNOME 3 projects. I really liked some of the lasso ideas, but this would be much more effective using larger thumbnail images with a brevis text description tagged to the one(s) at the center of focus.
Do we really need to compare computer desktop/workbench/whatever to real things? It’s not like it’s really that hard to understand and most of the users who begin to use their computers do that by memorising things. After some time it just becomes natural.
But it reminds me of how we have been busy creating the computers in the 70’s and 80’s so we could store information in an efficient, searchable manner and now we are busy moving back again by creating desktop metaphors that inherit all the bad things about a physical desktop.
This takes some real rethinking, because there are things that don’t apply well here:
– Having thousands of files on the desktop. Where are the nested folders?
– Things can get lost. It doesn’t help that you clean up things by piling them. I know that for a fact when I just look at the piles and piles of paper right next to me. You need a quick method of searching rather than browsing (or I’m just too used to Spotlight).
– File names? Associating a name with an icon makes it infinitely easier to find among dozens of identical icons.
– Readability? The desktop is skewed into a 3D perspective. Files at the top become smaller than the ones at the bottom. This demands a pretty highres display and good eyes.
It makes for an impressive tech demo. I have to say I liked the part with photos and apps piled up, because they were at a size where you can read them clearly. But I think it needs support from many of the techniques we are used to with the normal desktop metaphor.
I think it’s very innovative, and I can imagine using it as a replacement for thumbnails while sorting through a few images or videos. However, for anything more than a dozen items it is counter intuitive. For example, there’s no way I’d like to see the contents of my Projects directory, with thousands of .cpp, .h, .obj, .tif, .exe files in it. It’s more like a cool user interface for manipulating a couple of your current documents that you’re working on. It could be used for the currently open application windows or documents, or the most recently taken pictures from your digital camera, or as a photo organizer. It’s probably OK for your desktop (mine has only a couple of icons), but I can hardly imagine it as a replacement for a full-blown file manager. I can imagine this user interface in a computer game for handling you inventory that you carry in your backpack.
I think it should have features like folders, drawers, probably shelves.
Overall I’m impressed, but I take it more like an awesome toy than a productive file manager interface.
For example, there’s no way I’d like to see the contents of my Projects directory, with thousands of .cpp, .h, .obj, .tif, .exe files in it.
That is right, but those types of files are better comparable to the nuts and bolts of a machine. They aren’t what a user usually interacts with, they are what an _engineer_ usually interacts with. Traditionally “computer user” has been almost a synonym for engineer, but these times are changing and I absolutely wouldn’t mind to use more traditional tools like terminals, IDEs or file managers to work on my software projects.
In fact, a really clear split between the two types of application (using vs. developing) might help us to create better interfaces for the common user. Because for those people computers can still be very scary and that’s only good news when you are an elitist.
Most of the people responding poorly to this don’t know what they’re talking about. User interface design is a touchy topic, mainly because of the reasoning behind it–I’m a person, so I should be able to tell a good design from bad, right? Wrong. Uninformed people are a terrible source of comments, mostly because they don’t know how the brain works. Emulating real life doesn’t just make an interface easier to use, it makes it more efficient for the user, as well. Our brains have developed over thousands and millions of years working with physical objects. Abstract thought came long after this, hence it is a less refined portion of our brain. As such, abstract processes are much harder to reason about and take more effort and time to use.
I’m a person, so I should be able to tell a good design from bad, right? Wrong. Uninformed people are a terrible source of comments, mostly because they don’t know how the brain works.
Yeah never listen to the end user, never ask them to test it or give their opinion. What do they know, right ? Hope that works out for you.
Emulating real life doesn’t just make an interface easier to use, it makes it more efficient for the user, as well.
Yeah except it doesn’t. Lasso a lot of papers in real life, do you ? What they’ve done is gone from a simple learned interface, point + click, to a contrived complex learned interface, lasso + move over icon + move mouse sideways to move through stack.
I won’t work because they’ve taken one simple action like point and click and replaced it with a miriad of complex actions all of which would have to be learned by the user.
Lasso a lot of papers in real life, do you ? What they’ve done is gone from a simple learned interface, point + click, to a contrived complex learned interface, lasso + move over icon + move mouse sideways to move through stack.
You’re right – people don’t tend to lasso a lot of papers in real life. However this is not an interface designed for mouse input – it’s designed for a pen.
When using a marker pen drawing circles around things that you’re interested in is quite natural.
Yes, some pen gestures can seem a bit fiddly and odd. It seemed though from the video as if whilst it was possible to automagically do complex actions there were simple alternatives to get to the complex actions.
Pen-based interfaces with gestures can be very easy to use and learn. Just ask any Newton user. Newton supported quite a few different gestures.
It’s true, try the mouse gestures in frogs of war, the best games really are played with one hand, er, pen:
http://fullsack.com/frogs/
Most of the people responding poorly to this don’t know what they’re talking about
Oh god. Not another one of those “you don’t know what’s good for you” self appointed “usability” people. I am so sick of that viewpoint.
Physical objects on a real desk are easy to manipulate because I have a crazy high resolution view of them (eyes) and I have two very capable hands to interact with them. Using a 2D input device (mouse) on a plane perpendicular to the 2D display device (monitor) which is showing a rendering of 3D objects that don’t behave like real objects is not even close to realistic and will only confuse users.
This interface is cool looking, but it’s a completely stupid idea for everything but the most niche uses.
What’s the single most useful thing for identifying files? The file name. In this interface, the filename has been completely de-emphathised in favour of mostly meaningless icons. The various magical mouse gestures are also not intuitive in the least.
This kind of interface only works if paired with something like a VR system. Watch the videos here http://www.movesinstitute.org/~kolsch/HandVu/HandVu.html#screenshot… to get an idea of what I mean. Not perfect, but it at least shows promise.
<blockquote>
Oh god. Not another one of those “you don’t know what’s good for you” self appointed “usability” people.
</blockquote>
It’s the nature of the field. Unfortunately, unlike most other fields, people don’t realize they don’t know what they’re talking about.
<blockquote>
Physical objects on a real desk are easy to manipulate because I have a crazy high resolution view of them (eyes) and I have two very capable hands to interact with them. Using a 2D input device (mouse) on a plane perpendicular to the 2D display device (monitor) which is showing a rendering of 3D objects that don’t behave like real objects is not even close to realistic and will only confuse users.
</blockquote>
They were using a pen interface, not a mouse.
<blockquote>
This interface is cool looking, but it’s a completely stupid idea for everything but the most niche uses.
What’s the single most useful thing for identifying files? The file name. In this interface, the filename has been completely de-emphathised in favour of mostly meaningless icons.
</blockquote>
I agree. Filenames may be less optimal than a thumbnail of the document at hand, but unfortunately displays aren’t good enough for this yet.
<blockquote>
The various magical mouse gestures are also not intuitive in the least.
</blockquote>
Every interface requires a certain amount of knowledge to get started. You could say the same about using a mouse or any other interface. In the end, I find it a good idea not to dwell on the “ease of use” aspect; I’d rather have something that’s powerful for me, but also degrades somewhat gracefully for someone like my grandmother.
<blockquote>
This kind of interface only works if paired with something like a VR system. Watch the videos here http://www.movesinstitute.org/~kolsch/HandVu/HandVu.html#screenshot….. to get an idea of what I mean. Not perfect, but it at least shows promise.
</blockquote>
Eh. I’d rather use a touch-based interface, as seen in http://www.asktog.com/starfire/starfireHome.html“>Sun .
Ugh. I’m so used to replying on http://lambda-the-ultimate.org“>LtU instead of [quote] :/
Most of the people responding poorly to this don’t know what they’re talking about
Yes they do. While not everybody can come up with a great user interface, anybody can tell you if it works for them or not. The “uninformed” people simply can’t tell you why.
Here is a different analogy. When you watch an animated film with “realistic” chacters, you don’t need to be an animator to tell good animation from bad. Sometimes the characters just seem stiff and awkward even though the animation was created using motion capture of live actors.
That “real” motion often does not look “real” to the viewer because our brains have spent millions of years learning how to walk. The mo-cap data just doesn’t always translate over well into the digital world. But you take that mo-cap data and you tweak it a little, and that can make it a bit more real.
I find the concept behind this fascinating. What I saw on the however video doesn’t look like it would make me more efficient. I would have to try it myself to be sure though. I certainly look foward to seeing how this project turns out.
That doesn’t look useful, intuitive, or helpful at all. But man it looks fun!!! I bet I could play with that for hours!
I am not a fan of new desktop environments. I use the old 2k style in XP, and despise OSX because it’s ‘different’…and don’t even like the differences between 2k/xp and KDE or Gnome.
But I actually like this and could see myself using it. Sure it would take some time to get used to it, but everything does. WHile some moan about ‘intuitiveness’, the fact is nothing is expressly intuitive when it comes to interacting with a computer. It took everyone a while to learn the desktop style of win95.
Good work and I hope they push forward and refine it.
It does use the wheel menu, which has been around about as long as there was a graphical display. It also uses Apple’s patented piles in a more 3D way, so it should get some attention, although not much of it will be positive.
For those of us who’ve been working with a graphics tablet as much as a programmer’s editor, it’s not a bad attempt to go beyond 22+ years of commercial GUI.
Sorting/grouping looked tedious. There should be a menu option to stack or pile by type. There still needs to be a organising container, even if it’s not a folder.
Still, it’s just a desktop at this point, clever, but early.
I like the interface prototype they’ve come up with. I’m just not sure they’re applying it to the right thing. File icons aren’t really well-thought-out enough to be treated as virtual objects, because files can be absolutely anything. I’d prefer a system like this for more structured objects, like, say, a media player that stores cover art and opens songs by album, or, like the pile of magazines they kept showing, text and pictures arranged into virtual booklets. That would especially benefit from the multifaceted 3D-ness of it all, and it would be a real advantage over a static icon, which is unable to convey contents with something like a PDF. I don’t know what made them think this would be a good solution to clutter, but it’s still a very innovative idea with strong multimedia potential.
I wouldn’t be surprised if Apple implemented piles/stacks in Leopard (Mac OS X 10.5), since they already experimented with them in Aperture, where they are used to group similar pictures. A short explanation of how it works (along with a virtual table interface not too different from the one linked in the article) is at http://www.apple.com/aperture/compare/
Not real 3D though, just plain old 2D
3D interfaces are fun for a while, but not terribly efficient for general usage. Here is one that I like though:
http://www.steelskies.com/coverflow/ (a 3D album browser for iTunes)
It also uses ‘piles’, only stacked horizontally
As with most 3D interfaces, not particularly efficient or full-featured, but I find it interesting nonetheless, even if just by its ability to encourage ‘exploration’.
Friend of mine is recently teaching a computer familiarisation course to home users. This sounds unbelievable, and was to him, but its true.
He started by asking the class what they wanted help with. The number one problem was files, organising files, managing files and so on. Files seemed to just vanish, or be in a heap of other files which made it impossible to find them.
He found this a bit odd, but OK, it was their class. He first explained how the Windows file structure works, Drive C, Desktop and so on. Then he passed out some screenshots of desktops he had collected – a couple were of the desktops of well organised secretaries, another of experienced computer user.
These all showed folders on the desktop clearly labelled by topic. And there were one or two documents obviously currently being worked on.
He then offered a few guidelines. Use easy to recognise English file names, use the sort headers in the file manager and so on. Use list view if there are a lot of documents. Widen the window to display the full name. Try to keep documents on the same subject together…
At this point his class, increasingly puzzled, interrupted him to say that the only things they ever had on their desktop were programme icons, and how did you put folders on the desktop anyway, and why would you, and what was Windows Explorer?
It apparently turns out that they always managed their documents through Word or Excel or whatever. They really did not understand the concept of a file manager.
So, looking at the desktop metaphor in the article, one wonders if its hitting the real problem for the naive user. The real problem is not the visual metaphor. Its that no-one has explained to these guys how to do the most basic things with files with any metaphor.
Apparently this class is moving into session 3 or 4 now, which is to be entirely devoted to using the ‘find’ command!
I’m looking for an article, I just can’t find it:
It was one guy, who also had a video demonstration on his site. He developed a semi 3D user interface. I remember his demos to show the following:
* A 3D Traffic analyzer, where you could drag the different computers, and watch the packets flow across the network
This must have been on OSAlert a couple of monts ago, maybe somewhere between Jan and April.
Nice ideas ( for the most part ) but they are very much ahead of what is technologically plausable today. Several months back there was a demo of a touch screen display that was capable of handling multiple input points. Now something like this would justify the existence of an environment like the one shown here but using a mouse would be nothing short of a torture. Also the authors should rething the file organization as it does get very cluttered when you pile things up.
I watched the video and the demonstrator kept mentioning pen pressure so it seems to me that this “desktop” is best suited/designed for tablets. I do not see it working efficiently with a mouse or mouse pad, but I do like the idea.
and fun for almost an hour but I’m VH3+ so that’s probably only around 5 minutes for the MTV generation. I don’t see anything new there, the image/icons still represent the file type or application association and have no reflection on what they are, a glance at that desk tells me more about what software he’s bought rather than what he does or is doing. Can’t tell a travel brochure from a prospectus. Maybe try thumbnails instead of icons.
I totally agree with you. As an MD and PC user I find this project very usefull. On the other hand I think it needs to implement the “classic” approach of desktop environments, meaning the use of folders, etc.
Very interesting concept, but too complex for the novice user.