Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?

3D Window Manager 244

xmda pointed us to a website for 3Dwm which, as the name implies, strives to be a 3d window manager for X. They talk about hardware that it might be useful with, and show some screenshots. It looks very rudimentary, but its a pretty interesting thing none the less. I'm just wondering how long with have a good 3d display and 3d input device that would make this really fly. And for that matter, will flatland be better for coding anyway?
This discussion has been archived. No new comments can be posted.

3D Window Manager

Comments Filter:
  • Everyone in this thread seemingly ignores the fact that the window manager's intended platform is not a 2D computer screen, but a 3D CAVE, such as the 3D-CUBE mentioned on the homepage. Of course this idea seems stupid to all "flatland" users because you don't use X in a room where you are immersed in a 3D environment. Being able to run on Linux or any other unix is simply a by-product of using the X development environment.

    This is a very interesting development for the CAVE environment, as it allows users to interact with X windows without leaving the room and going to a terminal. A CAVE user can start from 3Dwm in the CAVE and switch to other simulations while still in the environment. There is a CAVE at Virginia Tech where I go to school that might be interested in such an application....

  • I'll all for trying new interfaces to make computing more intuitive. And I have to admit that the ability to view ordinary desktops and applications in three dimensions *is* pretty cool.

    But how can a third dimension make using a computer easier or more useful? The only thing I can think of right now is a different way to list inactive applications. Instead of listing them on a taskbar or a dock, it might be easier for people to remember the "location" of the app in virtual 3D space and go to it by "walking" there, the same way we do in real life.

    I suspect that the fundamental way we view applications will have to change before a 3D interface starts to make sense.
  • It looks like ppl are trying to get the desktop work the way as Gibson has described the Cyberspace. I doubt this will be useful until we are able to get away from 2D screens. Which, of course, will take at least a few more weeks. =)= Nifty but useless as long as "true" 3D is flattened by viewing devices.
  • I myself love the way these things look. The only concern I have about them is the amount of manual effort they seem to require for navigation. Whatever the input method, it seems to me that to do something equivalent to, say, a "drag and drop" operation in 3d space is going to require a lot more concentration and effort. As many of us know, RSI is already a big problem in the common "2d" world. I just hope it doesn't become much worse in 3d.
  • If we're going to compute in 3 dimensions, we may as well go the whole way and discard the notion of two-dimensional windows. I'm thinking in terms of, say, a cuboid with buttons on it -- sort of like, say, a gamepad, which has controls all over, and can be held with the hand. So, for instance, you can have cubes containing documents in some sort of representation. Of course, in order to represent text, you'd probably have to retain some sort of two-dimensional object.

    This could revolutionalize integrated development environments, where you could have a 3D representation of each module, and visualize how they fit in together. Also, things like database management systems (to visualize relations between fields across databases), file managers (viewing hierarchies and files within directories/folders), audio and video players (easier random access selection), presentation creators, multimedia creators, etc. would benefit from such an arrangement. I don't see how a word processor or spreadsheet program would be helped, because text is inherently two-dimensional, and for any kind of usability, one would need to keep these on a plane.

    One requirement of such a system would be to build input devices for this sort of interaction. Perhaps some sort of cameras or laser-based tracking devices in all the corners of a room tracking, say, head and arm movement? This is just going way beyond my imagination now. :-)

  • I rather think a spaceball would be a better input device than either a mouse or VR gloves (ugh). Especially since it was designed for 3d in the first place.
  • I was thinking of taking advantage of the limited viewing angle of a TFT display. Why not construct a TFT type display where the light channels were etched in such a way(and at such an angle) that each eye would see a different image of the same display.
  • A 3D interface for most 2D work is a bogus waste of CPU power and time. 3D is great if you've got to look at an exploded view, a brain tumor, or similar things. But to use 3D to browse flat webspace is stupid.
  • The only way 3Dwm's will become useful is if they display more that just 2d windows in the 3rd dimension. You've got to be giving the user more information in the 3rd dimension. eg. If you're surfing the web you could see the page you're on in the context of the page links, displayed in the 3rd dim. Or if you were writing a program perhaps subroutines could appear in added dimensions. That would be cool! (and perhaps useful)
  • Here are a few things, off the top of my head...

    Greater clarity in shared environments. For example, if you call up an app from some other server, you could see a link from the window back to that server. Or, if you are sharing an application window with someone (think something like VNC) you might be able to see which resources on the network (ie users) are connected to it.

    I could image a virtual office, where everyone's windows can be seen, and easily sent to another user. Imagine looking around and seeing another uer's applications floating around their presence in space... (of course, you'd want to be able to pick which windows actually appeared in the shared space....)

    Of course, you also have the ability to treat things on the computer as you would in real life. A background process would literally be in the background. It wouldn't distract you much, but just a single glance would tell you what it's up to, in general.

    I think 3D is just one of the elements that will make the user interfaces of tomorrow. Voice, force feedback, 3D audio, intelligent agents, interfaces that are adept at reading moods and emotions, etc. all have their place. It's all about increasing the bandwidth in and the bandwidth out. Taking advantage of natural, physical affordances, and adjusting to how the user expects the system to act, rather than forcing the user to conform to the system is what will eventually make computers effortless to use.

    We could also take advantage of natural habits to make computers easier to use. Ever notice that people look up when they try to remember something? It'd be natural for you to put your reminder databases above you in the user space. Simple things, like having alert noises from your apps use stereo so you have an idea where on your screen they came from, can add to ease of use.

    I guess a 3D window manager on a 2D screen is a long way from any of these... but it is perhaps a tentative step in the right direction.
  • Objective Reality is cool, because the utilites are 3 dimensional also. With 3dwm, you have a 3d interface for 2d applications.

    The thing that I don't like about Objective Reality as that if they are going to make 3d applications, why not make them so that they take advantage of their 3d environment? They mimic their 2d counterparts in 3d. Why does the file manager look like a window? Wouldn't it be better if it looked like a room? There has to be a better way to take advantage of the 3d environment, other than making windows look 3d.
  • A certain little girl saying "This is Unix! I *know* this!"?
  • Yeah, we could get away from the "point and click" metaphor and start catching widgets in little fishnets instead.

    It's October 6th. Where's W2K? Over the horizon again, eh?
  • by Hard_Code ( 49548 ) on Wednesday November 03, 1999 @05:12AM (#1567182)
    I think 3D can be usefull for some things. Lot's of research goes into human interfaces and factors. One thing 3D could do, for instance, is to "push back" or "fold away", applications which were being used less. Say you have 5 windows open, but only are using 2 of time most of the time...the other 3 slowly move backwards on the z-axis so they cause less clutter. Instead of having "desktops" you could simply have "areas", regions of 3D space housing applications.
  • i think one of the things really holding back some of these 3d window managers is really good, fast, full screen antialiasing. People have spent a long time developing nice smooth fonts and pretty slick little icons to make window managers look nice.

    When you translate that into a 3d window manager, you get all the aliasing effects associated. Now, once 3d cards get to the point where they can do high-res with fs-aa, we might be able to see 3d window managers be a reality.

    I'd love to see my windows in 3D with full screen antialiasing at 1280x1024 (my monitor only goes that high)

    3dfx's next part supposedly out in february-march, code named "napalm" will most likely have the fill-rate and full screen antialiasing capabilities to do some pretty good 3d window-managing.

    also, with their t-buffer's depth of field and motion blur, you could get some pretty cool windowing effects... windows out of focus could literally be "out of focus" and windows could blur as you move them. neat!

    i can't wait.
  • I have actually seen the Chalmers 3D cube at an open day. I have to say that compared to some other VR type projects I have seen before, it wasn't so impressive. However, the idea of building a window manager for such a space is quite interesting.

    It's certainly not somewhere one would want to work normally - it's no desktop computer replacement. However, imagine working on a project with a group of people - let's say an architectural project.

    Now can you see the use of being able to work in a 3D space which it feels as though you are immersed in? Sure, this is not something which would be possible with the current implementation, it would need some specific applications for working in such an environment, plus some sort of control mechanism for people to affect the environment.

    Other possible uses could include product mock-ups, collaborative work on molecular modelling, educational presentations and more.

    While the cube wasn't so interesting as a plain VR app, it is REALLY interesting as a new possibility for immersive computing.
  • by jilles ( 20976 ) on Wednesday November 03, 1999 @05:13AM (#1567185) Homepage
    I can see the point of having a 3d windowmanager. I find that 2d windowmanagers get really messy when you open more than say 4 or 5 windows. Basically you have to resort to tricks like virtual desktops or minimizing the windows since the average screen is too small to put more than a few windows next to each other.

    A 3d desktop together with some other features could solve this problem. It would for instance also be nice if you would never have to start an application explicitly. Many palm top computers already have this feature and I think it is time to introduce it to the desktop also.

    Of course just a enormous plane with flat windows sticking out of it is not very usefull. People are bad at keeping track of more than say 7 or 8 things at the same time. That's why menus are tree like structures (i.e. you don't put your zillion options in one big menu but you use multiple menus and submenus to organize your menuoptions). The same should apply to windowmanagers. I want to be able to organize my windows in a hierarchy. I also want to be able to have one window in multiple places in that hierarchy. A 3d structure could help to organize this.

    The screenshots are really cool, though I don't see the point of working in the gimp window while looking at it at an angle of 45 degrees. It is great for finding the window, though.

    A 3d windowmanager would also allow for 3d widgets, I didn't see any of those in the screenshots. Now coming up with usefull 3d widgets would really be a smart thing. Implementing them is probably not so difficult.
  • Y'all need to check out Actuality Systems [actuality-systems.com]. These guys have some stellar technology which is just on the verge of being available. I actually went to college with the founder of the company, and this summary in the online-WSJ [wsj.com] (no subscription required) of Gregg Favalora's technology is an incredible read. Among other things, he's won the BFGoodrich Collegiate Inventors Competition and the MIT $50,000 Entrepreneurship Competition. Watch this guy...and get in on his IPO. :)

  • I think designing my objects models for complex projects would benefit from 3D desktops. Not just 3D desktops though, a real 3D monitor would be quite useful.
    A full object model nearly always goes behond the boundaries of my desktop/paper which implies different abstraction levels on each page. This not so bad, but a full 3D model representation would be really nice.
  • Sure beats a talking paper clip eh? TAKE THAT, MICROSOFT!

  • It goes beyond that even. The entire use of the computer in its current state practially screams 2 dimensions. We've replaced paper, pencil, paintbrush, canvas, books, and letters written to loved ones with virtual counterparts on a computer screen....but all of those things are 2 dimensional. Even the computer display is 2-D. I don't see a 3-D window manager being useful for that precise reason. ... we don't even envision ourselves using computers for 3-D stuff really....sure there are fancy VRML mockups of virtual libraires and museums that look 3-D....but that's about it.

    Sure is cool to look at though :) Maybe they'll turn it into a screensaver. Windows has that annoying "flying windows" screensaver. It would be cool to have it do that but with the actual windows that were open on your desktop...and in 3-D...


  • ive tried and tried, but i just can't think of anything a 3d enviornment can do BETTER than a 2d enviornment. besides this just being kind of a neat thing what are the advantages of a 3d gui? Can anyone think of any? (besides being able to shoot processes with the plasma rifle i mean :) i DO like the fact that they aren't content to leave gui's as they are There have got to be ways to make a more productive enviornment... any ideas? ive got a few...

    what i would REALY like to see is a vector based gui. now that would provide some serious functional advantages...

  • Goggles?

    Best idea I've read recently (can't remember where; probably on one of the boards here) is for HUD contacts.

    Full '3D' (i.e. perspective altered for each eye) coverage of your eyes' viewable areas, with controls to make the 'background' variably translucent/transparent.

    That would be cool. (If not without its problems, e.g. power/RGB input)
  • I had a dream like this!!! I swear to God, it was exactly like this window manager. I was using RedHat, and I was visiting Slashdot and all sorts of internet sites, and Netscape just sort of floated on the screen on a black backround. It was scary as hell, too.
    Not to mention THIS isn't scary. Whoa. I'll have to tell my shrink about this : )

  • Microsoft has made mice with scroll-wheels ubiquitous in the PC world. Why not use the mouse for the two most common dimensions of movement (up-down and left-right, for a window manager) and use the scroll wheel to move forward and backward? Then, if you're stacking your apps one behind another, you'll use the mouse like normal to move around its window, and the wheel to jump from one app to another (although with that somewhat-disorienting walking-through-walls effect).


  • "However, the reason that most things are in 2D is simply because they were designed for 2D only systems."

    The reason most things are 2D is because they are in fact 2D! I'm old enough to remember a world without computers so I can assure you that the reason most data is 2D only is because they are truly 2D, and not because someone hasn't invented a 3D interface.

    Take another look at your home. Even though it is a 3D structure, by and large, everything is organized as 2D. Your family pictures are placed upon a 2D wall, your dinnerware is stacked in rows, your music CD's are stacked upright or along a shelf. Even drawers are merely 2D containers stacked atop each other. The reason for this is simple. If you have to reach around plates in your cupboard in order to get a glass, you'll end up breaking a lot of plates.
  • Yet utterly cool.
  • I was discussing this with my boss, and he said something to the effect of how thats all well and good, but they need something different. He thought that actually making a type of 3d "office" with a door, a desk, and anything else you would need in it, and used it as the GUI it would be better. It would, of course, need some type of 3d engine, Quake or what have you, but I'd like to see something like this done. Maybe someone can pick that idea up and make it into something. I just wish I could code well!
  • My question is, in terms of "navigability", would the Doom Desktop Environment be a quicker, easier, space to navigate? That is, is it easier to walk over to the web room from the desktop publishing area, or to hit ALT-TAB? Well, since Quake, the concept of teleports is a natural tool (to Quake gamers at least) you use to get somewhere fast. Pressing Alt-Tab could either a) teleport you to the room with the application that is next in the focus list or b) pop up a bunch of teleporters, each labeled with the particular app and you just walk through the teleport of your choice. I believe this would be a very natural analog to Enlightenment's focus list or MS Window's Alt-Tab and Ctrl-Tab. Indulge me as I take this a bit further. With the advent of "see through" portals in the latest incarnations of First Person Shooters, you could also provide a means to manipulate windows through the portal. So, just as I can drag and drop a window from one virtual desktop to another, I could shoot my grappling hook through a portal, yanking a window back into my current room, and optionally drop it through another portal. This may sound silly, but after hours and hours of playing Quake, I can imagine myself using mechanisms like this and eventually being able to work and navigate productively in a 3D environment. It is also my opinion that flat, 2D windows will always have a place in 3D virtual environments. In meatspace it is convienient to be able to shuffle off a pad of paper to some other area of my desk, but when I have the pad of paper in my "foreground", I tend to position it and my head in such a way that I'm looking at it head on. In a 3D environment, I anticipate that any 2D window I'm actively using will be oriented as to be flat. Additionally, I imagine much usefull overlap between applications with 2D and 3D interfaces as there currently is between CLI and GUI apps. Sometimes it might just be more efficient to type 'ps' followed by 'kill pid' than to hunt down a rogue process and shoot it to death with my phaser. Conversely, I might prefer the metaphor of shooting a monster with my phaser "set to stun" for 'kill -20' and "set to kill" for 'kill -9', especially if I don't understand the concept of PIDs and signals. The moral of the story is, 3D virtual environments offer some unique opertunities for innovation in user interfaces. IMO, First Person games are indirectly leading the way by providing concrete usability data on (successful) 3D virtual environments.
  • Take up a large portion of the viewable area, and follow the POV, so, it'd be like 'Locking' it in front of you, no matter where you turned..
  • My apologies for the poor formatting of my previous post. Despite the fact that I used valid /. HTML formatting tags, despite the fact that I did use Preview (several times), and despite the fact that I submitted my post with "HTML Formatted" selected, my post had all formatting stripped :/

    Now, excuse me as I test Extrans and see if I get any better results.

  • Having a standard bs'ing session with some of my old college buddies we talked about this sort of thing. However, I am of the firm belief that a 3 dimensional window manager is just a crutch to getting to the 4 dimensional window manager. Then the 4 dimensional window manager is a crutch to a 5 dimensional...

    What we really need to do is determine what an 'n' dimensional window manager is, what it would be and how it would work. I just can't see a 3d wm doing anything for me that I couldn't do with a 2d wm. Ah well, just my two cents worth.
  • You can't visualize 4D much less code a wm for it.

    That's a bogus statement.
    Note that a `dimension' is defined as an `axis' or `direction' or `way' along which something is measured. Most of us have likely navigated in 4+-space (meaning `four dimensions or more'), which may or may not have been graphical.

    Transcending 3-space graphically can be a rather diffcult task, but I'm rather convinced that that's so simply because of our self-limiting to the mundane n-space that we've spent our entire lives experiencing.

    I've visualised, graphically, in more than three dimensions (more than four, including time-changes), and it's really not that trying, until I think something like, `how do I reduce this to 3-space', whereupon it is reduced back into 3-space.

    As for writing code that lets the computer `visualise' 4-space..., it's been done. Code has been written that lets the computer visualise higher dimensions, and, for the most part, it's no more impressive than the computer dealing in 3-space, because these machines' computations are not inherently based in 3-space.

    The thing that I have most problems with, as far as undersanding (n>3)-space is how trigonometric functions work, but, hell, I can understand how boolean operations can work when the operands can have more than 2 (true/false) values, so it shouldn't be too bad, if I spend some time on it;)

    Whether an (n>2)-space-based environment has any use, and to which kinds of applications it really would be useful, I'll address elsewhere...
  • Oops, you're right. That is VIC running, not Rat. Sorry about that, but that text was written in a hurry long after we actually took the snapshot.

    As for the Linux and the Open Source people beating us to the 3D-dimensional desktop, well, as far as we are concerned, we are the Open Source people. 3Dwm is still under heavy development and in the initial "cathedral-like" mode before being introduced to the bazaar, but we do aim to release the source under a free license. Instead, be wary of what http://research.microsoft.com/ui/TaskGallery/defau lt.htm">the opposition is doing.

  • by rde ( 17364 )
    I've already got a 4-D window manager. I had great fun coding it; I spent most of next week on it.
  • ok.

    3D window managers are useless

    nuff said

  • i wasn't talking about literally imitating real life, i was talking about metaphors. now you're taking stereotypes of implementations and just saying they won't work because former implementations were poor.

    the point is that you don't actually productively use cubes in information processing (visual, thoughts, etc.). what the hell do you expect to do with it? it doesn't make sense. books are NOT six sided things with text on every side. cubes are idiotic. you don't read on the sides of your car, it is inefficient, wasteful, and awkward.

    i'm talking about a virtually empty environment, you HOLD the windows, you can move them around, but in realistically physicalities.

    i'm NOT TALKING ABOUT MICROSOFT BOB. you completely mistook my entire idea!!!

    i agree that replicating physical objects in virtual space is stupid. i wasn't talking about imitating it. just using the metaphor of papers. the double sidedness (and our familiarity with the ANALOGUE) and an ubiquitous stylus as an interface (pseudo-cursor?). you don't have to go any farther and i regret encouraging that you should (because it obviously came across incorrectly).

    all i'm saying is that windows don't have to be floating in midair, being all wacky and "virtual". that mindset is wishful, awe-shocked, and impractical. it'd be amusing for a while, but you wouldn't actually get any work done.
  • Of course we have a Linux port! The development machine I use at home (and yes, there is a lot of that kind of development going on) is a Linux box using Mesa [mesa3d.org] and GLUT. Still, I must say that the Onyx2 does push the polygons a tad bit faster...

    (Of course, we have yet to release any code, but I'm hoping that will change quite soon...)

  • by Enoch Root ( 57473 ) on Wednesday November 03, 1999 @04:47AM (#1567218)
    This is not useless! It's just way ahead of its time!

    And if they complete it (which I hope they'll do... Hope these images are not just GIMPed up!), it means that the moment the holographic display monitor hits the market, Linux will run it first!

    Quick, someone begin coding the drivers! :)

    "Knowledge = Power = Energy = Mass"

  • by Anonymous Coward
    Finally now you don't have to start up a copy of Doom to go hunt down a misbehaving process. Well ok kill -9 would work too, but that has no style. ;)
  • Me, I hope if goes further than the desktop idea altogether. Roll on the Metaverse from SnowCrash [slashdot.org]! :)

    Actually, I was wondering if CORBA was an appropriate vehicle to start implementing such a thing. A proper 3D environment (read: window manager level) might also be useful. Now for some networking and graphical *speed*!
  • I guess the point was to have window manager, ok a talking head would be cool, but that's not what a window manager is all about. I think it would be much cooler to know what benefits you could draw from a 3dwm. ie the windows that aren't in focus are moved 3 dm back so they aren't as large instead of using that little thingie with the different desktops and windows on it(I lost my mind and can't for the world figure out what it is called).

    If you have any other ideas about the benefits, please post them..at least they can't be patented by any evil company then. ;)

    1. Download Mesa or OpenGL
    2. Examine GL-MAME, to see how it handles projections of non-GL bitmaps.
    3. Implement the window manager in such a way as it projects onto a surface, rather than dumping basic bitmaps into a rectangle.

      AFAICT, that is -all- these people have done, in essence. Produced an OpenGL Window Manager, with some basic support for projections. That's it. So much for the "amazing new technology". But, there again, it's genuinely innovative. Yes, it's been done before with MAME, but not (AFAIK) with a window manager before.

      If the Linux folks want to beat this to being the first working, open 3D environment, just add OpenGL support to Gnome and/or KDE, and you're there. It's nothing fancy, after all.

      BTW, that's VIC they have in that third window, not RAT. VIC's the video tool, RAT's audio. Honestly! I'm not sure how far I trust people who can't tell the difference between a picture and a sound. (Well, other than for those folk who have cross-wired senses, and hear images, but that's different, and even then, they could tell the difference between RAT and VIC.)

  • But mapping the 2-D image onto a plane is a trivial solution to the problem. Pretty much useless if you ask me. (which nobody did)
  • by nanick ( 36401 ) on Wednesday November 03, 1999 @05:27AM (#1567225)
    Check out the screenshots :


    to download the demo check out the following :



  • Aren't we all using 3D window managers already? I can send windows to the bottom of a stack and view several at once. Sure seems like 3D to me.

  • to move into a 3d WM we must first lose the windows metaphor for aplications. the 2 just arent compatable.
  • A 3-D window manager is probably as useful as that hacked up DOOM that lets you kill processes with a a shotgun!
  • Ignore what I said about mwm, as the pager doesn't work that way after all, but as a virtual desktop.

  • A lot of the mouses (most of them) that you can find in a store have already a little wheel or up-down key, useful for scrolling inside a window... Well, how would you like having a mouse that displays thin rays to the X, Y and Z axes (in order to help you locate your reality in a 2D monitor). The wheel would move you front/back, and the mouse would be up/down/left/right. OK, maybe not 100% a 3D input device - But at least much more confortable than holding up your hand in a glove 8 hours a day!
  • Any hierarchy I imposed on these would be entirely arbitrary, mostly unhelpful and probably a waste of time. Granted, your use may be different, but I don't see any advantage to 3d desktops. Of course, when one with a proper 3d interface (one that ignores all prevailing 2d paradigms) comes along, I may well change my mind.

    Someone else posted somewhere about the idea of organizing work topics by 'rooms.' This appeals to my sensibility; creating more of an office instead of a desktop. However, I think that you cannot simply ignore all prevailing 2d paradigms, as you suggest, because there are certain tasks which are inherently two dimensional. Typing text comes to mind, as does graphic manipulation of 2d figures.

    I think partially abandoning the concept of windows may help. Suppose in your 3d space you have a ball which represents your word processing application/document. Navigate to it in 3d, grab it, and it switches your view to the canonical 2d representation. A key combination returns you to 3d space. Furthermore, some applications will be inherently better in 3d; CAD and some graphing functions would be best represented that way. You could grab the object in 3d space which represents those apps and have your relevant data expand out into the 3d space. A windowed interface becomes cumbersome in this case.
  • You can get along perfectly well in a 3D environment with one eye. If you couldn't, the FAA wouldn't issue pilot's licenses to people with only one eye. It is fairly easy to get a one-eye waiver from the FAA. You just have to demonstrate that you can fly the plane safely.
  • -> Hey, how does this work out? I was under the
    -> impression that the WM just handles the border,
    -> resizing, etc. of Windows in X. If
    -> that's the case, then how does 3DWM show a
    -> window from a different angle? Wouldn't it be
    -> more like a replacement for the X
    -> server?

    yup - what you see here is NOT a Window Manager the ONLy way to accomplish this is for the "so called wm" to be a proxy Xserver - that means it likely advertises an X Display on :1 and you run aps there - it acts as a full X-Server and thus gets all the client requests for drawing, input etc, and translates them. This means an extra layer of abstraction that makes things considerably slower since now all rewquests have to go through another application then later go through X. I toyed with this idea once for the sake of being able not to do 3D but to be able to lkeep miniature icons of applications (their windows) AND keep them up-to-date snapshots in E so you coudl see all programs running miniaturised - thus being able to see when something interesting happend abe be able to maximise them again... but the overhead of having to proxy ALL cals and then maintian a whole X-Server to boot ditched the idea.

    I'd like to see how far this gets - but moving 2D apps into a 3D world isnt that useful. I sytill maintain we need a whole 3D windowing system from scratch where apps create 3D objects and allocate 3D areas of space for these objects, not 2D windows of pixel data like its done now. That's where the future is for 3D - but this also hangs on the need for commonly available cheap 3D input devices too - the mouse has become the 2D input device - we ned a 3D one (powerglove or somehting) thats cheap, accurate, and everone has on - also 3D displays (stereoscopic so we can get deth cueues) are also necessary IMHO.
  • Is fill rate really that much of a consideration? I mean, you have one polygon for a window, plus whatever decorations you choose... not that much at all.
  • What I would like to see, expanding upon the CAD comment, is a 3d environment, with the typical title bar, except it would have four faces, then a cube of a window, not flat, but see through. That way, say a CAD object, would be represented in 3d space, being rendered from whatever angle you wanted to see it from. It would kill the top, front, side view that plagues many programs, IMHO. Then a person could rotate the "window" which is more a cube, or rotate around the window, take your pick, allowing a better grasp of the object. Perhaps for familiarity sake the top, front, side view could rotate along so you always see them flat, and presto! a nicer CAD program. I think, at least for the casual, untrained user of such a program, ad 3d window manager might be a boon.
  • Wow, this sounds...

    ...exactly the same, to a point. Quite frankly, if I understand the 3Dwm site correctly, they're using a modified Xfb server to render the windows, then passing a bitmap to the actual server. From a technical point of view, much the same.
  • Well, I concede that, but the point is, there are ways.

    Sadly enough, you're forgetting that there is one sector of the computer industry that has researched and gotten damn close to perfection on using 2D devices and screens to control 3D environments efficiently: the gaming industry! While I haven't played with it much, Homeworld seems to have a (kinda) nice interface.

    What about the current situation we're in? If one doesn't have a taskbar or some sort of windowlist present, one still has to "dance around" if several apps are running...I remember what a pain Win3.1 was. And don't tell me that root-window menus are ideal either...they're not. And adding a rollup-window feature is just a copout.

    True, we'll still need menus and taskbars in a 3D environment...but so what. It won't be as hard as you think. In fact, it will be easier than what you do now. Just use your imagination.

  • I would like to see some documentation. I am very interested in visualization of 5 dimensions by a 3 dimensional being. That would be a revolution of the science world. That is nobel prize type stuff. Could someone please post a link to this guys work.

    I am standing behind the theories of Einstein, Hawking, and Chin. No 3 dimesional being can visualize 4 dimensions of space.

  • A solely 3D desktop would become a serious drain on my patience, I'm afraid. While it may be *fun* to navigate through your desktop in 3D, you would run into some impediments that would prevent many users from being able to make practical use of it. Making it so you could alternate quickly between a useful 2D interface and an enjoyable 3D one I should think would be an acceptable compromise.

    I won't go into detail on where I feel it would not only be useless, but would also hamper many procedures, such as in word processing and image editing applications. The reasoning is too obvious, and I'd just be passing out hairnets to the bald, so to speak.

    I would however like to explore its uses on a topical basis, where a direct metaphor relation would be a boon, rather than a burden. Factory management, building security, and online shopping are a few fine examples where direct metaphor in a 3D workspace would be welcome. Virtual environments where people can interact with eachother could be interesting, although the capacity for misleading would be increased in parallel with the capacity for representing yourself more accurately.

    To wit, I believe that while 3D environments have their place, their value as an interface between user and desktop pc should not be grasped at simply for the sake of novelty. Perhaps in time, through further developments in advanced workspace metaphors, and the inexorable advances in 3D accelerated display adapters and such, I will be obliged to eat those words. I just don't see the prominent elements of the two dimensional desktop metaphor becoming obsolete anytime soon.

  • Virtually all data that is represented by a GUI is one or two dimensional. Ascii is one dimensional. Formatted text on a page is 2D. Source code is 2D. Most images are 2D. Sound is 2D. File system hierarchies are 2D. Most CVS trees are 2D, but there are as many 4D as 3D ones.

    Look at the objects in your home. What interfaces are 3D? None are in mine (push buttons are trivially and realistically represented in 2D, so they don't count). Your VCR and TV have 2D interfaces. It doesn't matter how wrapped it is over a 3D surface, your automobile still has a 2D interface.

    Look at history for an example of what is natural for 2D versus 3D interfaces. Writing has always been 2D, so keep it 2D. Architecture is always 3D, so 3D CAD would be very useful. Art is either 2D or 3D, and should be displayed as appropriate.

    Here's what would be bad in a 3D interface: those unreadable 2D windows in the 3Dwm screenshots, stuff that moves, in either 2D or 3D (don't give the user vertigo).

    Here's what would be good in a 3D interface: 3D widgets (cool), representing z-order by depth (naturally), representing 3D data in 3D (of course).

    What we still need for a decent 3D interface: a precise 3D input device.
  • In his defense, I'd like to point out that you could use a wavelength of light as a specifier. We perceive an approximation of visible-light wavelengths as combinations of three in particular, but they're actually separate wavelengths. Remember your elementary-school "physics" where your teacher took you outside on a sunny day with a prism... :)

    But anyway, I can see where he's coming from. If you have points at, say, 1,1,1,1,1 and 1,1,1,1,2 (assume 1=red, 2=green, 3=blue) then it'd be visible as a yellow point at 1,1,1 (assume it's a holographic display or something).
    "'Is not a quine' is not a quine" is a quine.

  • Right. The only reason we called it "3Dwm" was because people generally know what window managers are (I guess we could have called it 3Dws for "windowing system"...) and thus would be more likely to understand what 3Dwm was all about. And yes, it is indeed a proxy Xserver (a hacked version of Xvfb) which communicates with the actual 3Dwm application using shared memory.

    As for the relevance of simply displaying plain old 2D applications in 3D, again, you are right. But 3Dwm is really a prototype for im3D, the Immersion3D User Environment (no homepage yet, but there are some docs on the 3Dwm website [chalmers.se]), which will be a full 3D windowing system but which will still retain support for running normal X applications, so hopefully, things will improve here...

  • I just thought of this. We have 3d desktops. If you move one window in front of another window the moved window lands on top of the original. Hence 3d. Its incredibly rudimentary, but by definition its 3d. Just think about it. Just cause you can't look at it from an angle doesn't mean there aren't different planes of windows. All 3dwm does is look nice, and let you view your windows at an angle. This won't be truly useful until we have 3d displays.
  • you could get some pretty cool windowing effects... windows out of focus could literally be "out of focus" and windows could blur as you move them. neat!

    Yes, that would look kinda cool... but how practical is it? Look at windoz' attempts at an active desktop... it looked cool (at least to some people) on paper and in proposal, but in reality it sucks. It would look cool to have out of focus windows blurred (which we can already do without 3d, btw) but it would be a major pain in the ass when you're trying to compare things in separate windows. Same goes for motion blur.

    I do think that 3d worlds have enormous potential to enhance computing... but let's be realistic. It must serve a purpose, because 'neat!' gets old really fast.

  • It has to do more than that. It actually has to change the way each window displays itself on the screen. The Window Manager is not in charge of displaying a window on the screen, just managing it -- so this WM would either have to include it's own X server or somehow hook into an existing server. Otherwise it couldn't perform the angular transforms on the contents of windows.


    "You can't shake the Devil's hand and say you're only kidding."

  • The other issue is that people visualize 2d space very easily. The degree to which people can internally represent 3 dimensions depends on the individual, of course, but typically it's a lot lower. 4d's even worse. :)
  • There was an article a few months ago on /. on a holographic LCD display which did that sort of thing. Probably easy enough to find with the search form, but I'm too lazy right now :)
    "'Is not a quine' is not a quine" is a quine.
  • Bah, the T-buffer is an overmarketed and underpowered accumulation buffer. Other 3D cards have had this capability for quite some time, but nobody uses it because it's such a killer on your framerate.

    You don't need artificially-induced focusing effects. The fact that the only part of your eye in-focus is the fovea is blurring enough.

    As far as antialiasing: that's not really necessary either, except along the edges of objects. Within objects, mipmapping and bilinear/trilinear filtering take care of the other bits. (Mipmapping actually does a lot more for aliasing effects than filtering, btw.)

    Motion blur is a sucky effect, too. It'd just eyecandy which would only get in the way in the case of a windowing system. THen again, Enlightenment has all that stuff for annoying (and slow) windows-flying-everywhere and the like which just makes your system slow.
    "'Is not a quine' is not a quine" is a quine.

  • God, I already see *enough* of Netscape wherever I turn... ;)
    "'Is not a quine' is not a quine" is a quine.
  • Actually my friend is working on a sun, a pretty old one too.
  • Basically you have to resort to tricks like virtual desktops
    I don't consider virtual desktops to be a 'trick'. I did as a windows user, but with linux (perhaps simply because I've gotten used to them), I find them indispensible and ridiculously easy to use.

    I want to be able to organize my windows in a hierarchy
    Hrm. I'm currently using four different desktops for four discrete tasks; there are multiple windows on some of these but most of them have only a single window. Any heiararchy I imposed on these would be entirely arbitrary, mostly unhelpful and probably a waste of time. Granted, your use may be different, but I don't see any advantage to 3d desktops. Of course, when one with a proper 3d interface (one that ignores all prevailing 2d paradigms) comes along, I may well change my mind.
  • Wow, someone has actually out-eyecandied Enlightenment. ;-P

    I wonder if there would be any plans in the future for this to also support some of the rudimentary 3D glasses that are out there, and some sort of 'glove' interface. I would simply LOVE to be able to just 'grab' a netscape window and move it somewhere in VR space, grab another window, make it full screen, etc..etc..
  • Don't expect much out of this until we improve 3d interfaces, both input and output. Once motion tracking and holographic technology is improved enough to be usable, we'll be able to move our floting windows around on our literal desktop by physically grabbing them. Transparent computing at it's best - floating in midair :)
  • I think my favorite feature of this would be to push a window away from me for a minute ... making it smaller but still visible, etc. It would also be nice if there were clear definitions in window creation of peer windows vs. modal child windows because a child could cause the parent to be "pushed away" a little, making "stacking" obvious. Peer windows, OTOH could sit on top of another window normally.

    - Michael T. Babcock <homepage [linuxsupportline.com]>
  • As long as you're restricting CORBA to the relatively high-level stuff ("draw a button") rather than putting it in front of the equivalent of XLib ("draw a pixel here, paint it red"). The project to watch for this approach is probably Berlin; they already support the notion of CORBA communication between client and server, where the "toolkit" code is effectively on the server. It's a small step from there to writing a server that creates 3d objects, with no application rewriting necessary (unless your app wanted to manipulate things in 3d itself).
  • "I find them indispensible and ridiculously easy to use"

    Sure its much better than without windowmanager. But nevertheless it's a solution to the problem that you can't keep track of 20 or so windows in one screen. Basically the solution consists of hiding the windows you don't need at the moment. I know one guy who routinely has 30 to 40 xterms spread all over his 16 or so virtual screens. Putting that in one 2d screen would be rediculous. Spreading them over 16 desktops only works if you can remember where to look for each window.

    Most computer users also open and close applications. Suppose you wouldn't have to do that. A good linux pc doesn't need rebooting so the only reason to close an application is to save on memory resources. Suppose you could automate this. I.e. if memory is needed, apps that are not in use are swapped to your HD. When you need them again they are swapped in (of course this would be nonsense for commandline apps). If you take into account that one app may have several windows, the number of windows in a system increases quite a bit so you need a good way to organize them.

    Of course you can argue that you don't want to see them all at the same time. But you still have the problem that from time to time you are looking for a particular window. 3d workspaces may help in finding your window faster.
  • by SurfsUp ( 11523 ) on Wednesday November 03, 1999 @06:47AM (#1567317)
    None of this is going to be too useful until we get some good input and output devices to use with it. Think about it, how well does a mouse that moves around in 2d work in a 3d environment?

    It works just fine. I know, because I have done it, quite a lot. BTW, don't think of my remarks as criticism - I'm just addressing some common misconceptions about 3D interfaces.

    Same with your monitor, it is very limiting in a 3d environment.

    Think of your eyes. Close one eye. You can still function perfectly well in a 3D world, right? So, no, a 2D monitor is not limiting in a 3d environment. Your eyes present a 2D map of a 3D world to your brain, the same as a 2D montitor does. The actual limitation is good motion and viewpoint control - this is a software issue much more than a hardware issue.

    What we need are gloves and goggles.

    Bzzzt. No. Have you ever tried it? You look like a space alien for one thing, the goggles shift around on your head, your hands get sweaty, and your arms give out after a few minutes. To convince yourself of this, try holding your arm straight out in front of you for 5 minutes. Your fingers are also a lot less accurate as positioning devices then you might think, especially without kick-ass force feedback, which won't be out of the labs any time soon.

    Then this will become somewhat usefull.

    It's going to become useful even without those things. Again, I'm not speculating because I've been using this technology for a few years. Once you've used it you'll find good ol' 2D desktops as restrictive and uninteresting as a green 25x80 text mode display. Probably.

    Think of all the desk space you could have. All you have to do is turn your head to a blank area. Don't like where an application is? Grab it with the glove and move it.

    Yes, correct. These are reasons why 3D desktops are good. Also consider: no more scroll bars - you just move your point of view instead. For graphs of functions, no need to pick a scale or limits - you just move closer if you want to zoom in, and the graph goes off to infinity if it wants to. For 3D graphs, no need to pick the viewpoint - you pick your own viewpoint, and fly around if you have to, to see the details. No need for zoom in general, for anything, you just move closer and further away.

    There aren't really any disadvantages to a 3D interface that won't be solved in time, and by this I mean not very much time, which you might suspect from the screenshots you're looking at. To prove this to yourself, consider that any 3D interface can be turned into a standard 2D interface, e.g., by pointing you directly at a 2D rectangle containing your screen view at a distance that maps texels one-to-one to pixels, and keeping you from moving or turning.
  • "[The mouse] works just fine. I know, because I have done it, quite a lot."

    Could you give some more info on that? I use my mouse to play q3test a lot, but it doesn't have any depth to it. I just point it in a particular direction and I can hit whatever happens to be closest to me in the line of sight. How could I select a 'window' that sits behind another 'window" in my line of sight.

    "Think of your eyes. Close one eye. You can still function perfectly well in a 3D world, right? So, no, a 2D monitor is not limiting in a 3d environment."

    Function perfectly well? Um, no not really. Try doing that and judging how far away that football that is flying toward you is. And how fast it is coming. But that is not really that big of a problem. The problem with the monitor is the narrow field of view. Trying to view a 3d world through a monitor is like having your vision limited to a small square about a foot in front of your face.

    I do agree, however, that these problems will be solved over time. And I certainly look forward to my gloves and goggles. :)
  • It looks very cool, but before it moves from being cool to actually being useful, I think the whole GUI metaphor would need to change to something that makes intelligent use of 3 dimensions. Then we could produce apps which used that metaphor and the 3D window manager would be more than just eye-candy.
  • Sounds too much like the movie "Hackers". So, when we're bored, we can pull up a file manager, and spend 15 minutes trying to find X configuration files? Sounds... interesting.

    Well, I suppose it would be fun to have. As long as I can make my X session look like something else. Like an old castle, with each room being an icon. Yeah yeah.
  • I remember reading about Synapse [oreality.com] on Slashdot before, and it looks somewhat better. I haven't used either of these, though. Still, on the 3Dwm page, it looked like all of the screenshots were just 2-D screenshots horribly distorted to be viewed at an angle. These weren't 3-D applications. Personally, 3-D window managers won't be ready for prime time until there are plenty of decent 3-D applications that people can use.
  • I remember this game on the Amiga, later ported to other systems, but not with the same quality. In any case, the initial sequence of the movie showed a guy using a '3D desktop', which was done using a holographic cube.

    This looks very cool. But I do agree, the current computer paradime is very hard to put into a '3d mode' for nearly all applications.

  • by jandrese ( 485 ) <kensama@vt.edu> on Wednesday November 03, 1999 @10:33AM (#1567334) Homepage Journal
    Of course you run into nameing problems.
    4Dwm is what SGI uses as the window manager on its Irix systems, fortunatly it is not a true 4D windowmanager, or I'd have to replace my brain with one from a 4 dimentional being (and they're hard to find!) just to use it. :)
  • by Foogle ( 35117 ) on Wednesday November 03, 1999 @04:51AM (#1567335) Homepage
    Hey, how does this work out? I was under the impression that the WM just handles the border, resizing, etc. of Windows in X. If that's the case, then how does 3DWM show a window from a different angle? Wouldn't it be more like a replacement for the X server?


    "You can't shake the Devil's hand and say you're only kidding."

  • Cool- this almost exactly duplicates some funky graphics I did long long ago on the Apple IIgs. There was some sort of program that let you 'perspective' and composite bitmaps. Using it to invent weird interfaces seemed like a glimpse of a weird future (don't think I'd read William Gibson by then, but that was the gist of it really)
    Now somebody is making it actually happen. I say, cool :) no way will I want to use it, but come on, just as an art project this rules. Let's see it with Enlightenment, with windows that are 'panes' with the window borders and widgets being actual 3d shapes. _that_ would be very cool.
    As for useful 3d, I'm with the person who observed that layering of windows was like 3d. I'd also note that all my favorite xterms have some feature for changing font size- and since they are terms and assume a fixed size, they scale all over and take up more or less of the screen, and THAT is where real useful 3d effects will come in, and THAT is something you can't usefully do on Windows or Mac or indeed with GUI-toolkit Linux, and the only thing it's really missing is the presence of fonts like Flyspeck 4 or whatever, to allow 'distant' xterms to be made to look really distant. I'd suggest Flyspeck 1. Each letter is one pixel :) naturally, you can't read a thing, but who here _hasn't_ put a window with some laggy but active content like an idly IRC channel, into the background, and kept half an eye on it for signs of _movement_? Also, if you had some system log with colored text for messages, and it was in Flyspeck 1, you'd say 'Wait a sec, that's a red line! Better look at what that sentence says'. All this can be done _now_ with normal, existing X terms and WMs. My suggestion would be Window Maker with titlebar and resizebar turned off so the 'window' was just a simple rect- ideally you could have a WM that drew the titlebar to scale with the resizing xterm, perhaps fading it into the background by assigning lower contrast text and background colors to give a sense of distance.
    In this way you would have a fully 3Desque _CLI_ environment with vastly greater capability to display text on the screen, at anything from large readable letters to tiny semi-obscured letters to greeked text where the information is in the motion of the blocks of text, or in the color of the greeked text. And it's all available now and needs only a bit of integration to get together. :)
    And now I've told you (almost) all of my secrets- for I'm plotting a window manager _too_ ;) so wish me luck. My goal is to make a totally, relentlessly CLI-esque environment that's pretty enough to compare with E (mostly through background pictures and having the actual text windows blend attractively), can contain orders of magnitude more information than your usual X desktop, and is commonsense enough to compare with a Mac regarding the moving and adjusting of windows.
    Wish me luck, and I hope to have concept pictures up pretty soon! The really rough part will be writing the window manager, though I am quite willing to restrict everything to _only_ color xtermland and not make any provisions for supporting GUI toolkits at all. Why not be radical? Somebody else (hah, everybody else) will look after the GUI toolkit users :)
  • by Anonymous Coward
    Hah! I'd been wondering why no one had done this yet... Well, its good that this has been taken care of, now I don't have to make myself do it. ;-) Check out this screentshot, of Xggi running with 6 screens, each screen mmapped to the texture of a 3D cube which can be rotated : http://www.ggi-project.org/images/cube_withxggi.jp g
  • OK. First, to declare my bias, I work for a company called ReachIn [reachin.se] and all we do is this sort of stuff.
    3D is definitely worthwhile for modelling, visualisation and simulation. How worthwhile is a good question. But even these CAVES, at about ~$1 M have been bought by the oil and automotive industries and paid for in a few weeks.
    3D window managers are interesting, but how usefull they are remains to be seen. Without computers, for most things people actually work in 2D. Books and paper are 2D. Whether this is because the technology is deficient or because this is the most efficient way is a really tough question.
    I'm not yet convinced that you can get a whole lot more from a 3D UI, but, this is probably like some UNIX hacker in 1975 saying, what Windows? Nice gimmick, but what do you NEED them for, I can switch TTYs fast.
    But you should try shaping something in 3D, it's really a new experience, anyone who has ever tried to use 3D Studio Max or Rhino or whatever will appreciate that there is some manipulation of 3D stuff that is painfull in 2D 'cause it's SOOOO counterintuitive. What we do is have a PHANTOM from SenSable [sensable.com] and co-locate it so you can see what you can touch. It is really quite neat. We can then do medical training and shaping like you wouldn't believe. If you can, get to SIGGRAPH 2K in New Orleans and have a look around, they'll be stuff from us and hopefully some others that will show that 3D is really capable of being worked in.
  • by NickElm ( 9556 ) on Wednesday November 03, 1999 @07:28AM (#1567360) Homepage

    Unfortunately, I was notified by this thread a bit too late, so I guess this entry is getting a bit stale, but I thought I should post my thoughts here nonetheless.

    As many of you might have guessed by now, 3Dwm is a 3D-User-Interface research project at Chalmers Medialab [chalmers.se]. We're currently in crush-mode, as we're going to host a demonstration of the system on Friday (if you happen to be in the vicinity, be sure to drop by! Check out this page [chalmers.se] (Swedish only) for more information.), so we're currently putting a lot of time and effort into the system.

    I see a lot of concern about 3Dwm just being a fancy way to display plain 2D applications in 3D. This is true. Yet 3Dwm is more of a prototype than a full project, and we will be addressing the questions of fully three-dimensional applications in our current core project, im3D, the Immersion3D User Environment. If you think 3Dwm is cool, then imagine having apps that are built for three dimensions. A modelling program might look like a workshop you may step into and use when designing your 3D-models. Your plain ole' debugger (gdb) might have a fancy 3D-dimensional interface to allow you to look at different threads of execution, stacks and heaps in an intuitive way. The plain 2D-VRML browser is now a gateway into the actual model which allows you to step into and actually explore the world from within.

    Well, that's some of the hype, anyway. 3Dwm and im3D is still under heavy development, but we hope to be able to release the code under a fairly free license (as in GPL, but we don't know at this point) and post it for the rest of the community to enjoy. Yes, it helps if you have a CAVE or a HMD, but we'll support (and already do) normal desktop systems as well.

    If you have any specific questions, comments, flames or criticism which you really want to make sure reaches us, then mail us at 3dwm@medialab.chalmers.se [mailto]. Thanks for all your feedback!

  • There's a difference between 3D, the illusion of having 3 dimensions, and a perspective change. 3Dwm is like holding a piece of paper and turning it away from you, which is pretty silly and if you're trying to read it VERY silly. What 3D should be is building applications that are polygon based and using that extra dimension in a practical way.

    Its just eyecandy.

  • If it does the sizing, it is doing positioning. It's also responsible for choosing how to draw borders around windows, so doing angled lines shouldn't be that hard.

    But what next? Instead of "autoraise" timing, will there be a "flyby" setting so that you point at it and it goes flying past your head?

    If people start using that thing that blows holes in windows in X apps, it'd be at least one step closer to my dream of having a quake-like interface as a WM.
    Imagine that set up like an office - whiteboards for knotes, and when the manager walks in... ;)
  • Why would Linux run it first? It doesn't even sound like these guys have a Linux port of this system. Yes, the speculated that it would run on a 3D accelerated PC running Linux, but the machines they were testing on were SGI's: An Onyx and an O2.

    So I guess when those holographic display monitors pop up, Irix will run it first :)


    "You can't shake the Devil's hand and say you're only kidding."

  • by Dast ( 10275 ) on Wednesday November 03, 1999 @04:59AM (#1567372)
    None of this is going to be too useful until we get some good input and output devices to use with it. Think about it, how well does a mouse that moves around in 2d work in a 3d environment? Same with your monitor, it is very limiting in a 3d environment.

    What we need are gloves and goggles. Then this will become somewhat usefull. Think of all the desk space you could have. All you have to do is turn your head to a blank area. Don't like where an application is? Grab it with the glove and move it. With even more input devices, the implications for pr0n are unthinkable. ;)
  • I've been kicking around this idea for a while and though it's cool, it'll only really come into its own with 3D displays and 3D input devices. You could probably cobble something together with some of the optical displays popular in wearables these days and perhaps a web-cam set up to track hand motion. It'd be kind of neat to reach "into" the display and manipulate a program. It'd also be kind of neat to map 2D programs like Netscape onto objects like cubes or polyhedrons which you could then rotate to find the program you're looking for.

    It's a cool idea though, gotta admit :-)

  • Could X be hacked in such a way that it was double-headed, but displayed nearly the same thing on each monitor?
    Next all you would have to do is take apart 2 pair of crummy LCD glasses (where each eyepiece is the same screen) and reassemble them so each eyepiece goes into a different video card.
    Therefore, IIRC, you could have a truly 3D Linux system, and change the depth by altering the relative positions of windows on each monitor.

    What do you think sirs?
  • coding 4d is way easier than visualizing it.
  • Interesting idea. An enhancement would be to have applications present different views at different distances. Another idea would be to take a user on a tour along different windows to let him/her perform a task (also great for training since people are good at remembering routes).
  • I once had an intense argument with a fellow CS dork friend of mine, and the conversation was about this point: UI development is all about *text*. Regardless of how fancy multimedia gets, and how good are computers are at actually speaking to us, when it comes down to straight hard work, text is the way to go.

    Remember Snow Crash? What happened when the main character did real work? He brought up a straght 2D environment, did his coding, and then went back to 3d funland. From the screen shots, the tech looks cool, but it just makes everything harder to read.

    3D is cool, and I could see some uses to "folding" or "pushing" away apps that you are not using, but when it comes to text, we all want it nice and flat right in front of our eyes.

    Do you read books sideways? Why does everyone drool over trinitron flat screen monitors?

    So I fail to see how a full 3D window manager actually makes X better for us. I understand the argument that 3Dwm is designed for this Cave thing, but the 3D aspects of the Cave have much to do with the structure of the Cave itself, and how a person interacts with it.

    Regardless of how I feel about its uses, 3D UI's definitely need research. Only from total knowledge on a subject can you truly determine what will work, and what will not.

    "You want to kiss the sky? Better learn how to kneel." - U2
    "It was like trying to herd cats..." - Robert A. Heinlein
  • Actually, you can in fact write a sort of 4-dimensional window manager. Here's how:

    Most regular window managers have the concept of multiple 2-dimensional "workspaces" (which, being 2-dimensional, might better be called workplanes). Now, and in fact Enlightenment actually does this, you can think of these workspaces as stacked on top of each other, with z = n for workspace n. Changing workspaces would then mean setting the viewport so you view the plane z = n.

    I might add that this system can in fact be thought of as 3-dimensional because it takes 3 numbers to specify completely a window's position in the environment.

    Now, if you have a 3-dimensional workspace, and you have multiple workspaces you can select, then it will in fact take _four_ numbers to completely specify the window's location (workspace, x, y, z). Therefore, this is a FOUR-dimensional window manager!

    Depending on how you number the workspaces in the (Mo/Less)tif window manager, it could also be a four-dimensional window manager. Numbering them as:

    1 2 3
    4 5 6
    7 8 9

    makes it three-dimensional. If, however, you take advantage of the fact that the pager is itself two-dimensional, you can number them as:

    (1,1) (1,2) (1,3)
    (2,1) (2,2) (2,3)
    (3,1) (3,2) (3,3)

    Then you need 4 numbers to specify the position of a window again, making the workspace 4-dimensional.

    One last thing: depending on the workspace structure, windows that are presistent over all workspaces could be thought of as 3 or 4 dimensional themselves...

    As far as visualizing 4-dimensional objects... the standard procedure is to show many 3-dimensional slices of it over time.

    Note by this definition, space is 4-dimensional because you need the time, as well as X, Y, and Z coordinates, to locate things.

    Rudy Rucker is probably the best at writing books about dimensions greater than 3. The two I know of are "Geometry, Relativity and the Fourth Dimension: Toward a Geometry of Higher Reality", which is out of print, and "The Fourth Dimension: A Guided Tour of the Higher Universes" which you can still get from Fatbrain.com.
  • OK this really is kinda stuck as so many people are saying. I mean all of its apps are 2D, when it really is 3D. And what good is a 3D window manager if you don't have any apps for it. But how can you develop 3D apps if you don't have a window manager to run them in. Ohhh circle logic.

    But slightly different if you want to start getting into theoretical ideas about it let's see. Imagine you are an engineer and designing something (or even to make Hemos like the idea a nano-something :). Well now you can have a real walk around kind of idea. And maybe there are chalkboards around that you can write notes on that are saved. Or even a calculator floating by you whenever you need it. But the 3D app really is the modeler of whatever. This would be great for graphic artists (imagine being in the world you are designing), engineers (seeing what you are building), doctors doing virtual surgery, etc.

    The ideas are there you really just have to stop thinking in terms of flat info. Which I'll admit is difficult as everything today is 2D. Newspapers, books, computers, tv, just about everything is 2D. Except the real world.

    The biggest thing this 3Dwm needs is actually a good way of displaying it. You really do need a virtual environment to use it effectively. Which they have in the Cave(?) but not every home user will have, well untill they come out with holographic monitors.

  • The whole window idea just doesn't seem to work for a 3d environment. I don't see any good way of porting it either.

    Probably, the only way anyone is ever going to get a true 3d gui to be accepted is to write a ton of new "3d" apps for it. The 3d object idea is a good one.

    Another problem is controllers. The mouse is a pretty poor 3d manipulation tool.

    Another problem is the screen. The screen is a 2d window into your 3d gui? No, until some form of holographic display, you'll still have a hard time getting this accepted.

    Here's what it all boils down to: Everything you currently have in terms of hardware and software is linked to that 2d window metaphor. None of it fits a 3d metaphor. Trying to link them together is trying to use the 3d metaphor with hardware that doesn't fit and with software not made for it. Just doesn't work.


Always leave room to add an explanation if it doesn't work out.