Slashdot Mirror


3D File Manager on Linux Wins NSF Prize

MadFarmAnimalz writes "Science Magazine's reporting on the results of the NSF's Science and Engineering Visualisation Challenge and the first prize in the Illustrations category has been claimed by the Innolab 3D File Manager, which was developed on linux. Apparently this involves arranging data in a ferris wheel type structure." The data is arranged by its relationship with its content, rather than by its physical position on a hard drive or its file system.

23 of 283 comments (clear)

  1. OT: 3d file manager by ArmorFiend · · Score: 5, Interesting

    No, I didn't RTFA, and I'm sure I'll get modded Offtopic, but the thought occurs to me:

    Why are we, the free software community, busting ass to integrate pseudo-3d technologies to the desktop (AA-fonts, SVG-icons, real alpha blending), while it seems obvious that the next step is going to be a fully 3d-enabled desktop, with 3d icons placed in the current 2d-metaphor? Already new computers with new accellerators can push so many polys that the overhead is not measurable by users.

    1. Re:OT: 3d file manager by Nooface · · Score: 3, Interesting

      This article in Reuters describes the Heliodisplay, a device that creates a two-dimensional image which appears to hover in mid-air and can be seen from several angles. Similar to the Fog Screen, the Heliodisplay projects the image into a cloud of "benign" particles that it sprays into the air. The developer states that he was directly influenced by the hologram communicator shown in the "Star Wars" movies. Here is a set of video clips demonstrating the device in action, and there is more detail about the design on p. 14 of Emerging Display Review (PDF).

      --

      Nooface
      In Search of the Post-PC Interface
    2. Re:OT: 3d file manager by Pope+Raymond+Lama · · Score: 2, Interesting

      I would agree with you!

      Actually rather than fully poligonal icons and windows, I'd like to see texturing and lightning effects going realtime on current X' windows. I mean.., actual lightning effects, not pre-rendeered stuff. And of course, nice 3D transformation effects on the oppening and closing of windows.

      Maybe this could be achieved with a couple more people working on the transluxent project, and making it go beyond the extremely outdated alpha effects. For Heavens..it had been cool to show off a desktop running WindowMaker and half a dozen Eterms open, but we need to leave taht behind.

      Maybe an extension to X to use OpenGL, and a quick hack on KDE to animate -and post render - windows as they open/close go in/out of focus is feasible without that much manpower. A K3D could then raise. (Not to be mistaken with the existing K3B cd recorder, or K2D drawing tool)

      --
      -><- no .sig is good sig.
  2. Can it be downloaded and taken for a test drive? by Proudrooster · · Score: 3, Interesting

    This looks really cool. Anyone know if it can be downloaded so we can take it for a test drive? Please post a download link if you have one. The article doesn't provide any links except to a static image of how the program visually organizes the files.

  3. This is strange by rkz · · Score: 4, Interesting

    Rather than having a software layer which groups files by content rather than tree structure, why not impliment a SQL type of system to access ReiserFS after all it is a database underneeth.

    Doing this through the filesystem strikes me as alot more efficient than a quick hack of a filemanager.

    Even Microsoft are working on a file system based

    1. Re:This is strange by Anonymous Coward · · Score: 1, Interesting

      Oracle's Internet File System does it since years ago; it maps entire filesystems on a database, allowing SQL operations on data.
      It's slow like hell though: nothing that I'd use on my system without being forced to.

  4. 3D GUIs? by immel · · Score: 2, Interesting

    Some people have wondered in the past "What happened to the 3-D GUIs that were promised to us in the past from movies like 'Jurassic Park'?" Well, here it is. But really, what are the advantages of this system that cannot be offered by a 2-D GUI? It's really cool and all, but don't you think this would be a slight waste of CPU or GPU power?

    --

    10 Bits= $.25
    100 Bits= $.50
    110 Bits= $.75
    1000 Bits= 1 byte
  5. 3D Directories for OS X by afflatus_com · · Score: 4, Interesting

    If you are on OS X and would like to sample 3D navigation of disk drive content, there is a nice free project that does this, aptly named 3DOSX.

    It uses Open GL to make the file system into 3D rotatable platters, and the platters are linked together. Can swim around the platters looking at the different documents.

    Some screenshots are here:
    3DOSX Screenshots

    The project homepage is here:
    3DOSX Homepage

    It is an interesting look into alternative ways of doing things.

    --

    -----
    Cast a Cold Eye
    On Life, on Death
    Horseman, pass by
    --W.B. Yeats' gravestone
  6. Pretty-printing by ewn · · Score: 5, Interesting

    This is more pretty-printing than real innovation. They claim to arrange data by relation but the thing still knows active folders, parent folders and subfolders. And the color scheme (subfolders are blue) focuses on the hierarchical structure of the folders and not the relation of the data. So they took one way of organizing and presenting files that works for most people most of the time but has a few big shortcomings, pretty-printed it in a somewhat confusing way and added relational sugar that can only add to the confusion.

    Pretty, but not impressive.

  7. That structure looks unwieldly by jea6 · · Score: 4, Interesting

    One problem with this type of arrangement is that it requires thoughtful meta-description of all content (which scientists do but PHBs don't). What you have an interesting way of representing "degrees of separation", not a "triumph of Linux on the Desktop." The challenge ( http://www.nsf.gov/od/lpa/events/sevc/overview.htm ) was:

    "This new international contest is designed to recognize outstanding achievements by scientists and engineers in the use of visual media to promote understanding of research results."

    So for the visual representation of linked data structure, sure this looks great. As a GUI, heck no. "File Manager" seems like a misnomer here.

    --

    sarchasm: The gulf between the author of sarcastic wit and the person who doesn't get it.
  8. When can we download it? by Majin+Viper · · Score: 2, Interesting

    i want to give this a go, can i download it , no links to official site or anything

  9. Secret of Mana Anyone? by Mister+G · · Score: 5, Interesting

    Does this look like a souped up ring interface from the classic Secret of Mana published by Square-Enix (nee Square)?

    It does to me...

  10. Re:ls -R / by nutshell42 · · Score: 4, Interesting
    Well, if you look at the image in the article, it *is* one hell of a ferris wheel and while I can't say how effective it is without trying it, it really doesn't look less cluttered than a normal file-list but it could be useful in distinct parts of a file system where a maximum of *visual* organisation is necessary (in a cvs-tree perhaps, to see what files influence which others if you change them, just an idea)

    And just to get my daily flamebait rating: Who modded the parent offtopic? It's a valid questioning of the usefulness of the program mentioned in the article

    --
    Don't think of it as a flame---it's more like an argument that does 3d6 fire damage
  11. Don't forget 3Dwm by axxackall · · Score: 2, Interesting

    3Dwm is the most promising to really alterate our human-computer interaction.

    --

    Less is more !
  12. Memory Palace of Simonides by handy_vandal · · Score: 2, Interesting
    Summary of the competition principles, from the NSF web site:
    "Photographs, pictorial and diagrammatic illustrations, computer graphics, and animations are now an essential aspect of communicating research findings. These new avenues prompt discussion of different techniques, and encourage innovative approaches to visual communication. This competition was created to reward these new techniques and ways of communicating."
    It's interesting that the ancients were well aquainted with and made extensive use of similar principles of communication, in the form of mnemonic metaphors used by orators:
    "In the ancient Greek arts of rhetoric, memory was a science. The science has an origin in what is surely myth. The poet Simonides of Ceos was hired by the noble Scopas to attend a formal banquet as a paid performer, singing a poem of praise of his host. As was the custom, Simonides began by first praising a pair of gods. After the performance, Scopas informed the poet that he would only get half of the agreed-upon fee, the other half he should get from the gods who had stolen the limelight.

    "At that point, a messenger came in and told Simonides that a couple of athletic men on horseback were outside waiting for him. Simonides went outside, but nobody was there. But, while he was outside, the gods destroyed the banquet hall to teach Scopas a few lessons about respect. (The lessons being pay the poet; don't mess with the gods; and, memory palaces are a gift from above.)

    "The banquet hall was so badly destroyed that none of the diners could be recognized. Simonides was able to remember the exact location of every guest at the banquet, using the principles of the Method of Loci, the science of memory. Later, Cicero (106-43 B.C.) wrote a few pages on the science in his classic work, De Oratore. [See De Oratore, II. lxxxvi. 350- 353]. The definitive treatment in Greek literature, however, is the work of an unknown author previously attributed to Cicero in the classic work Ad Herennium.

    "The principles of the science are fairly simple, at least using our modern hindsight. A person who wished to memorize a large work, say an address after dinner or the closing argument of a legal proceeding, would begin by constructing a memory palace. While novices constructed a palace by going to a real one and memorizing the rooms, the memory palace could just as easily be any structure that can be imagined."

    Source: Mappa Mundi
    --
    -kgj
  13. Re:Dock by justsomebody · · Score: 2, Interesting

    Sorry, my bad.

    After reading my post again I realised where the trouble is.

    What I ment was:
    With or without magnification this approach to file system is still stupid. I just can't imagine troubles people would have with 3D fs layout, when there's a lot of users that have problems with 2D, which is far more simpler to imagine than 3D.

    This representation is feasssible in some movie to produce high tech feeling, but in real life is unusable

    --
    Signature Pro version 1.13.2-3 release 83.5 beta3try7 after-breakfast edition
  14. Re:Not New... by goatbar · · Score: 2, Interesting
    No kidding... not too different than what Fourth Planet was selling with their nScope product. Did that about 1997/1998 or so. Also saw a number of programs like this in the Stanford Computer Graphics labs back in 95/96 in Levoy and Hanrahans' classes. Ah... the joys of watching network traffic with etherman (??) on IRIX 4.x. Also very similiar to the winner.

    Okay, okay... I entered and didn't win with xcore. Not quite as flashy, but I thought I'd have a good chance.

    Congrats to all the winners. Those finalist projects are pretty amazing.

  15. the 3d interface you love by sniggly · · Score: 2, Interesting
    Lessons learned from gaming?

    One of the few 3d interfaces I love to use is the Homeworld / Homeworld2 interface for rotating and zooming in space.

    The build & research manager in Homeworld is 2d though.

    For most types of data representations the 2d tree interface is ideal. Maybe we are far too used to it; we don't now really see what we can do with a 3d interface that we can't do just as efficiently as in 2d. Even in a lot of movies 3d is just an enhanced use of 2d displays.

    What we do most is deal with text. Text is very typically a 2d thing because its on paper or a representation of paper (slashdot textarea box). Text in 3d space... doesnt make sense. We'd have to learn a language of 3d space to understand references. Once we learn such a language it might be extremely efficient though, I guess time will tell.

    --
    Of those to whom much is given, much is required.
  16. Re:Dock by Minna+Kirai · · Score: 3, Interesting

    1) Providing reinforcement
    Yes, it gives reinforcement. But there are other ways to do that- color change, make it bounce or vibrate, etc. But any change which alters the size of their area sensitive to mouse-clicks should be a big no-no.

    2) Fitt's law - the button you're trying to click on gets bigger when you get near it, so it's easier to hit.

    That doesn't really work... it's circular reasoning. After all, the computer doesn't know which button you want to hit. Some button gets bigger and easier, but not necessarily the right one. If it knew which button you wanted, it could be large all the time.

    Changing the size or position of GUI elements in response to mouse motion should generally be avoided (unless you've moved to a whole other paradigm than the regular "windows, buttons, and scrollbars" layout. OSX has made no drastic transition like that. Besides the Dock and Apple menu, it's all the same).

    The user should feel assured that moving the mouse doesn't do anything- only clicking (or drilling) it has an effect. The GUI should partially emulate a consistent, physical world- predictable cause/effect, etc.

  17. Re:WTF? by Minna+Kirai · · Score: 2, Interesting

    How is this easier to use than this?

    The picture you reference (the white and yellow boxes in a big circle) is a classic example of a computer algorithm mistake: naive connectivity graph generation.

    What happens is a programmer notices that some set of data has relationships between the elements, so he decides to draw them onscreen for the user as boxes connected by lines.

    But it turns out it can be quite tricky to construct a graph layout that'll be easy for a human to understand. You'd want to minimize both the length of connecting lines, and the times they cross each other, which is a tough problem. So programmers tend to skip working on it and just space the boxes around the edge of a circle, completely ignoring linkage for purposes of placement.

    Here is another bad example of this lazy graph-layout in action. At least using a circle is better than putting the nodes on a 2d grid!

  18. Re:Obligatory FSN/"Jurassic Park" reference and li by Roguelazer · · Score: 2, Interesting

    fsv.sourceforge.net Almost identical program, but open-source and for Linux. Now if only there was a way to run apps with it...

  19. Re:Evolution's VFolders -- for file systems by digitalhermit · · Score: 2, Interesting

    Really interesting stuff there. I remember reading about the GNU "Storage" filesystem and something like that would be quite useful.

    What I'd like is something like a CVS based filesystem; i.e., one that can automatically track changes to my documents/files/etc.. If I perform some upgrade and everything breaks, I could then retrieve a 'tagged' version of the OS. The same would apply to individual files; per-document versioning systems would be obsolete as the filesystem itself would take care of everything.

    One of my biggest data problems is that I have about a dozen systems throughout my house. All of them have different data in different states. Yes, I use a file server and CVS to track, but it's still unwieldy, especially since there are some large files on each system that would quickly fill up about 150G.

    It's funny, but some sort of P2P file sharing system -- something that's being tarred as a piracy tool -- would actually be very useful for my situation. I could run the p2p server on each machine and the client could transparently query each one for the file I require. Yup, more useful technology that may get banned.

  20. Sometimes 2.5D is enough by stefpub · · Score: 2, Interesting
    The best new idea about data organization I've seen so far comes from Jeff Raskin's Humane Interface. The user zooms in or out on an infinite 2D plane where information is stored in groups, the more general the bigger, the more detailed the tinier.

    Here's a demo.