eftychia: Me in kilt and poofy shirt, facing away, playing acoustic guitar behind head (Default)
posted by [personal profile] eftychia at 05:25am on 2006-02-12 under

"I am a firm believer in the people. If given the truth, they can be depended upon to meet any national crisis. The great point is to bring them the real facts." -- Abraham Lincoln

eftychia: My face, wearing black beret, with guitar neck in corner of frame (pw34)
posted by [personal profile] eftychia at 06:27am on 2006-02-12 under

The house that's burning is near enough that I can smell it plainly in my bedroom, but the snow is so heavy that I cannot see the smoke. Only a few sirens (whoops, there's another), sounding like it's a short distance west of here. Unless the snow seriously changes how the smell moves, it can't be more than a couple of blocks away.

Judging by what's piled up atop the chimney-cap so far, this is a bit more than what I heard in the last forecast I caught. And it's still falling thickly. Very pretty (and here in Baltimore, just as [livejournal.com profile] anniemal pointed out in Arlington, a blessed reduction in traffic noise, though someone did just go past with a car stereo cranked up to nine-and-a-half in the pre-dawn), mood-setting, relaxing ... I'll wait until it's time to start shovelling it to worry about having to move more snow than was forecast.

Yesterday, before the snow started to fall, it "smelled like [approaching] snow" inside my house, which I found rather interesting.

I think I hear a neighbour wielding a shovel now. Earlier than I'd expected for a Sunday, especially with it still coming down.

eftychia: Lego-ish figure in blue dress, with beard and breasts, holding sword and electric guitar (lego-blue)
posted by [personal profile] eftychia at 07:49am on 2006-02-12 under ,

Nah, this ain't workin'
It's no way to do it
Just puttin' numbers on a CRT
This ain't workin'
There's a better way to do it
Show me my data in full three-dee

I'm using a spreadsheet to sort out options, and to compare the costs of different combinations of services from different providers. This is all well and good -- it's a more suitable use of the tool than what I usually do with it (like most people: use spreadsheets as quick & dirty databases, one table per sheet, using vlookup() and hlookup() to create relations[1] if the database is more than just a flat file).

The problem is, this is the second problem in two weeks for which I want a three-dimensional spreadsheet. I want to see a transparent cube, with values in cells addressed by row+column+plane. I'll settle for slicing my data up and linking chunks with references and lookups, since that's what I've got, but I want that three-dimensional representation. Intuitive. Convenient. (Actually, for one problem I really want to see a tesseract[2].) Instead, I'm sitting here trying to figure out the most reasonable way to slice things up to be laid out on one or more two-dimensional grids.

Y'know, I've said before that a lot of representations of VR in fiction strike me as silly -- nifty, dramatically effective, but silly -- because they'd use resource-intensive renderings for things for which we already have very efficient interfaces. So I mostly see VR as something for games (and other entertainment), and for specialized applications. It makes sense for a biochemist to play with modelling software that lets her "walk around" and "grab" virtual molecules. Depending on what they're making, it could be useful for mechanical engineers as well. The "everything you see is a metaphor for a chunk of code or a real-world interface" approach from some cyberpunk fiction strikes me as ... well, difficult to get right (in the sense of being more useful than a more straightforward but less interesting -- in a novel/movie context -- interface, unless you're tweaking complex information streams to take advantage of human pattern-recognition abilities[3], but even then it seems as though it'd be more useful for alerts and general status reporting than for detailed information and hands-on interaction[4]. So I've had trouble believing in worlds where full-sensory VR is the way all tech-savvy characters interact with computers. But as of this morning, I'm ready for consumer-priced VR just for the sake of a three- or four-dimensional user interface for Gnumeric or Excel.

I'm ready for my VR multi-dimensional spreadsheet now. As soon as I can get ahold of the goggles and the data-glove, and find out where to download the demo software... But I wanna try it out for a while before I decide whether to get a jack implanted in my neck for a neural hookup.


[1] Yeah, I'd prefer having a real query language and greater efficiency (enough vlookups and Excel gets pretty slow), but when I'm creating a database for myself I'm usually impatient to start using it, and misusing a spreadsheet means being able to take advantage of certain data-entry shortcuts without having to design custom input forms. Obtaining better front-end tools should probably go on my to-do list, I suppose. I use real database tools when I'm solving someone else's problems (because then I'm much more concerned with doing it right than just getting the first bunch of answers).

[2] It's just more intuitive to me, to represent data in the number of dimensions the data have, as much as can be conveniently rendered. I hated Karnaugh maps in my Computer Architecture class until I realized that the three-variable map was a 2x2x2 cube "unfolded" by putting a hinge in the middle of one face, and the four-variable Karnaugh map was a similarly unfolded tesseract. When I drew a cube on the board to solve a problem in class, the professor asked, "Okay, wise guy, what do you do with four variables," so I showed him: I drew a tesseract (okay, a two-dimensional representation of one common three-dimensional projection of a tesseract, if'n y'wanna get all technical at me), I'd been using my own notation to map six-variable symbol tables onto a tesseract for homework problems, because I never worked out drawing pictures of more than four dimensions in any useable way[5]. The instructor found me annoying.

[3] I must remember to dig up the InfoWorld column in which someone suggested an "environmental sounds" approach to data-center status monitoring. I know from experience that subtle changes in one's acoustic environment can be a very effective way of receiving status information about one's equipment: I remember being able to hear a paper jam in the Xerox 1075 copier and usually know which section it was in before the machine had detected it, while my attention was focussed on something else entirely. I remember being able to tell when someone had logged in on the Xenix box I administered, from the sound of the hard disk. Now imagine engineering that sort of thing intentionally, so that data center staff know "the web server is experiencing unusually heavy load" because of the rough equivalent of the sound of twigs snapping, or "the WAN has started having latency problems" because "the birds stopped chirping". This is where I can see VR interfaces being especially powerful, but this doesn't require VR to implement. I've been meaning to write a separate entry about it for a while, and may get around to it yet.

[4] And yes, geek (and SF fan) that I am, whenever I make this point my brain spawns a background process trying to come up with more ways in which it would be useful. Let's just say that I've simplified my argument for the sake of brevity, and we can discuss when and how I see a metaphorical VR UI being more useful than decorative another time. And that my thoughts on the matter are far from finished.

[5] In middle school I was told that humans could not visualize four-dimensional objects (obviously, by someone who hadn't seen enough three-dimensional projections of tesseracts to find one that "folded up in his brain" the right way to make the idea click). In high school I was told, "of course we can visualize four dimensions with training, but not five." Several friends took this as a personal challenge. One claimed to be able to picture six-dimensional geometric solids. I managed to visualize a few simple shapes in five but nothing complex -- just enough to convince myself that there had to be other people who could do geometry in more dimensions than that "visually" in their heads, if "mere ordinary me" could reach the edge of five with a struggle and hold four dimensions in my head long enough to sort out diagramattically whether I was doing polar<-&gr;rectangular coordinate conversions correctly. Not that I find four-dimensional visualization easy -- I do have to slow down, and keep reinforcing bits of it that try to slip away -- but it's "there" enough to be useful to me, and four a four-dimensional representation of four-dimensional data to be a good way for me to spot the patterns and relationships in those data. (Note that drawing four-dimensional diagrams to show somene else what I'm thinking remains quite a problem.) How to build four-dimensional representations into a VR interface is something I haven't thought about yet.


Oh bother. The footnotes took over again, didn't they? And I just sprayed for those last week.

Links

January

SunMonTueWedThuFriSat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24
 
25
 
26
 
27
 
28
 
29
 
30
 
31