Woke up at 5:45 yesterday to get an early start on the summer. Drove up to Bear Mt. with Alex and friend Matty (sp?) for a little hiking and introduction to bird watching from the man himself. Like most other pursuits these days, the technology really makes a difference. Good binoculars are a must for enjoying this very peaceful and meditative endeavor. And looking through them all day is not unlike staring at a screen (although a very high resolution one.) Thought a lot (again) about using natural landscapes to model data. Also I realized (although this may not be true) that xml has some serious problems. It might be the case that the data (the content) is itself the most consise (meta-data) description of itself. Or, in other words, it might be that we will not be able to decide on any but the most trivial schemas to share in common. Oh well. Still more thinking to be done.
>Thought a lot (again) about using natural landscapes to model data.
>>what do you mean?
Using natural forms to model data is a solution. The problem is having too much data. The idea is that the human brain is very good at navigating the "real" world because those are the skills we've been selecting for millions of years. Staring at spreadsheets filled with numbers is something we can do, but we're not very good at it (in comparison, to say, selecting a familiar face in a crowd.) Amex (I think it was them) had a big project a few years ago to model the stock market using a forest as a metaphore. Each company would be a tree in the forest, and different visual aspects of the tree (height, leaf pattern, coloration, ect...) would represent different financial facts about the stock (changing in real time.) Their hope was that allowing analysts to fly around and through the forest would promote intuitions about stock movements more than pouring over those same numbers in giant tables. And that's if you even could "look over" the raw data. With the amazing storage available these days, this will often not even be possible, because there is just too much to print out. You can't look at a billion pieces of data and make any sense out of it, but you could walk around on a virtual beach where each grain of sand represents a point of data (with, say, the color and courseness of each grain representing two variables associated with each data point.) The hope would be that although a million page book containing the billion points of data would be almost useless, you could get an immediate sense of the data as a whole if the data were a landscape.
I first encountered these ideas in Howard Reingold's book Virtual Reality (or was it Virtual Realities?) from way back around 1990. William Gibson does a good job of explaining this in his (even earlier) Sci-Fi book Neuromancer. Also, Neil Stephenson's SnowCrash has some good explanation as well (plus the great word "metaverse" to describe this virtual world where data is represented as objects.)
Virtual Reality 1991--good book--wish stuff stuck in my siv of a brain--i think i will read Neuromancer again on the way to Cape Cod next weekend!!--and bring "Walls Come Tumbling Down" RAW--just in case of rain--did you read??
Yes I read the RAW. It's really good if you can deal with the screen play format. Quick read once it gets going. You'll like it. Great ending. I wish someone would make it into a movie.
|
- jim 6-22-2000 4:20 pm
>Thought a lot (again) about using natural landscapes to model data. >>what do you mean?
- dave 6-22-2000 4:25 pm
Using natural forms to model data is a solution. The problem is having too much data. The idea is that the human brain is very good at navigating the "real" world because those are the skills we've been selecting for millions of years. Staring at spreadsheets filled with numbers is something we can do, but we're not very good at it (in comparison, to say, selecting a familiar face in a crowd.) Amex (I think it was them) had a big project a few years ago to model the stock market using a forest as a metaphore. Each company would be a tree in the forest, and different visual aspects of the tree (height, leaf pattern, coloration, ect...) would represent different financial facts about the stock (changing in real time.) Their hope was that allowing analysts to fly around and through the forest would promote intuitions about stock movements more than pouring over those same numbers in giant tables. And that's if you even could "look over" the raw data. With the amazing storage available these days, this will often not even be possible, because there is just too much to print out. You can't look at a billion pieces of data and make any sense out of it, but you could walk around on a virtual beach where each grain of sand represents a point of data (with, say, the color and courseness of each grain representing two variables associated with each data point.) The hope would be that although a million page book containing the billion points of data would be almost useless, you could get an immediate sense of the data as a whole if the data were a landscape.
I first encountered these ideas in Howard Reingold's book Virtual Reality (or was it Virtual Realities?) from way back around 1990. William Gibson does a good job of explaining this in his (even earlier) Sci-Fi book Neuromancer. Also, Neil Stephenson's SnowCrash has some good explanation as well (plus the great word "metaverse" to describe this virtual world where data is represented as objects.)
- jim 6-22-2000 11:31 pm
Virtual Reality 1991--good book--wish stuff stuck in my siv of a brain--i think i will read Neuromancer again on the way to Cape Cod next weekend!!--and bring "Walls Come Tumbling Down" RAW--just in case of rain--did you read??
- Skinny 6-23-2000 4:38 pm
Yes I read the RAW. It's really good if you can deal with the screen play format. Quick read once it gets going. You'll like it. Great ending. I wish someone would make it into a movie.
- jim 6-23-2000 6:16 pm