I agree with you, and I also think this raises major issues around the extent to which we want the government to have access to information. The truth is a double-edged sword and hopefully we will someday have a system where the government has the data and analysis tools to catch the Enrons of the world, and the public has the data and tools to catch the governments when they abuse power.
The magic pixie dust is the realtime collaboration. That's why it matters that the data is in the virtual world.
Also, I have to respectfully disagree with your assumption that using three dimensions to display data gets you no more understanding than using two. Just the ability to see time, grouping, and magnitude all at once exposes relationships that were not obvious in two dimensions. This is why scientists have been using 3+D visualization for a long time. Why not apply the same techniques to business and to government?
Arkowitz
Yes, garbage in garbage out. There has to be a mechanism to vet data. Let's see, I think in the scientific community it's called peer review...
Arkowitz
Tufte unfortunately did not have an actual 3D world in which to experiment. He was stuck on the page.
Screenshots are bullshit no matter what you do; 3D requires motion. Try it someday.
Arkowitz
The fuss is that Google has the potential to control far more than Microsoft ever has. I think there are significant differences between Google and the typical defense contractor. Maybe they are good differences; maybe not. We'll see I guess.
yeah I was referring to Microsoft Office as opposed to Google Docs.
Zune!
Windows Live!
The Google Docs angle is very dangerous because it provides an alternative to Microsoft other than linux+openoffice... an alternative which is even more closed than Microsoft Office because you can only run it off Google's servers and they see EVERYTHING THAT GOES ON. I can't believe this Google fanboy Obama brought in would even consider this.
Arkowitz
Yup. The rezzing is fun to watch for a little while... then it is just plain slow. There are ways to have the objects rez eachother like a swarm but the grey goo fence eventually kills it.
The graphs are still useful in SL though; I've given presentations to lots of people in there and they like looking at data this way.
Wonderland blows SL away architecturally. Simulating everything on the server, like in SL, has major drawbacks; the Wonderland implementation of CICP is so fast because the description of an entire graph goes into a file, the virtual world server sends a message to all the clients saying "there's a new cell at such-and-such place", and then each client fetches the definition file via HTTP and adds everything to the scene graph locally.
Arkowitz
Why don't you try actually running the code? You sound like my high school computer teacher. He was all about comments but couldn't program to save his life. My code is usually very hard to understand, but it does cool things. If you can hang with that then open up the pdf, copy the file, put it into a prim, and then tell me how to improve it. I'd really appreciate it if you would send me a mod that spits out multiple "rezzer" prims that then each rez out a portion of the children, making the whole graph rez twice as fast.
Arkowitz
You can go see a couple sample graphs in SL at the dataviz sandbox in scilands... there's a slurl here http://sldataviz.pbwiki.com/
Or you can grab the script from greenphosphor.com and rez your own. Also, Elmo Dynamo has a video of one at http://www.youtube.com/watch?v=io62ENVxkQw
Croquet would probably work just great for this; I would need to have a SmallTalk programmer implement the protocol however; I'm not into SmallTalk.
Arkowitz
Well, here's what I'm going to use it for:
http://www.youtube.com/watch?v=i1OgIZ2mVhk
Other uses include placing buildings into worlds, which are interactive and linked back outside the world to an actual architectural or cad system; or allowing AI's to place avatars into worlds as well as build things as requested by other users in the worlds.
Arkowitz
That is exactly what I have been thinking. Something like this, were it to be adopted/used/spread, will need to be defended.
Thanks for the feedback; this helps.
You would use the protocol if you are a program which wants to put some 3d objects into a virtual world and have users within the virtual world interact with your objects, resulting in messages coming back to you.
Virtual worlds can be great platforms for interacting with people and things but they need a standard, cross-platform way to accomplish that interaction.
Slashdot Mirror
I agree with you, and I also think this raises major issues around the extent to which we want the government to have access to information. The truth is a double-edged sword and hopefully we will someday have a system where the government has the data and analysis tools to catch the Enrons of the world, and the public has the data and tools to catch the governments when they abuse power.
An OLAP database is the perfect backend for Glasshouse. What OLAP has been missing is a 3D user interface; now it's here.
LOL those "user interfaces" in CSI are so cheesy - Arkowitz
The magic pixie dust is the realtime collaboration. That's why it matters that the data is in the virtual world. Also, I have to respectfully disagree with your assumption that using three dimensions to display data gets you no more understanding than using two. Just the ability to see time, grouping, and magnitude all at once exposes relationships that were not obvious in two dimensions. This is why scientists have been using 3+D visualization for a long time. Why not apply the same techniques to business and to government? Arkowitz
hey man that's one of my favorite movies.
Yes, garbage in garbage out. There has to be a mechanism to vet data. Let's see, I think in the scientific community it's called peer review... Arkowitz
Try the new beta! There's an applet which runs via Java webstart, and gives you a UI for defining graphs - no more typing commands in chat. Arkowitz
Tufte unfortunately did not have an actual 3D world in which to experiment. He was stuck on the page. Screenshots are bullshit no matter what you do; 3D requires motion. Try it someday. Arkowitz
The fuss is that Google has the potential to control far more than Microsoft ever has. I think there are significant differences between Google and the typical defense contractor. Maybe they are good differences; maybe not. We'll see I guess.
yeah I was referring to Microsoft Office as opposed to Google Docs. Zune! Windows Live! The Google Docs angle is very dangerous because it provides an alternative to Microsoft other than linux+openoffice... an alternative which is even more closed than Microsoft Office because you can only run it off Google's servers and they see EVERYTHING THAT GOES ON. I can't believe this Google fanboy Obama brought in would even consider this. Arkowitz
I for one have a problem with our government documents and processes being hosted by a private company. At least Microsoft just sells software.
Yup. The rezzing is fun to watch for a little while... then it is just plain slow. There are ways to have the objects rez eachother like a swarm but the grey goo fence eventually kills it. The graphs are still useful in SL though; I've given presentations to lots of people in there and they like looking at data this way. Wonderland blows SL away architecturally. Simulating everything on the server, like in SL, has major drawbacks; the Wonderland implementation of CICP is so fast because the description of an entire graph goes into a file, the virtual world server sends a message to all the clients saying "there's a new cell at such-and-such place", and then each client fetches the definition file via HTTP and adds everything to the scene graph locally. Arkowitz
Arkowitz Jonson what's yours mon?
Why don't you try actually running the code? You sound like my high school computer teacher. He was all about comments but couldn't program to save his life. My code is usually very hard to understand, but it does cool things. If you can hang with that then open up the pdf, copy the file, put it into a prim, and then tell me how to improve it. I'd really appreciate it if you would send me a mod that spits out multiple "rezzer" prims that then each rez out a portion of the children, making the whole graph rez twice as fast. Arkowitz
You can go see a couple sample graphs in SL at the dataviz sandbox in scilands... there's a slurl here http://sldataviz.pbwiki.com/ Or you can grab the script from greenphosphor.com and rez your own. Also, Elmo Dynamo has a video of one at http://www.youtube.com/watch?v=io62ENVxkQw Croquet would probably work just great for this; I would need to have a SmallTalk programmer implement the protocol however; I'm not into SmallTalk. Arkowitz
Well, here's what I'm going to use it for: http://www.youtube.com/watch?v=i1OgIZ2mVhk Other uses include placing buildings into worlds, which are interactive and linked back outside the world to an actual architectural or cad system; or allowing AI's to place avatars into worlds as well as build things as requested by other users in the worlds. Arkowitz
I figure if I patent it, and assign the rights to the patent to EVERYONE, then nobody is an OTHER. Ark
That is exactly what I have been thinking. Something like this, were it to be adopted/used/spread, will need to be defended. Thanks for the feedback; this helps.
lol yes I did put source into a pdf. why not? open your mind.
You would use the protocol if you are a program which wants to put some 3d objects into a virtual world and have users within the virtual world interact with your objects, resulting in messages coming back to you. Virtual worlds can be great platforms for interacting with people and things but they need a standard, cross-platform way to accomplish that interaction.
nice. this seems like actual good advice.
well if we changed the name we would change it to "bubbles". I've read SnowCrash twice, btw, and it is overrated. I prefer Gibson's vision.