Here are my goals for the next two weeks:
1) Figure out how to create a platform for the back row of the people
2) Set up the room with the lab tables
3)Create a beta of the Hispanic person
4)Figure out and run a test of the people appearing
I hope this goes well. Once this things are done, I’ll basically be done (except for making the signs and coding in real data).
As promised, I have two screenshots to show everyone. The first is of me working with my clickable US region map on the wall. The second is of two potential lab table designs. Currently, I can’t really decide which is better. The one on the right seems more lab-table like than the one on the left, but that one allows you to see under it and see more of the clothes, which can help to distinguish the people (though this might become moot because I might use the front of the table to put my sign with the data). I figure I’ll have everyone voice their opinion here and at the meeting and go with the general consensus. As far as the ring thing goes, I plan on making an elementary one tomorrow (unfortunately, Dr. Polack didn’t give me hers, so it’s currently in the way).
Screenshot 1:
Screenshot 2:
This week has been a bit better for me. I’ve successfully created the invisible overlays for the map on the back wall, so it’s now responsive to user interaction. This will lay the ground work for the object communication. Tentatively, I’ve planned it so the clicked region sends each lab table a message saying it’s been clicked, the lab table itself figures out how many people to make, and then people either go visible or start spawning in place. Each step in this equation is not very hard to do. Mostly, I’ll have problems making designing things in an aesthetic, functional way (and maybe dealing with the overflow prim problem too).
I’ve also tinkered a bit with making a lab table and will soon try and make my own ring platform system using the ShapeMaker program Dr. Polack gave me. Tomorrow afternoon/evening I’ll post a screenshot or two so everyone can get a better idea on what I’m designing.
So far this week, I haven’t been able to make much progress (beside some minor tweaking of the Asian mannequin and contact with the mannequin guy). He sent me some parts to create my own mannequin (some sort computer heating problems keep him from consistently staying online), but I haven’t been able to sign on to Second Life to try them out. Because of homework obligations, I really couldn’t spend much time in Second Life on Wednesday and Thursday. I was away for the weekend, so I had no time to do it then. Finally, I’ve been trying all day today Sunday, but power/internet problems have kept me off Second Life (both the Trinkle labs and my room cannot connect to Second Life…I can’t even keep a consistent connection for purposes of writing this blog!). Despite these problems, I hope to sign in and make some progress to show off on Wednesday.
This semester, that is Fall 2007, I’m doing an independent study based off my summer research project (for anyone who doesn’t know, it was titled Visualizing Data in Three Dimensions Using Second Life). I’m going to be creating a new visualization and solidifying my earlier test results.
Currently, I’m planning on this visualization being centered on the education data in the Census 2000 data set, with a focus on the racial makeup of undergraduate education. What I’ve done so far is to build 4 different races mannequins out of the 6, those being the Caucasian, African-American, American Indian, and Pacific Islander/Hawaiian. This pastweek, I’ve done two things: I’ve designed a beta version of the Asian mannequin, and storyboarded two possible arrangements of the room I’m displaying the information. I have reservations about the mannequin for the Asian, but I’m not really sure how to change it. Arguably, one of the most distinctive traits for Asian-Americans is the eyes/general facial makeup, but with these mannequins I’m not including any facial detail. I’m posting a screenshot of the mannequins for everyone to see. Regrading the story boards, I’ll just show them on Wednesday (I drew them on my own paper).
The transportation visualization is finally done. Everything has been checked and double-checked, so the visualization should ready to go. Last night, I went over every single bar/pyramid, and fixed every minor inconsistency found in the text output system, so every bar/pyramid should display the correct percentage now. Also, I made sure that everything percentage is visible from the elevator area without moving around.
Today, I added a couple bells-and-whistles for the final touch: 1) As per Dr. Polack’s suggestion, I added one invisble overhead light for the room. It does make things a bit brighter, but the colors become too washed out and bright if I use anymore. 2) I added an automatic door to the entrance of the room. Basically, the main reason I added it was so the room became it’s own separate entity while the test was being done (also it just feels more natural to have a door). The fun and exciting thing is that I made it so that it greets you/says goodbye depending on which way you come from!
One thing that might be noticeable is that I didn’t add a second mid-level functionality for the elevator. After coding half of it and playing around with it, I realized that it just won’t work for many, many reasons: 1) The elevator itself blocks the view of the bars (even when mostly clear), and if it’s moved back it becomes too far away. 2) The mid-level actually hides and skews the differences between the heights. 3) It makes the visualization a bit too complex, and in a Tufte manner of thinking, it seems to shift too much focus away from the data and onto working the visualization. 4) It wouldn’t be used much/at all, since the point of this 3d visualization is to walk around and feel the data.
Basically, for the transportation side of things, all that needs to be done is create the questions/instruction sheet and then start the testing. Here’s my general outline for the test: There should be two sections: 1) Data hidden (for data retention), 2) Data present (data ease of use). Data hidden should probably go first (that way we can capture the first reactions to both), so I think the test would first be reading the instructions/getting acquainted with Second Life/website, and then being told to “play around” and look at all the data for 2 minutes. Then the monitor would be turned off and the first test given. After that, the monitor is turned back on, and the second test is given with the data available for use.
The questions for each test should have two sections, being: 1) Understanding, 2) Interpretation. There should only be a few understanding, which will be multiple choice questions like “What is the percentage of people who drive in a car alone in the Middle Atlantic?” a. 57.2%, b. 10%,…, “Which of the following regions has the highest comparative percentage for Car Pooling?”a. Middle Atlantic, b. Pacific, …. The interpretation questions should be more interesting/harder, such as “Which region seems to utilize the least amount of Car Pooling/Public Transportation?” a. New England, b. Mountain, “Estimating, where do you think the largest number of people use a Motorcycle/Other means of transportation in…” a. East North Central, b. Middle Atlantic, New England, …”.
Hopefully we’ll be able to progress soon.
For the lack of a better descriptions, things are plainly just going along. I’ve finally finished creating all the signs/pyramids for the “Compare Regions” half of the visualization! There’s basically nothing else to report on, so here’s what I need to do in the next 2-3 days to finish the entire visualization off: 1) Check/rearrange the room so every pyramid /sign is visible, 2) Click on every pyramid/bar and fix any display errors (output not matching sign), 3) Create the 45 signs for the “See Region” half of the visualization (which unfortunately might end up being easier if I delete all the bars and start over…), and 4) Edit/Add on to the elevator to accommodate two levels of sight.
Minor elaboration on the four points: 1) Becuase of the size/design of the pyramids and signs, the ones in the foreground will cover the ones in the back. Thus, the pyramids have to be specially arranged that way this problem can be avoided. 2) The size of the pyramids is scaled as such: take the compare regions percentage and scale it down by 20 (10 to convert into SL meters, 2 to make it reasonable in size). Unfortunately, this process sometimes loses a fourth decimal point in rounding, so I’m going to have to check over every pyramid and make sure it returns the right size when clicked. 3) All the bars still have the black SL “SetText” floating above their heads, so I need to convert those into the signs used in the “Compare Regions” area. The way it has to work, however, might make it so this is a very inconvenient operation. If this is true, I’ll basically have to recreate a couple days work and delete all the bars and recreate them with the sign default over their heads (it’s much easier to copy a bar/sign combo then create a new sign and place it over each bar). 4) Though the pyramids/bars are visible from the ground, a slightly higher perspective helps to make distinctions in the height. With that being so, two levels need to made for the elevator (the other being for the extremely high “Car – Alone” bars. I’ve decided the most intuitive approach is one of two things: first, clicking the elevator brings up a dialog box, which give you the floor buttons, and second, having the floor buttons printed on the elevator itself and just clicking them. If this option is so, I’ll either print “G 1 2″ or “1 2 3″ (I haven’t decided whether to start with the Ground floor or first in either option).
6/20/07
*Displayed final version of white male mannequin
*Made the pyramids/signs for all the NE, MA, SA, ENC in “Compare Regions”
Spending:
*$10 – Upload of ne22 image
*$10 – Upload of ne23 image
*$10 – Upload of ne24 image
*$10 – Upload of ne25 image
*$10 – Upload of ma22 image
*$10 – Upload of ma23 image
*$10 – Upload of ma24 image
*$10 – Upload of ma25 image
*$10 – Upload of sa22 image
*$10 – Upload of sa23 image
*$10 – Upload of sa24 image
*$10 – Upload of sa25 image
*$10 – Upload of enc22 image
*$10 – Upload of enc23 image
*$10 – Upload of enc24 image
*$10 – Upload of enc25 image
——————————————
6/21/07
*Created a couple images for the signs
*Payed for the white mannequin
Spending:
*$1000 – Payment for the white mannequin
————————————————–
6/24/07
*Created the rest of the images/signs/pyramids for the “Compare Regions” half of the transportation visualization
Spending:
*$10 – Upload of esc22 image
*$10 – Upload of esc23 image
*$10 – Upload of esc24 image
*$10 – Upload of esc25 image
*$10 – Upload of wnc22 image
*$10 – Upload of wnc23 image
*$10 – Upload of wnc24 image
*$10 – Upload of wnc25 image
*$10 – Upload of wsc22 image
*$10 – Upload of wsc23 image
*$10 – Upload of wsc24 image
*$10 – Upload of wsc25 image
*$10 – Upload of m22 image
*$10 – Upload of m23 image
*$10 – Upload of m24 image
*$10 – Upload of m25 image
*$10 – Upload of p22 image
*$10 – Upload of p23 image
*$10 – Upload of p24 image
*$10 – Upload of p25 image
So today was going a lot slower than I would’ve anticipated. Even though I anticipated this part of the project going slowly, I didn’t realize how slowly it would really be. Basically, to reiterate, I have to make little signs in paints and then upload them to my pyramids/bars. To save time as I make new pyramids, I’m also making the signs at the same time. Thing is, no matter what I do, this job passes really, really, realllllyyyyyy slowly. Honestly, I’m foreseeing myself being lucky if I finish the pyramids/bars by the end of the week now (not that I think it’s hard, it just takes forever). I’m not worried right now, it’s just a lonnng road ahead. So far, I’ve managed to make the signs/pyramids for 4 of the 9 regions of the “Compare Regions” area. Still to go: 5 pyramids/signs for the other regions, and then the 45 signs for the “See Region” bars already made. The past several logs:
6/18/07
*Lowered the room by 5m
*Built a path leading up to the room
*Added an elevator/standing object
*Built first group of pyramids for “compare regions” on
*Searched for and displayed a text display system
Spending:
*$500 – Text Display Device
—————————————————————-
6/19/07
*Deleted guardrail from platform
*Made replacement signs for text on first batch of pyramids
*Displayed the white male/american indian prototypes
Spending:
*$10 – Upload of ma21 image
*$10 – Upload of ma21 2 image
*$10 – Upload of ne21 image
*$10 – Upload of sa21 image
*$10 – Upload of p21 image
*$10 – Upload of m21 image
*$10 – Upload of esc21 image
*$10 – Upload of enc21 image
*$10 – Upload of wsc21 image
*$10 – Upload of wnc21 image
————————————————
6/20/07
*Displayed final version of white male mannquin
*Made the pyramids/signs for all the NE, MA, SA, ENC in “Compare Regions”
Spending:
*$10 – Upload of ne22 image
*$10 – Upload of ne23 image
*$10 – Upload of ne24 image
*$10 – Upload of ne25 image
*$10 – Upload of ma22 image
*$10 – Upload of ma23 image
*$10 – Upload of ma24 image
*$10 – Upload of ma25 image
*$10 – Upload of sa22 image
*$10 – Upload of sa23 image
*$10 – Upload of sa24 image
*$10 – Upload of sa25 image
*$10 – Upload of enc22 image
*$10 – Upload of enc23 image
*$10 – Upload of enc24 image
*$10 – Upload of enc25 image
Today was not spent like I thought it would be. Meeting up with Dr. Polack allowed her to express concerns with the visualization, which eventually got me thinking: before today, I really didn’t get the essential difference between displaying data in 3d and data in 2d. In 3d, everything becomes more real, for the good and the bad. Even though it seems really nice to be able to display data using real objects and better design, there’s a huge drawback: 3d limits you in some ways like real life! You run into problems like height, perspective, line of sight, draw distance, etc., when in 2d you get a “birds-eye” view of things you could never see in real life! Obviously one of this projects goals is to mediate the two, but before now I never realized why 3d visualization is a lot harder than it sounds.
Anyway, I got a bit done today. I moved my room 5m down, and then built a ramp leading up to it. Inside, I have two main displays for everyone to check out: first, on the “See Regions” area, there is now an elevator (which points you the correct way now), so you can see the (now contained) tops of the “Car – Alone” bars. Also, Dr. Polack expressed interest in having “signs” with the percentages, so for the “Car – Alone” I posted what one sign could look like. Basically, I bought a text maker program and wrote that in it. Unfortunately, I can’t do words very well with it, and the numbers have a predefined minimum size. As long as the signs can work out for both the big and small bars, I don’t mind making them that much. The only problem, however, is that I’ll have to make every single texture in paint (90!), and then upload every single one to second life onto a block. In other words, it will be a super tedious operation that could possibly take the focus of an entire day or two. Alas, if it’s best for the visualization, then that’s what really matters I guess.
The second (and bigger thing), is that I implemented the first of the “Compare Regions” pyramid visualization system. The first batch is a test, and works only for the “Bicycle/Walk”dataset . They look sort of neat (I don’t think they’re as good as the bars overall, but are a different way to display data, if for no other reason that to see how people respond to them). If Dr. Polack likes them and gives me the go ahead, I’ll probably start making the rest as soon as I can.
Oh, and one more thing before I forget. Shiv contacted me a day or two ago, and sent me the white person mannequin. Unfortunately , it doesn’t have any hair yet (he outsources that to someone else), but hopefully he can get that done soon. Also, he sent me two different versions of the American Indian mannequin, which I’m not really sure I’m completely happy with either one (shirt wise). I’ll put all three upstairs in the UMW area tomorrow afternoon for everyone to look at and give me feedback. Tomorrow should be quite the interesting day…
So it’s officially 4:40am, but I could care less at the current moment, because…my “See Regions” part of the visualization is completely functional! I’ve spent the last 4.5 hours on it or so, but it’s finally done. Thought it seemed like “auto-pilot” after my meeting with Dr. Polack, that only turned out to be half-true. Coming up with all the extra bars/data wasn’t so bad, but fine-tuning it turned out to bit more involved than I thought. Believe it or not, almost all my problems were created by the data itself! The pesky “Car – Alone” are figures are extremely high (they tower over everything else). Among the many problems they caused, here are a few: blocked other bars from being clicked, blocked vehicle choices from being picked, text was too high too read, won’t fit in the room. Besides the room part, I was able to fix everything else with a bit of ingenuity (i.e. inspiration from other people on Second Life lol). Because objects can be clicked even when phantom/clear, that was making those bars obscure certain choices. To solve this, I ended up moving all the bars 20m below my room into random space. This causes a .4 second delay in their appearance, but that’s definitely small enough to not really matter. Also, since the text is so high, I attached an invisiprim displaying the text halfway up each bar, that way the text is visible from the ground.
Unfortunately, I can’t do much about the room height right now; my room can’t be moved down for the time being. Apparently, a guy named Chris in one of Dr. Polack’s classes added a bunch of shields/floors for her. Unfortunately, because of permissions problems, I can’t modify/get rid of the extraneous ones in my way. Dr. Polack, if you could, please e-mail Chris ASAP and tell him to get rid of the extra walls right as soon as possible. I noticed he got rid of some, but not all the ones I need to be gone. Tell him I’ve left a box outside of my room with instructions as to what I want done with the area (and the reason I say sooner is better is that’ll it probably take me at least half an hour to an hour to realign 45 bars, 5 vehicles,5 vehicle bubbles/signs, the platform/rails, the signs, walls, etc., if not more time, and that’s not really what I should be focusing my time on).
Now that I’ve gotten all that out of the way, I’ve got a couple theoretical ideas to muse on. First, I think it might be a good idea to have some sort of “start” area that the person could walk to so they know where orignally look to see everything. I was thinking of something akin to a large red “welcome mat” type thing at the front of the floor map (otherwise, it might be very confusing to explain to the person where the optimal beginning standing position is). My second idea is, what if I also turned this mat into an elevator like device (like the on Greta has). The reason I suggest that is the pesky “Car – Alone” bars (again!!). Because they are so high, you can’t see the tops of the bars, so there’s no effective way to make distinctions between them beyond the numbers printed halfway up. However, a simple elevator could solve this problem pretty well.
Finally, I’ve been thinking about these “piles” to make for the “Compare Regions” choice. I really can’t figure out what they’d look like. Currently, the only thing I can imagine a pile to look like is a tiny pyramid, and I’m worried that wouldn’t be a very good visualization device. Would there be much distinction at all between 15% and 30% as pyramids? The 30 might be a bit more dense/taller, but the distinctions really aren’t anywhere near as close as the bars would be. And if instead it was less like a pyramid and more like an extended flat surface, I’m not sure the distinction would be very noticeable then either (since a lot of the eye is going to detect is height-related). Also, because of the limited prim count, I was wondering if it would be wise to have 900 prims sitting on top of my map just for piles…I think there’s a command to rez objects out of the inventory (and I know how to delete items) so I wouldn’t have to store them on the map at all times, but I’m still not sure how wise making the piles would be (the education one is going to burn up A LOT of prims due to the people objects). I think I might go ahead and try to make a pile tomorrow, but Dr. Polack/Greta/others, what are your thoughts on the piles? Do you still think they’re a good idea, or should I head in another direction? If nothing else, I could cook up more bars without extreme amounts of trouble.
Anyway, it’s about 5:17am, so I think I’m heading to bed soon. If everyone who reads this could provide me with comments/suggestions, I’d really appreciate it. We’re pretty close to testing time, so I want to make sure to get everything done (and well). Here’s my activities log over the past couple of days (which, by the way, if you haven’t noticed before, also provides an exact change log as to how the room looks/works day by day):
6/14/07
*Attended the ASL Scripting 203 class
*Turned all the vehicle signs into platforms
*Changed the coloring of the see region/compare regions signs to blue
*Created and coded the first set of bars
Spending:
*$20 – Tip for ASL 203 scripting class
*$10 – Upload of blue see region image
*$10 – Upload of blue inverse see region image
*$10 – Upload of blue compare regions image
*$10 – Upload of blue inverse compare regions image
—————————————————————-
6/15/07
*Lowered the platform and raised the map a bit
*Coded and implemented the communication between the choices and bars
*Changed the coloring of the see region/compare regions signs to blue
*Got rid of the platform
*Turned the floor map into the standing area with white surrounding region
*Built guardrails for the map standing area
Spending:
*$10 – Upload of navy see region image
*$10 – Upload of navy inverse see region image
*$10 – Upload of navy compare regions image
*$10 – Upload of navy inverse compare regions image
——————————————————————-
6/17/07
*Added in all of the data for the cylinders of “see regions”
*Made it so the cylinders instant message the person who touched them instead of public chatting
*Changed the motorcycle sign to now read “Motorcycle/Others”
*Made the motorcycle platform a bit bigger and moved the car one over to the right to accommodate for the new text
*Made it so the humongous “Car – Alone” bars disappear under the screen to prevent interference
*Created extra prim objects to display “Car-Alone” text halfway up the bars
Spending:
*$10 – Upload of white motorcycle others image
*$10 – Upload of white inverse motorcycle others image
