Social Group

5-4-2008

I had the wonderful opportunity to go purse shopping for the first time in my life today. After three and a half of looking around Chinatown, Anaid and I sat down and decided the best game plan for which purses to buy. It came down to two options: high-class and high-fashion bags that would be worn by snobs, or extremely different bags that would have very stylized personalities and would show more social group contrasts. I decided the latter would allow me to make it obvious the bags are reacting to each others proximity and are networked, and are not merely playing audio.

Here are the current bag personalities I’m working with:

Grunge - a silver bag I’m going to spray paint, tape, and rough up
Diva - a shiny olive colored Versache bag (Versache logo has been removed)
Cocktail - very small pearl colored bag with a solid metal strap and dangling thingies
Girly - pink leopard print with a pink feather boa
Colorful - red, orange, and yellow thin fabric bag with a flower that I’m going to add beads and other colorful pieces to

5-2-2008
I’ve decided even though a networked structure of bags that contain a narrative structure and story line that continues throughout the bags may be my desire, that’s not my best option knowing the venue I have next week - the Spring Show. The show is far too loud to be able to hear speakers inside of the bags, and having headphones means the participant can either only hear their own bag, or I have to play mixed-down versions of all conversations possible to get the back-and-forth interaction of the bags. Instead, I’m going to use the bags as personal relationship items to the person holding them and play audio through headphones in the shoulder strap(s). This way I can also make inferences about the other bags and their “owners”.

all hate each other - truly proves the “bitchy” point: a drawn out narrative is too subtle for the environment of the ITP show
bags act as fashion conscious: tell person what to do (”oh my god that purse is hideous - just say you like it” or “green? they think that looks good with those pants?”)

2 voice options:
each have their own voice based on bag
all are same voice (would be able to upload all same albums and just not use some)

4-28-2008

In an attempt to figure out the best way to communicate with all of the nodes, here are the options I have come up with (assuming iPod over serial works).

Creating the interaction:

Work with 8 digit code in Arduino for proximity to nodes:
(leaving 4 extra spaces in case i want to expand the network later)

bag1/bag2/bag3/bag4/etc
2 = near node (low signal strength)
1 = next to node (receiving high signal strength)
0 = not receiving signal from node

Bag 1
22220000 = high all 4 nodes
21110000 = low to all 4 nodes
20000000 = alone

Bag 2
22220000 = high all 4 nodes
12110000 = low to all 4 nodes
02000000 = alone

Bag 3
22220000 = high all 4 nodes
11210000 = low to all 4 nodes
00200000 = alone

Bag 4
22220000 = high all 4 nodes
11120000 = low to all 4 nodes
00020000 = alone
Example:
120120000 = near bag 1, next to

4-21-08

After meeting with Clay I discovered although my Flash prototype was extremely helpful in helping me realize I needed to determine the number of connections and clips to use, it wasn’t entirely useful for my audience - namely being exhibited in the Spring Show. Having 3 “zones” is less useful than identifying event-based connections and playing audio clips by these. I’m currently working on using either three or four (3 bags and the base station/computer) nodes. Here is a diagram of the possible connections.

Now the work is in deciding how often the sound clips will run (Clay suggests 15 seconds as my audience won’t have the bags for over 2 minutes) and how many will be based on actions moving from one connection to another (for example if there are three bags together and one leaves and they start talking about that bag).

A couple quick thoughts: I think it makes sense for the node to act the same if its independent from all other nodes, no matter if they are connected at the time or not. Secondly, from the complexity of having four nodes I will probably not worry with trying to have 2 vs 2 interaction in this version.

4-16-2008

To make sure I practice what I preach I’m working on prototyping and doing some testing for the interaction of the bags. It really is amazing how just a little testing and extra thought in the foundation levels can make a huge difference.

Some things I’ve already come to find out or hadn’t thought of:

• ability to focus on one node as the primary, and the other nodes’ relationships fall in accordingly
• there are many more possibilities for interaction than i had originally thought.
• units can relate to each other differently if within different signal strength areas (i.e. if 2 are in 1 and one in another, or all three in three separate)
• because the sound clips are time delayed and aren’t needing exact readings, having the varying signal strength from XBee should be okay
• adding more strength areas would be simple, but quickly adds complexity in other areas (number of connections, sound clips needed, etc): I need to be very specific about whether adding more nodes/bags with fewer conversation (7 bags that don’t interact as much), or adding more cross-talk between fewer bags (3 bags that communicate a lot) would present the project better

This quick drag-and-drop visualizer simulates bag 1 as the primary/focused node (it doesn’t move) and the other two’s relationship to it. This is done under the assumption the using the values from the signal strength of the XBees are broken into chunks of value: no strength, low strength, and high strength (example: via analog this would be 0-600, 600-1200).

4-14-2008

Looking at the connections with adding multiple nodes to the social network. Adding more nodes obviously makes it incredibly more complicated (a la Clay Shirky), but I also notice it’s easier to focus on one particular node if you begin taking away some of the more extraneous connections between nodes. Basically, if I’m wanting to focus on one node and not worry so much about how much some of the other nodes interact with each other as much as how much they interact with the focal point, this is the way to go. But if I’m wanting all weighted evenly, I should stick with only a few (3) units.