So below is a piece from Tim Stutts at OpenProcessing.

The idea is, you go and click to a place that you have been. The guy in the application goes to that point and turns back to ITP on the google map screen shot, leaving a trace behind. I am going to put it on the floor to collect real data, but till then, here are some screen shots with random points clicked, and the emergence of the patterns through out time (shown without the map).

Just a beautiful side of visualizations, it was very interesting for me to see that the auto-random-wandering guy never crosses Greene Street and Mercer Street. First, I though it was some kind of bug, but then I figured out that, seriously, we don’t need to cross these two streets since the places one would go from ITP are either below or above these two streets; we never need to cross these streets if we are turning back to ITP.

In fact, creating the algorithm for the guy to find its way through the streets to reach its destination had been quite tricky, and kept me awake till 5am this morning (oops, daylight saving time.. so I guess it was 6am when I was in bed). But simply, the algorithm is:
the boy (boy.class) checks the adjacent corners and its own corner to see which corner is closer to the destination point
if it is the corner he is currently at
go to destination point directly
if it is not/if an adjacent corner is closer
go to that corner

I loved this piece, and trying to get it work as an applet on the webpage. Then it will be up for you guys to try.

Why a user should spend time on being a part of OpenVisuals?  This also brings more complicated questions: why she should spend time tweaking her visualizations with the API? Will she be interested enough to make her visualization working on the website?

I have this given as the mission in the front page :

OpenVisuals aims to provide an Open Source Visualization Framework, in which the datasets can find their own visualizations, and visualizations already out there can be shared to be used with other datasets. Dataset owners and Visualization artists - datasets and visualizations - in the same pot.

I wonder how it stands with respect to Linus Torvalds’s proposal for Linux - a new but small operating system, undertaken principally as a way to learn together -.

improving communications as group activity -  “suggest a visualization to this dataset” feature, “post this dataset to the ’singles’ wall”

In reference to the section of Caterina Fake, greeting first ten thousand users in flickr, I did have my first registered user, but there is nothing I can do to greet him. He didn’t post any visualizations/assets, hence nothing to comment on. So getting a user registered gives them just an empty portfolio page. So I guess something is missing here.

On small/large groups issue, the initial idea is having a large group interaction within OpenVisuals. But since last week, I have been trying to implement a way to get future Nature of Code applications to be posted and gathered to have a small group interaction. It wasn’t the way I was looking for, but this opportunity of Nature of Code class posting their stuff itself questioned some small group interaction in the website? Can Nature of Code class students have their own space in the website? My current solution for that is to having them to tag their stuff with Nature of Code, so that in the browse page they can focus on Nature of Code works. (actually, I even added rss function, so that they can subscribe to the feed of that tag)

• What did you do last week?
  • I implemented that tagging function for visualizations, so any visualization uploaded can be tagged (not sure about datasets yet). Using these tags in the database, I created the new browsing page, which a user is able to search the visualizations through tags. I am going to replace the browse url soon, but till then, the url is www.openvisuals.org/browse/index2.php, otherwise it is www.openvisuals.org/browse
  • I also implemented some security measures, because I had a user who found out the way the delete all the visualizations and datasets within the website (and took the related action, of course). This had been an interesting issue because it disproved one of my assumptions; I was thinking that till a website starts getting an amount of users, the developers don’t necessarily need to spend time on dealing the security of the website and database (preventing users to access some functions). I was thinking it could be a secondary priority.
  • Actually, after I saw some user was deleting all the assets from the website, before implementing the security stuff, I implemented logging. I wanted to observe how he was behaving. I was able to log his deletion in the morning of March 1st. Then I took that portion of the logging table, and posted in openvisuals as a dataset. Maybe I can visualize it one day..
• What do you intend to do next week?
  • I am going to attend the Nature of Class, thought by Daniel Shiffman, on Wednesday, and I will present OpenVisuals to the students of that class. They are working on quite interesting processing applications, any of which can be a visualization candidate. I will invite them to use OpenVisuals to upload their applications, share it with their friends through the website. OpenVisuals API is not ready yet, so I am not going to ask them to play with it, unless I can achieve to post a stable version of the API.
  • I am going to work on API, and going to convert couple of my processing applications to visualizations using API, and upload them to the website.
  • I will think about not storing the datasets issue.
  • I will try to put the reference page up, explaining the functions of API.
• Any recent concerns and unexpected obstacles
  • one of my concerns is the datasets. Currently storing the datasets within the website brings many issues:
    • Uploading datasets requires a good troubleshooting feature, because csv files might not be properly setup, field types (text, date, integer, etc.) is a messy thing to deal with, and some couple of more technical issues.
    • It makes OpenVisuals a rival with Swivel, Many Eyes. One of my principles reminds me that ‘every competition brings opportunity for collaboration’. I started thinking on making OpenVisuals just a visualization counterpart of Swivel and Many Eyes, which the datasets from these websites can be integrated into OpenVisuals and be used with the visualizations within. So that those websites deal with the datasets part, OpenVisuals deal with the visualization part.
• Any recent insights and surprises
  • I have implemented Google Analytics and I am excited to observe user behavior through the website.
  • I started to question the dataset thing on another aspect. When I look at the visualizations around, their source are not necessarily data tables. Take the visualization of Lance Armstrong’s last tour, its source is hard to have on an excel table. There are many examples like that. Will OpenVisuals include such visualizations, and if so, how some other person can use his own source (not dataset in this case)?
  • I had one non-ITP, non-New Yorker user registered already. It is nice to have an interested person, although nothing in the website is in shape. I searched him on google, and he come out to be a visualization artist. This makes me smile :)
  • I had an attacker. This also makes me smile. His attack was quite a nice user feedback indeed.
  • In a google search of “open source visualization”, OpenVisuals shows up in the second page. This is quite nice too, considering the very limited number of hits (hence its popularity) it is currently taking.

Touchies is my experimental project for Telepresence Class. Touchies, (two cubicles with a soft section to finger) (soft section is made out of toilet paper) (how lame!), allow you to send your touch of finger real time to another touchy. You can send your warm touch to your lovely who is far away from you. Or even E.T. doesn’t need to travel million light years to touch your finger anymore! Touchies are here!

Touchies from Sinan Ascioglu on Vimeo.

• sample a generic type of datasets.
• should have different but generic formats, which would be a format outcome of copying/pasting, exporting from excel, etc.. (right now, I can think of differences in comma separated values (sometimes comma, sometime tab, sometimes space).
• should imply comparison among its data.

I will use these datasets to make sure that the website (hence the applets) will be able to work fine with them. I currently fetched couple of datasets that would be suitable:

MDF (wood type) Production for Continents per years (gathered from FAO United Nations website)

years	subject	commodity	America +	Asia +	Europe +	Oceania +	

1995	Production Quantity	MDF	2398000.00	1582000.00	3363300.00	540000.00

1996	Production Quantity	MDF	2725000.00	2190000.00	3604300.00	786000.00

1997	Production Quantity	MDF	3285036.00	3737000.00	4587300.00	853000.00

1998	Production Quantity	MDF	3899694.00	3484000.00	6557913.00	899000.00

1999	Production Quantity	MDF	4635000.00	3854000.00	7288500.00	1025000.00

2000	Production Quantity	MDF	4773000.00	4652000.00	8380493.00	1241000.00

2001	Production Quantity	MDF	5101472.00	8002300.00	9163590.00	1350000.00

2002	Production Quantity	MDF	5601924.00	10135100.00	10386690.00	1419000.00

2003	Production Quantity	MDF	6327229.00	13970200.00	11110990.00	1627000.00

2004	Production Quantity	MDF	7729545.00	18658026.00	11846105.00	1639000.00

2005	Production Quantity	MDF	7751384.00	24133523.00	12704400.00	1622000.00

2006	Production Quantity	MDF	8467770.00	27017523.00	13383800.00	1635000.00

In the table above, the generated dataset set is not delimited with comma, but tab spaces (although the file is .csv). This also happens when you copy paste from Excel.

Also below, is a very typical example of a dataset with empty cells (unknown data), though the content is one of the most popular ones lately.

"State","Barack Obama","Hillary Rodham Clinton  ","John Edwards","John McCain ","Mike Huckabee","Mitt Romney","Ron Paul"
"Alabama",10,19,0,16,21,0,0
"Arizona",21,25,0,50,0,0,0
"Arkansas",3,10,0,1,29,1,0
"California",155,195,0,146,0,3,0
"Connecticut",26,22,0,27,0,0,0
"Delaware",9,6,0,18,0,0,0
"Georgia",22,12,0,3,45,0,0
"Illinois",68,35,0,54,0,3,0
"Kansas",15,6,0,,,,0
"Massachusetts",38,55,0,18,0,22,0
"Missouri",36,36,0,58,0,0,0
"Montana",,,,0,0,25,0
"New Jersey",46,54,0,52,0,0,0
"New Mexico",12,14,0,,,,0
"New York",80,121,0,101,0,0,0
"North Dakota",,,,5,5,8,5
"Oklahoma",14,24,0,32,6,0,0
"Tennessee",14,24,0,19,25,8,0
"Utah",14,9,0,0,0,36,0
"West Virginia",,,,,18,0,0
"Florida",,,,57,0,0,0
"South Carolina",25,12,8,19,5,0,0
"Michigan",,,,6,1,23,0
"New Hampshire",9,9,4,7,1,4,0

Here is also the imported version of this dataset in OpenVisuals.org .

Open Source Visualization Framework.

I have a dataset -> pie charts is not enough to visualize it..
I go to openVisualz.org -> pick a suitable visualization -> submit my data -> customize the view -> put the result to my blog…

I am a programmer and produce data in my application in realtime (eg. I read sensor input from Arduino) -> I want realtime visualization -> I import the openVisualz.org library -> submit the data real time to the openVisualz.org -> it is visualized on the website in real time.

I am a visualization artist -> I create and share a visualization of (eg. movie dialogs) using the openVisualz library -> I share it at openVisualz.org -> some media company is interest in using it, but needs to customize it for their own purpose -> they hire me to improve the visualization for this specific purpose.

After giving a second though, I started to think a revision is necessary. First comes first, I would take out the realtime part, since the other parts are complicated enough to work on within 4 month’s time.  And project needs more focus.

This user scenerio is not exactly a statement, so here I should devise a statement. So, new version is:

Open Source Visualization Framework. A website which provides dataset users various open source visualizations which are created and submitted to the website by visualization artists. Framework provides various APIs, to let visualization artists to tweak their visualizations to make them ‘OpenVisualz’ compatible; compatible datasets can be imported and applied to the visualization in the website.

so this is the jump start point of the project. the first post. done.

Having said that, here is a project proposal for Telepresence class:

Do you have your loved one away? Are you bored of his business trips? Are you most often sleeping alone in your king size bed?

PillowSolus is your companion in your bed, giving you the real time presence of your partner right next to you! The pillow, which your loved one uses when he/she is away, transmits the owners head position to the pillow next to you, which shapes itself as if your partner is sleeping right next to you! You will see his/her sleeping, waking up, turns in the bed, stress, and maybe even sweat! (Simply, one pillow is transmitting the position of the head, and the other is taking the shape as if it has the head in the same position)

Anyway, I prepared official web pages for the projects:

Eyes on You, Tarantino!  - http://www.wiredpieces.com/projects/eyesonyoutarantino

John Train - http://www.wiredpieces.com/projects/johntrain

I will probably post further developments there, so,   so,   this might be the last post for this semester for the blog….. omg…. the end is coming….

Both JohnTrain and EYT got accepted to ITP Winter Show! See you there:

http://itp.nyu.edu/show/winter2007/

“Let’s get physical, physicall, ooo”

EYT prototype on the column drawing from Sinan Ascioglu on Vimeo.

There is very little left to do on the software side, and weekend is getting ready for building the physical controller! (I call it ‘EYT columnHugger’)

A screenshot from the software is below. So what is new? Now three movies are displayed, which the user can switch between. Also, their eye tracking plots are displayed below them for easy comparison! Actually that is the point of all project; to see and experience the difference in eye movements among these movie scenes!.

Click on the image to see the actual size.

EYT project prototype from Sinan Ascioglu on Vimeo.

EYT project prototype from Sinan Ascioglu on Vimeo.

I will post upcoming codes under this post I hope. Explanations can be found in the files.

eyeCapture3XML.pde - processing code