iwobot in happy state

iwobot in sleeping state

First possible design I came up with, is a TV-like robot with basic body parts which can help iwobot to have human-like gestures. It has 2 legs, 2 eyes, a mouth a screen and hair with a hair clip ^^

I thought the screen can be a data transfer interface between the robot and the user. So in case the user asked iwo what her favorite picture looks like, iwo can actually show it to the user by putting the image on her screen.

Anyway,  it’s Feedback time!

The Idea
I want to make a virtual friend (iwobot) who can:

• act as a usual interface between me and the people who are visiting my website
• be a guide for the visitors of my website in the pre-defined cases
• let the visitors of my website interact and play games with him

Functionalities
Here are the use cases and functionalities I want my robot to have:

1. iwobot will use 1 twitter account (iwobotTwitterAccound) as a data feed
2. The user will be able to feed the robot with input text through the user interface
3. The user will be able to choose actions for iwobot to take (like jump, walk etc.) by clicking on buttons in the tool box of the application
4. The user will be able to interact with the robot by giving him objects (by dragging and dropping an apple from the tool box to the canvas where the robot is drawn)
5. There will be pre-defined questions and answers set, like a Frequently Asked Questions, if the input text the user enters looks like one of these questions or simple commands, then they will be interpreted right away by the iwobot engine and the user will be able to see the reaction quickly
6. Otherwise the idle animation will play until the admin receives the question in his twitter account and answers it (The idle animation may be thinking or sleeping etc.)
7. The text that the user entered (as told in number 2), will seemlessly be sent to iwobotTwitterAccount from iwobotTwitterAccount if the engine couldn’t handle the message itself

Pixel Bow

I made an image viewer using Flex and Actionscript 3 where you can view the images as thumbnails and select an image and annotate it and even download that image with your annotations on it or save it to my server. Since I can view these images later, this can be a good way to send me a secret message ;)

Anyway here you can view my pixel bow application or download it and embed it in your website. Or even better, you can download my source code and play around with it.

Your comments are appreciated :)

Me and my project mate Mies developed 2 different versions for this assignment.

• Version 1 – uses the GeoIPService, the Ruby programming language, Sinatra, soap4r, the jquery Javascript framework and the Yahoo Maps API and can be found here:
  iwa-wsdl2ruby
• Version 2 – uses the Fraudlabs IP2Location webservice, the PHP programming language, nusoap, Javascript and the Google Maps API and can be found here:
  DisplayGeographicalLocation.html

Version 1:
Since Version 1 was fully developed by Mies, I will not post the source code for this here. Maybe he can give a link to his code if he wants. So here is the explanation for the version 1:

The GeoIP WSDL-to-REST service is built using Sinatra, soap4r, the Ruby programming language, Yahoo Maps and the jquery javascript framework. Sinatra is an open source web application framework and domain specific language written in Ruby. As opposed to other web frameworks such as Rails (also in Ruby) or Django (written in Python), Sinatra is very minimal in its design and implementation. URL Routes in Sinatra are HTTP methods combined with some pattern matching and logic. The simplest URL route and thus web application in Sinatra is:

get '/' do
"Hello World"
end

which basically prints out “Hello World” if invoked via the web server using the root url (http://localhost/). As such Sinatra is ideal for building small REST services, API’s and web hooks, and thus a perfect candidate for completing this assignment.

All application logic resides in the ‘main.rb’ file. Using the soap4r library (an implementation of Soap 1.1 in Ruby), a request object is created using the provided .wsdl file. The IP address as provided by the user is passed as parameter to the getGeoIP method of the request object. Second a result object (in fact a key-value pair, or dictionary) is created using the getGeoIPResult method, invoked upon the same request object. The actual country name is the value of the CountryName key within the result key-value pair.

Two forms of URL routes are provided, one for the html representation and a second route for the REST API. The html representation is a simple .erb (a ruby templating language) template with a form. The IP address can be submitted and the result is displayed.

The second URL route provides the methods for the RESTful API. The following methods are supported: /api/location.?ip= where is the desired output format.
Current options are .xml and .json

For instance the url:

/api/location.xml?ip=62.69.184.53 will render a response NETHERLANDS

whereas

the location.json?ip=62.69.184.53 equivalent will render a {”CountryName”:”NETHERLANDS”} response.

The ajax response and yahoo maps calls are handled using jquery; a javascript framework.

Figure 1 displays an overview of the application’s architecture:

The final web application is deployed using Heroku, a cloud-based hosting environment for ruby applications at http://iwa-wsdl2ruby.heroku.com/. The initial version used the Google Maps Geocoding API for requesting a latitude and longitude and subsequently render the actual map. However as Heroku uses a reverse proxy (Nginx) and http caching layer (Varnish) to service and load balance requests, and being able to dynamically scale deployed web applications, the IP address used for this web application is not static. As such the Google Maps API key can not be coupled to this web application and thus an error message appears and the actual map is not rendered. The Yahoo Maps API is less strict in this sense about its usage and thus we ended up using this service in rendering the final map.

Version 2:
This version of the application makes use of Javascript in the front-end and PHP at the back-end and only works for IE  for now. This can be avoided by using a framework like prototypejs.

When the user browses to DisplayGeographicalLocation.html 2 new Ajax objects are created as the page is loaded(one for getting IP asynchronously and the other for getting the geographical location asynchronously).

The logic of the application resides in ipToMap function. If the ip input text field is empty, it makes an HTTP GET request to getUserIP.php which returns the user IP to the client.

The Javascript code for this would look like this:

The PHP code for getting user ip is as below:

After the ip is successfully assigned to the ip input text field, a new http GET request is made to getGeoLoc.php and ip is given as a parameter.

This php file makes use of nusoap library in order to create a soap client using the WSDL file provided by fraudlabs: ip2locationwebservice.asmx?wsdl and to call the fraudlabs IP2Location webservice with a message including the ip and licence key needed to use this webservice.

A sample call to getGeoLoc.php and its result would look like this:

And the source code for getGeoLoc.php which turns WSDL webservice into REST.

The result is then processed and turned into XML format within getGeoLoc.php and returned to the ajaxObject as a responseText. This text is than recognized as an XMLDOM and parsed and the latitude and longtitude are retrieved. These latitude and longtitude variables are then used for creating a map using the Google Maps API.

Please note that the free version of fraudlabs includes only 90 calls to their webservice per month, so when the credits are finished the result will not be accurate, it will return a latitude of 0 and longtitude of 0.

These examples are bound to external services though.
In practice, if it is a crucial functionality for your projects, I would recommend keeping the ip to geographical details on your database and write a service of your own, so that when these external services are down or you are out of credits, you wouldn’t be angry at other parties :)

Our first assignment for the Multimedia Authoring course was to make an interactive logo which navigates back to the home page when clicked. Nothing special, just an exercise to get familiar with Flex and AS3.

Since it had to be a portfolio page in the end, I wanted to use my name and initials. And while doing this I wanted a mascot for my logo, which happened to be an owl by using 2 different fonts. About the effects I used, I move the eyes of the owl according to the mouse position. And add blur effect when the logo is clicked.

You can download the source code for my logo component or check out the demo flex application.

Oh, and the documentation of my code is already embedded as comments in my files. Let me know if you have problems running the application or have any questions related..

• my school assignments / projects
• things I’ve learned / discovered during my studies
• may be some tutorials
• and sample code you can reuse..

Click here to view the embedded video.

Anyway, here is the final edit of the video, so here you can watch and comment on it. So that I can make it better. And thanks for watching!

Can’t see the movie? Download it here

Think of 24.500 people, visiting a small village which is decorated with various fantastic themes, dressing and behaving in anyway they like for two days. Anywhere you look, there is always something interesting to see. Lucky that I had my camera with me, so that I could make a short video out of it.

Elf Fantasy Fair takes place in Haarzuilens(The Netherlands) every year around April. It’s a fair where people go and have some fun, listen to some concerts, shop for all kinds of uncommon things they can’t find in their daily shopping trips. People dress as they like, behave as they like. It is a nice weekend to escape reality.

Can’t see the movie? Download it here