Search

upswing, Computer science teachers continue to strive for new examples and problems to interest millenials. The Media Computation approach (Guzdial 2003) has proven successful in attracting students in contexts from community colleges to R1 universities - students are clearly excited by writing programs that make images. In this project, take Media Computing in new directions: we have students build images by interactively scripting an application, which means that they can more easily experiment and mix work that they create "by hand" and work that they create "by programming"; we work collaboratively with studio art faculty, so students build works that must meet underlying design criteria; we teach using the workshop approach, so most classes involve students working in small teams on a set of problems; and we use a multi-paradigm approach - students make images using functional, declarative, imperative, and object-oriented techniques.This approach has proven quite successful - our enrollments are up (at least 33% in CS1, 50% in CS2). We are attracting more women (now 1/3 of our students). And comparative data suggest that our students gain significantly more confidence in their abilities than do student in other introductory science courses.

Consider a middle-school student "playing" with a computer. What does she do? She may chat with friends. She may surf the Web. She may even play a few games. But she is also likely to make things with drawing programs, word processors, or other applications that make things. Can one leverage that creativity to encourage her to want to learn to program? Projects from Alice (Moskal et al. 2004) to Scratch (Maloney et al. 2010) have certainly shown that tools that tie creativity and programming can encourage students to think algorithmically. Despite the success of these programs, there is still room for new models that link children's artistic interests with programming. In this project, we explore the link between creativity and programming through a different model, one in which the drawing application "self discloses" underlying commands (DiGiano & Eisenberg 1995). As students draw images with the GNU Image Manipulation Program, a Photoshop-like application, a small window displays code that represents their actions. For example, when they draw a line, a "draw line" command appears that includes the coordinates of the line; when they change brushes, a "set brush" command appears that includes the name of the brush. Eventually, students refer to these commands to redraw their figures, to copy their figures, and to change their figures. Slowly they move from copying and pasting code, to modifying, to writing their own instructions, and beyond.

Computer science teachers strive for new examples and problems to interest millenials. The Media Computation approach has proven successful in attracting students in contexts from community colleges to R1 universities - students are clearly excited by writing programs that make images. In this workshop, we show how to go a step further and have write scripts and plug-ins in Python for open-source graphics programs such as the GNU Image Manipulation Program and Inkscape, open source alternatives to Photoshop and Illustrator. With this approach, students not only make images, they also write filters and features that they can share with others, even with non-programmers. For example, students have written filters that "fractalize" vector graphics or that turn images into something that looks like stained glass. The ability to make programs that others will use is particularly motivational.We will explore ways in which students can interact with both Inkscape and GIMP, from working with interactive programming consoles to writing plug-ins that can easily be shared with others. We will also consider the ways in which different models of image making can be used to motivate the major programming paradigms.The approach has been used with Python, Scheme, and C and for everything from a one-week module to a full course.