4sep13 From Letter to Zach et al Yes (and pretty soon we'll have to move this and other mentor discussions to some common BBS or listserve .... suggestions?), I plan to make the HTML5+CANVAS (we need a catchier name) a project (see below). The overriding "lesson plan" goes something like this: The original Hippocket Programs had two purposes: 1. Explore the graphics capability of the given thing (AppleII, C64, OpenGL, TkInter, Python, HTML5+CANVAS). 2. Illustrate an interesting mathematical idea at the same time. Because humans perceive and depict 3D (and higher) first and foremost in 2D by drawing pictures (line drawings and polygon filling) its appropriate to do 2D stuff first. If you can animate 2D drawings, you can draw in 3D. For example the Sierpinski Triangle. Its graphical hypothesis is a rectangle subdivided into square cells. Each cell contains a (small) amount of data, and displays it (usually) by a color. The metaphor is a piece of graph paper and a box of crayons. The mathematics is cellular automata, specifically Barnsley's idea of Iterated Function Systems (IFS), and the classical concept of Haussdorff dimension (fractals). My plan last summer (which failed because of my back problems and other interference) was to rewrite the document "BASIC Pocket Graphics Programs" for Python and H5C. It's current (1997 !!!) text was such a translation from the Apple2 BASIC and graphics to C/IrisGL and then OpenGL. The project can be structured horizontally or vertically. Either a team of qualified and interested students work on this together, each taking one of the pocket programs (old or new) and writing a "chapter". Or it could be a solo semester project for one or two people. The former is appealing because it would speed up the process and produce a workable scheme for "publishing" other projects. Like some of the more interesting DPGraph homeworks already handed in. [see below] The latter is appealing because with effort and ingenuity (198 students often have really good ideas I wouldn't have thought of) the PocketH5C project would be publishable. Enough for the moment. Concerning the DPGraph micro projects. To integrate this exercise into the course all three of us have to have access to them. On the other hand, it will be a while before the entire class, especially the novices, can handle it. The model I have used in my courses which has either TAs and or mentors attached to them, is to use a second SVN repository, called mentor198. This is accessible to me, to the "staff" of a course, and to some other people working with me on my courses. Such a private repository can be very effectively used to keep track of stuff for MA198, especially when it is still under construction. It's also where we can deliberate about which students go into what mentoring groups. What do you think? Zach 1. The DPGraph homework is the first work submitted. From it you can pretty much tell who's going to be do what in the course. Several additional exercise will confirm or vacate the impression. Since you 3 will be working with students on just about any level they request (within reason) you do need to know how they work. That's why you'll need access to the DPGraph homework. 2. Yes, let's meet at 2pm like last week and hash this out a bit better. (I know Gillian will be bit late, that's fine.) =========== If we use mentor198, I have to 1. clean it up 2. sequester private information i.e. not for the mentor's eyes. Maybe figure out how to revive/archive better. Manton is doing a class198f11, class198f12, etc. Alternatively: Use only the repo. Have the students put their private contributions in to their own folders. We can transfer stuff to the commons if distribution is advisable. We can use vpeters directory to discuss private class matters. Difficulty with this is: 1. There is hazard that she might access it. 2. It the repo gets messed up, our stuff is also interdicted. 3. It makes the repo heavier.