Mailman 3 May 2006 - Edu-sig

Fwd: [Only Python] 5/14/2006 09:29:20 AM
by Andre Roberge May 24, 2006

May 24, 2006

Hi everyone, I got the following message on my blog, aroberge.blogsot.com (more below). -------- Forwarded message ---------- From: George Wright noreply-comment(a)blogger.com I installed Crunchy Frog 0.3 and ran the included tutorial. Really liked the DocTest approach. *It forced me to write my own programs that would produce the desired results *. Up to now I had contented myself with simply reading and understanding other people's programs. I need more python tutorials like this. Where can I find them? ===== Given that I had included docstrings as an example, based on comment by others on this list, comments to the effect that some wanted to write such docstring-based tutorials, does anyone have any suggestions? Andre

3 8

Crunchy Frog 0.3 released
by Andre Roberge May 24, 2006

May 24, 2006

Version 0.3 of "Crunchy Frog" has been released. It can be found at https://sourceforge.net/project/showfiles.php?group_id=125834 Crunchy Frog transforms a "traditional" Python tutorial into an interactive session within your favourite web browser. Four modes of interaction are currently possible: 1. One-liner, similar to the typical instruction at the Python interpreter. *NEW* 2. Embedded Python shell, within an html page. This is a bit similar to Ian Bicking's HTConsole. 3. Multi-line Python code, as entered in a traditional editor, and executed by Python. 4. Solutions to doctests, as described by Jeff Elkner in a post on edu-sig. Even though it is an early release, Crunchy Frog could probably already be useful in a classroom situation. To use it, you will also need to have installed Elementtree and CherryPy. Suggestion: read the 6 page long "tutorial" included, using your favourite browser. Then, start Crunchy Frog, and go through the tutorial again. The whole thing should be doable in approximately 15 minutes (20 for Kirby, according to a previous message ;-) ... if you don't spend too much time playing with the new interpreter. :-) André

2 2

[edu-sig] Hi from Summer of Code
by Johannes Woolard May 24, 2006

May 24, 2006

Hi! I think I've posted on here before, but I though I would introduce myself properly: I'm Johannes Woolard, from the UK, and I've just been accepted into the google summer of code project to work on interactive python tutorials. My mentor (I believe) is Andre Roberge, and my work will be based on crunchy frog. For those of you interested, my proposal is currently in two parts: the main body at http://users.ox.ac.uk/~orie1763/coder-python.pdf and a revised schedule at http://users.ox.ac.uk/~orie1763/coder3.txt (The schedule actually describes everything I want to do). I've also set up a blog at http://pytute.blogspot.com to write about my progress. I look forward to working with everyone here, yrs, Johannes Woolard -- Johannes Woolard, Oriel College, Oxford OX1 4EW 07837880296 ox extn: 26611

2 1

RUR-PLE 0.9.8.2 release
by Andre Roberge May 24, 2006

May 24, 2006

A new version of RUR-PLE has been released. Actually, there's been quite a few releases since the last announcement on this list. Changes that have been made since the last announcement (from memory) are: 1. auto-indentation of code 2. possibility of importing user-defined images for the robot 3. improved code-highlighting and stepping through 4. spaces are visible (helps beginner with indentation, and avoiding partly blank lines which cause "invisible" syntax errors). 5. print statements in robot programs are now possible, with output appearing in special "window" 6. input_int() and input_string() instructions in robot programs allow to pass values to robot; this is not documented anywhere (yet) in the lessons. Enjoy! (and feel free to criticise) André

1 0

[Ann] DocTest solution: Crunchy Frog
by Andre Roberge May 23, 2006

May 23, 2006

Version 0.1 of Crunchy Frog has been released. It can be found at https://sourceforge.net/project/showfiles.php?group_id=125834 Crunchy Frog allows to transform a "traditional" Python tutorial into an interactive session within your favourite web browser. Three "modes" of interaction are currently possible: 1. one-liner, similar to the typical instruction at the Python interpreter. 2. multi-line Python code, as entered in a traditional editor, and executed by Python. 3. Solutions to doctests, as described by Jeff Elkner in a previous post. Crunchy Frog is more of a proof-of-concept than a serious application at this point. The code is rather messy, showing my total lack of experience with this type of program (web app). To use it you will also need to have installed Elementtree (included with Python 2.5?) and CherryPy. You do not need to install Crunchy Frog. To try it: read the 5 (short) pages long "tutorial" included, using your favourite browser. Then, start Crunchy Frog, and go through the tutorial again. The whole thing should be doable in less than 10 minutes. André

5 6

MakeBot release open-source
by Winston Wolff May 23, 2006

May 23, 2006

I have released MakeBot's source code with the MIT open source license. Vern Cedar has expressed interest in getting a Linux version running. You can now find the source at the project's Trac site: http://stratotools.python-hosting.com/ -Winston ______________________________________________________ winston wolff - (646) 827-2242 - http://www.stratolab.com learning by creating - video game courses for kids in new york

1 0

testing interactive loops
by Andrew Harrington May 23, 2006

May 23, 2006

We have been discussing testing in tutorials with doctest. There is one place that does not work, I think: testing an interactive loop, which is a construction I teach in introductory programming. It appears that doctest cannot handle interactive input. I would still like to be able to illustrate a session with interactive input and explicit output at the end, and check it automatically. If you just do input redirection and compare the actual output to the desired input, you get messed up because the input is not echoed when you redirect from a file. The following does not work, but is the idea you want, to force the echo when stdin become a file: orig_raw_input = raw_input def raw_input_echo(prompt=""): s = orig_raw_input(prompt) print s return s raw_input = raw_input_echo def input_echo(prompt=""): return eval(raw_input(prompt)) This fails because it makes raw_input a global for the current module, but not a built_in, so it fails when you import and use a module using raw_input. I guess we could make it work in little code fragment tests, where you append the user's fragment to the code above, in the same module, but it is not a general solution. Is there any better way to do this? A kludge for tutorial doc tests would be to mark the prompt strings somehow in the tutorial source, for instance making it always end with :: or ?? so the test string can be parsed to find the responses that come after the prompts, and put them in the input file and not in the output file. For example, the test string ... Sum numbers! ... Enter a number (or 0 to quit):: 3 ... Enter a number (or 0 to quit):: 4 ... Enter a number (or 0 to quit):: 8 ... Enter a number (or 0 to quit):: 0 ... The sum is 15! might be processed for web display in the tutorial undoubling the :: ... Sum numbers! ... Enter a number (or 0 to quit): 3 ... Enter a number (or 0 to quit): 4 ... Enter a number (or 0 to quit): 8 ... Enter a number (or 0 to quit): 0 ... The sum is 15! and the tutorial parser would also generate an input file 3 4 8 0 and output file Sum numbers! Enter a number (or 0 to quit): Enter a number (or 0 to quit): Enter a number (or 0 to quit): Enter a number (or 0 to quit): The sum is 15! Then run the code with redirection and compare the desired and actual output. There still is the issue of the program hanging with the given input ... How does regular doctest deal with infinite loops or waiting for nonexistent keyboard input? Better ideas? -- Andrew N. Harrington Computer Science Department Undergraduate Program Director Loyola University Chicago http://www.cs.luc.edu/~anh 512B Lewis Towers (office) Office Phone: 312-915-7982 Snail mail to Lewis Towers 416 Dept. Fax: 312-915-7998 820 North Michigan Avenue aharrin(a)luc.edu Chicago, Illinois 60611

3 4

More re Saturday Academy course
by kirby urner May 21, 2006

May 21, 2006

A useful aspect of mixing Python with some POV-Ray training, aside from the immediate access to graphics, is a golden opportunity to go over string substitution, especially using dictionary key arguments. Here is a tiny namespace that outputs colored balls, randomly positioned and sized. In class, I asked for a hundred of 'em. Note that lottaspheres simply dumps to stdout (IDLE in this case). Writing to a file might be the more sophisticated way to go, but I'm focusing on one feature at a time, and cut and paste works just fine for demo purposes: from random import randint def somesphere(): xyz = tuple([ randint(-5,5) for r in range(3) ]) coords = "<%s, %s, %s>" % xyz radius = randint(1,4)/2.0 color = ['Orange','Blue','Green','Yellow','Gold'][randint(0,4)] specs = {'theplace':coords, 'thecolor':color, 'theradius':radius } mysphere = """ sphere { %(theplace)s, %(theradius)s texture { pigment { color %(thecolor)s } } } """ % specs return mysphere def lottaspheres(n): for i in range(n): print somesphere() Actually, the output of the above left something to be desired, aesthetically. With tweaking, more could be done. Like, I'd suggest smaller spheres. Anyway, the point is it's easy to toss out any number of such tiny functions, all aimed at saving typing in POV-Ray and giving us pretty stuff to look at. And one point of doing *that* is students see string substitution in action. On a related note, I see POV-Ray as an excellent vehicle for introducing and/or revisiting XYZ coordinates as a topic. The Z axis is into the screen, with positive X to the right, Y up. Simply rendering three colorized axes, say from -5 to 5, and then test-placing various objects to get a feel for things. Important in ray tracing: to locate the point the camera. Somehow, our culture has devised these third person viewpoints that disavow their first personhood, e.g. when you're looking at some polyhedron in a math textbook, you're not usually encouraged to think about the question "from where am I looking?" Computer graphics, on the other hand, is refreshing in that it acknowledges the relevance of viewpoint or vantage point (camera location), even if the scene is quite devoid of persons. In this sense, we're closer to the world of movie directors. Likewise, ray tracing encourages to think about the so-called "secondary characteristics" of otherwise Platonic objects nesting in some Cartesian-Fermatian XYZ space. We think about color, texture, reflectivity, shadows. I like the interdisciplinary feel of this. "Pure math" (which historically turns up its nose at such special case details, or even graphics of any kind) is too confining and restraining, especially when we're trying to develop sophisticated and powerful imaginations in young people. That's why we should be using more ray tracer and less chalkboard. Beyond Flatland. Kirby

1 1

Re: [Edu-sig] Pythonic Mathematics (session 5)
by kirby urner May 21, 2006

May 21, 2006

For those curious, my fifth and final Pythonic Mathematics session for Saturday Academy in Spring 2006 is written up here: http://mathforum.org/kb/message.jspa?messageID=4726175 Kirby

1 0

Just ran into Kevin Altis...
by kirby urner May 18, 2006

May 18, 2006

Kevin is the main force behind PythonCard, which adapts the wxPython API to make it even more sane. His product features an OpenGL window, but it's just a pass-through the the wx OpenGL window; I don't think PythonCard has a bunch of OpenGL-specific classes. But these could always be added to a custom version (PythonCard for Geometers say). I ran into Kevin at Zupan's, the local food market just now. I was buying provisions for a BBQ (about to happen). Kevin reminded me I haven't submitted any talk to OSCON this year, and now the deadline is passed. I'll probably be a background figure in that event this year, given we're both still in Portland. Kirby

3 11