Mailman 3 August 2006 - Edu-sig

Mindstorm NXT and Python ?
by francois schnell Sept. 13, 2006

Sept. 13, 2006

Hello, I've just read the open source move of Lego concerning it's new Mindstorm NXT: http://slashdot.org/articles/06/05/02/111245.shtml http://mindstorms.lego.com/press/2057/Open%20Source%20Announcement.aspx The firmware sources and blue tooth protocols will be available. :) This definitely interest me and I'll buy one as soon as available (in August). I didn't find anything about using Minstorms with Python yet but it looks like there'a a SDK for Microsoft Visual Studio Express with VB, C#, J#, C++: http://msdn.microsoft.com/coding4fun/lego/default.aspx Even if Iron Python is not integrated in Visual Studio (I believe), using the classes with IronPython shouldn't be a problem (I imagine). Anyway I'm not a fan of .Net/Mono and I'll probably try to explore other ways . I want to see what can be done with Python and this Mindstorms NXT in particular in an "educational/constructivist" context. If you know any projects already using Python and the actual Mindstorms I'm also interested or your thoughts on the subject. Thanks francois

1 1

Wing IDE 101 for Intro Programming Courses
by sdeibel Aug. 29, 2006

Aug. 29, 2006

Hi, Wingware has been working with the University of Toronto on an experimental new member of the Wing IDE product line that is designed for teaching introductory programming courses. Our goal is to try to support more wide-spread use of Python as a language for teaching programming. The result is called Wing IDE 101 and has been released semi-publically here: http://wingware.com/wingide-101 It runs on Windows, Mac OS X with X11 Server, and Linux. Although Wing IDE is commercial, this edition of the product is free to use without charge. It is not, however, open source. As I noted, this is experimental, and we're not yet sure where the experiment will lead us. Please feel free to try it and if you are interested in using it for course work, contact us at support(a)wingware.com. Thanks! Stephan Deibel -- Wingware Wing IDE for Python Advancing Software Development www.wingware.com

1 0

Re: [Edu-sig] Python as first language - yet another paper
by Radenski, Atanas Aug. 28, 2006

Aug. 28, 2006

> -----Original Message----- > From: kirby urner [mailto:kirby.urner@gmail.com] > Sent: Wednesday, August 16, 2006 9:09 PM > Subject: Re: [Edu-sig] Python as first language - yet another paper > > On 8/16/06, Radenski, Atanas <radenski(a)chapman.edu> wrote: > > For those of you who might be interested, I have posted a recent paper > > on the use of Python as a first language. The paper is entitled: > > > > "Python First": A Lab-Based Digital Introduction to Computer Science > > > Thank you sir. Thank you for reading the paper. I agree with most of your comments; a thoughtful message indeed. Your reasoning on terminology needs some response, though (below). > I think your arguments will be understandable to people outside of CS, > which is important, as there's a lot of concern about how we're not > getting enough competent technical people across the board, and CS is > seen as a bottleneck, i.e. its inability to recruit seems > counter-intuitive given it has, in principle, access to a lot of the > best toys. CS-related careers have a lot of glitz and glamour. But > as you say, a dry course in nothing-but-Java doesn't give that > impression, and would-be stars wander off to other majors. > > Internally to CS, however, there's the more esoteric line we draw, not > between commercial and education languages, but between system and > agile languages, the former being used to write the latter. Classic > core Python is a very commercial project because it's written in > lickity-split C code, and that's about the fastest on the planet. I use terminology with a purpose in this paper. Speaking of educational and commercial languages serves my purpose, while speaking of system and agile languages would not. Here is part of my message: "In the not so distant past, introductory computer science education was dominated by languages that were specifically designed for education, such as Pascal, Basic, and Logo." I mean Wirth's Pascal (not Delphi), Kurtz's and Kemeny's Basic - the one also beloved by Bill gates (not VB.net), and the Logo of the Feurzeig and Papert flavor. Such educational languages are intended for small scale programming. They are simple and manageable by beginners. Also: "Today, the majority of introductory computer science courses are based on a commercial language, such as Java or C++." Commercial languages, such as Java and C++, are intended for large scale software development. Such languages are naturally complex and difficult to learn by beginners (who originally do only small scale programming). The full discussion is in the paper (http://www1.chapman.edu/~radenski/research/abstracts/python-iticse06.ht ml). My point is that this particular conversation cannot be carried out in terms of 'system' and 'agile' language. For one thing, Pascal, Java, and C++ are all considered system languages [http://en.wikipedia.org/wiki/Systems_programming_language]. Besides, it is difficult to decide if Logo and Basic (the original ones) are system or agile languages. May be they are neither system nor agile languages, yet these languages are quite clear educational ones. So I think that speaking of educational and commercial is the right thing to do in the context of the paper. > As for whether Java is the right follow-on for a CS2, I think a bigger > success in recruiting will open more options there too. > > Yes, the Java track is secure for the time being. But don't deny > those who wish to dive into C, now that they're hooked, now that they > know that coding is for them. Explore the guts of CPython, the study > the Java and C# implementations. ANSI C is a wonderful language. The only hurdle to a student is to understand that arrays and pointers are the nearly the same thing. Once this is realized, C should look like a really simple language to the student. C++ is a different story. It is the product of a pretty long evolution. The whole thing lacks uniform design and is messy. For example, is it really necessary to have C++ objects possibly allocated on the stack, not only in the heap? And look at exceptions - they seem pretty ugly linguistically. Java is now going the same way - evolution clutters and complicates the language. Until Java 5, Sun managed to hide the evolution inside the API only, though. Python, in its entirety is not simple, either. Hey, we got two styles of classes, for example. But Python has this beautiful kernel that one can use to teach beginners effectively, by arranging concepts in a reasonable hierarchy, without many cyclic dependencies. It is not like teaching Java, when you have to say all the time: "This concept is important, but we will study it later." The design of Python was initially motivated with educational goals, and it shows. Yet, the language caught up with industry, which is very fortunate. A wonderful hybrid language: good for beginners and good for advanced work. Functional, imperative, OOP - you name it :-) > Kirby Atanas Atanas Radenski mailto:radenski@chapman.edu http://www.chapman.edu/~radenski/ There are wavelengths that people cannot see, there are sounds that people cannot hear, and may be computers have thought that people cannot think -- Richard Hamming

1 0

slashdot: Teaching Primary School Students Programming?
by Paul D. Fernhout Aug. 28, 2006

Aug. 28, 2006

http://ask.slashdot.org/article.pl?sid=06/08/27/000248 Teaching Primary School Students Programming? Posted by Cliff on Saturday August 26, @08:25PM from the toddlers-as-coders dept. Education Programming NotesSensei asks: "Recently I was teasing the teacher who runs the computer club in my sons' primary school: 'You teach the kids only how to use software but not how to make software.' Today I got an email: 'OK, you're in: teach them programming.' Now I wonder what language should I pick? My first lesson will be the board game c-jump, but after that? The contestants are: Kids programming language KPL (ab VB.net derivate; Java using BlueJ; Greenfoot (and the BlueJ); and HTML. Does it sound like I'm on the right track or should I try something completely different? We are looking at primary 3-5 (that's 10-13 in this part of the world). Where can I find inspiration for the curriculum?" === Python is mentioned there, along with a lot of other suggestions.

6 9

ANN: PyLogo 0.3
by Ian Bicking Aug. 28, 2006

Aug. 28, 2006

With the recent mention of Logo, I realized I forgot to announce PyLogo 0.3 here: http://cheeseshop.python.org/pypi/PyLogo/ This release features the addition of object-oriented language additions (TELL, ASK, class definitions), a doctest mode (see docs/test* for examples of Python doctests), and slightly cleaned up interpreter interaction. Ideally going forward with PyLogo I'd like to keep it focused on the language implementation and look elsewhere for IDE/GUI support (removing what little is currently there). I believe (but this is not confirmed by experience) that it should be relatively easy to integrate into other Python-oriented environments -- it can call Python functions and deal with Python objects. Since there's already a lot of interesting projects from several directions this decoupling of language and environment seems more fruitful for experimentation. -- Ian Bicking | ianb(a)colorstudy.com | http://blog.ianbicking.org

1 0

An Urner update on doings/undoings
by kirby urner Aug. 27, 2006

Aug. 27, 2006

I did a quick coin of HP4E but haven't used it for much since: http://mathforum.org/kb/thread.jspa?threadID=1437142&tstart=0 Basically, I'm refining the "gnu math" concept, addressing concerns about the New Math resonance, which one teacher worried had "Titanic" connotations (as in "went down").[1] But I'm pointing to what we've saved from that shipwreck: an emphasis on bases (as in subtraction in base eight), and on Venn Diagrams (boolean algebra), and on Sets (unions and intersections). All that stuff is still in the mainstream, from Saxon to Singapore (those're textbook allusions, much talked about on the math-teach list I frequent). Gnu math uses a non-proprietary substrate on top of proprietary hardware (e.g. Intel's, AMD's...) to keep computer programming a "for everybody" sport, i.e. we won't wall them out on the basis of price alone (never a good way to promote a meritocracy). Yes, gnu software also runs on proprietary operating systems, and yes, we might someday have an open source chip, i.e. the designs are free and clear to anyone with a fab (still a huge barrier to entry at this point -- clean rooms ain't cheap). But until that blessed day, when it's all open source as far as the eye can see (the Debian ideal), I'm not averse to using whatever public venues they leave open to us laypeople (even as we keep fighting to make inroads). In the gnu math I envision, sets are just another data structure (added somewhat late to Python, as dictionaries with null values, only keys, did pretty much the same work). We're not as beholden to early 20th century philosophies of mathematics, which made the set front and center in some theory of types, with Russell's paradox about what might be "empty" or not. Per Wittgenstein, all that Cathedral talk does *not* underwrite what we do in the Bazaar (set theoretic logic provides a painted foundation under a painted castle, to paraphrase the guy). The genesis of mathematics is ordinary ethno-arithematic, and as such is "all in the open" i.e. doesn't depend on special Harry Potter powers to see in "higher dimensions" (another position of mine: contra the HyperCross Dogmatists (but I'm fine with Coxeter's polytopes and recognize their utility in higher mathematics (but a hypercube is *not* the same thing as a time machine remember, per 'Regular Polytopes' on page 119))).[2] So yeah, we've got sets (lists, dictionaries...), and math.pi, and pi generators, and let's not forget phi. Phi is actually at least as important as pi as a curriculum asset, once we're into the greek letters and spooky greek metaphysics ("infinitely extended infinitely thin planes" and all that ghostly stuff). People tend to dismiss phi, out of bias against anything to do with the pentacle. Holdover superstitions. But we can't deny the pentagon its due, phi being the reciprocal of its edge-to-diagonal ratio, although some call that tau. Five-fold symmetry *is* important in Nature. Who are we, the naked apes, to argue with her wisdom? In fact, we've done a lot of homework on pentagon math right here on edu-sig, including in the form of continued fractions (a perennial favorite, as phi is "just ones" all the way to the bottom, like turtles). Plus you've got this convergence to phi along the Fibonacci sequence, already paradigm generator in the Pythonic literature, right up there with Guido's cute little gcd function. I'm still liking the fact of fib.next()/fib.next() getting closer and closer to the most irrational irrational, every time we hit the Enter key (again, the bigger number will be in the denominator, given left-to-right evaluation, so the convergence is to 0.618..., not to 1.618...). >>> from __future__ import division >>> def fibs(a=0,b=1): while True: yield a a,b = b,a+b >>> genfib = fibs() >>> genfib.next()/genfib.next() 0.0 >>> genfib.next()/genfib.next() 0.5 >>> genfib.next()/genfib.next() 0.59999999999999998 >>> genfib.next()/genfib.next() 0.61538461538461542 >>> genfib.next()/genfib.next() 0.61764705882352944 >>> genfib.next()/genfib.next() 0.6179775280898876 >>> genfib.next()/genfib.next() 0.61802575107296143 >>> genfib.next()/genfib.next() 0.61803278688524588 >>> genfib.next()/genfib.next() 0.6180338134001252 >>> genfib.next()/genfib.next() 0.61803396316670656 >>> genfib.next()/genfib.next() 0.61803398501735796 >>> genfib.next()/genfib.next() 0.61803398820532507 >>> genfib.next()/genfib.next() 0.61803398867044324 A general thrust of gnu math is to use OO concepts to build awareness of "maths as extended type systems" (class hierarchies if you will). That's a bridge to our Fuller School geometry, as we regard the Tetrahedron as a kind of superclass object or parent, for the whole idea of Polyhedron in general. You have edges, vertices, faces (V + F = E + 2), and an inside and outside. Every subsequent Polyhedron inherits those elements, and the machinery to go with 'em. And we don't have anything simpler. The sphere is actually much more complicated to think about, given its reliance on a continuum hypothesis. The tetrahedron is friendlier to discrete math engines like Python. We don't need some infinitely perfect pi to think about our what's in our tree.[3] If you've followed my rbf.py thread's evolution, you know that it builds on a primitive vector class (of the kind any first year student could write), and face-tuples, to define a vocabulary of primitive prefrequency polyhedra. If I go: >>> from rbf import Tetrahedron, Octahedron, Cube >>> mycube = Cube() >>> mytetra = Tetrahedron() >>> myocta = Octahedron() and then ask for volumes I get: >>> mycube.volume 3.0 >>> mytetra.volume 1.0 >>> myocta.volume 4.0 Whereas if you *haven't* been following my rbf.py's thread's evolution, you'll probably just be asking yourself "Why?" at this point. Because that's how we've designed our little zen rock garden in Synergetics.[4] The tetrahedron is unit volume, given it anchors the Polys class hierarchy (introduces the whole idea of an inside versus outside for example). Two tetrahedra, per Kepler's stella octangula, define a Cube of thrice the volume. And that Cube's dual, with edge intersections at 90 degrees, is the Octahedron of volume 4, and anchors our 6-rayed XYZ coordinate system (which, notice, comes somewhere on down the hierarchy tree i.e. the primitive tetra is topological such that no coordinate system machinery need apply). OK, that's a lot of verbiage. You're likely happy I don't flood edu-sig with all these threads. When I'm blabbing about Bucky a lot, I tend to firehose in the direction of some Bucky-friendly e-list, like Synergeo on Yahoo. Or maybe I'm on wittgenstein-dialognet (pretty quiet lately), or just discussing stuff privately among Wanderers (wwwanderers.org). Finally, in addition to Glenn's global matrix, I've been caught up in Sam's and LaJean's scenario. They've been trying to cobble together a website, using a content management system. I've brought this up on edu-sig because I thought it relevant to our community, given Python has one or more of the flagship CMSs out there. We shouldn't over sell them (CMSs) to people who just need static web content and could get by with XHTML/CSS. Client side WYSIWYG tools make it easy to avoid server-side dependence. Unless you're hosting the server yourself, beware of getting captured by your own ISP. However, in the case of flextegrity.com, I think it's turning out OK. They're likely to need those fancy CMS features down the road (implemented in PHP in their case), i.e. it was never all that "static" a design (even though it may have started out that way). Kirby 4D Solutions Portland, Oregon Notes: [1] re 'Titanic' http://mathforum.org/kb/message.jspa?messageID=5055407&tstart=0 [2] Coxeter & 4d: http://controlroom.blogspot.com/2006/08/more-dimension-talk.html [3] More About Pi: http://groups.yahoo.com/group/synergeo/message/28550 [4] rock garden: http://www.grunch.net/synergetics/volumes.html

1 0

Re: [Edu-sig] where can I find xturtle?
by Gregor Lingl Aug. 26, 2006

Aug. 26, 2006

There was a server-crash at rg16 while I was on vacation. That has been fixed now. You can find xturtle again at http://ada.rg16.asn-wien.ac.at/~python/xturtle Sorry for the inconvenience Gregor

1 0

Appreciating Crunchy Frog
by Arthur Aug. 24, 2006

Aug. 24, 2006

Crunchy Frog is cool! Must be, since even grouchy Arthur thinks so. To me the attraction is not so much "web-based" - at least any more than the attraction of something like html docs is that they are necessarily web-based - in the sense of Web as in Net, as in Remote. 127.0.0.1 is cool too. Don't think that the computer obsoletes text as the primary vehicle for learning. And it seems pretty clear that reading straight text documents via computer has *disadvantages* over the dead tree approach. And in fact most web based interactive learning efforts I have seen seem lame to me. See Not Pi. http://www.sock-monkey.com/pi.html I do not see too much more than that in javascript turtles that folks seem to find so exciting. But I, and most everybody else, appreciate the fact that hyperlink technology for documentation - as an example -can add a level of truly useful functionality to text documents. Seeing Crunchy embedded in, and extending, a substantive text document as in the Crunchy .7 mini-howto helps me understand and appreciate its usefulness in extending text documents. Here, existing "how to" docs were extended with Crunchy. If they were written from scratch with embedded Crunchy they might have taken a slightly different approach. Might be a while before the possibilities here get appropriately shaped. Why is all the cool stuff being done by us "hobbyists"? Art.

2 1

Re: [Edu-sig] Another update from the field...
by kirby urner Aug. 21, 2006

Aug. 21, 2006

On 8/20/06, Dethe Elza <dethe.elza(a)gmail.com> wrote: > Eliot Kimber (aka Dr. Macro) has a great series on his blog about > building an XML-based CMS. It is really applicable beyond XML (and > most XML stuff is applicable to HTML). He advises (and is building > open-source and blogged all the way) a solution on top of Subversion > (versioning software). Does Eliot consider Zope 3 a good example of XML-based CMS? I've seen it presented as Model-View-Controller, with ZODB providing an API to the model, XML controlling the View, and Zope + custom Python being the controller. > If HTML is the scary bit, I suggest looking at reStructured Text/ > Markdown/Textile as alternatives (depending on need). All convert to > nice (X)HTML. reST is the most complicated, but can also be > converted to XML, PDF, PPT, and other formats, and you can add your > own hooks to bring in more advanced features. "Convert to nice (X)HTML" where though? You're still implying some engine doing the translation, whereas straight XHTML/CSS liberates you from anybody's Wiki or whatever framework. Per my friend Gene Fowler's vision, I think we should just plan on spreading XML fluency at least to a point where fear of XHTML is the exception, not the rule. The basics aren't that hard -- arguably as simple as all the rules reST gives you. Gene's vision: http://controlroom.blogspot.com/2006/08/more-cast.html > If getting the content up on the web is the problem, a wiki or blog > tool might be the answer, or simply showing them how to share a > remote directory to the desktop for editing. Too non-standard for a basic company website. This is a small startup that can't afford to look too quirky, given the invention itself is already on the wild side. For a sneak peak: http://www.4dsolutions.net/flextegrity/ (I'll be taking it down within 12 hours, so expect a broken link pretty soon -- sharing with you because you track the Bucky stuff some). The real site will debut on flextegrity.com in the near future, probably as a PHP site (i.e. I expect my good advice to be rejected). > A full-fledged CMS is often the wrong solution even for the high-end > corporations that they're generally targetted for, and always require > more effort than is anticipated. > > --Dethe > Yes. A CMS is a wonderful thing when it's the right tool for the job, but in too many cases it's promulgated as a way to make the web "friendly" to newbies, whereas just learning a little HTML in a WYSIWIG editor would be a kinder gentler thing to share with them (keeps them free of the CMS itself, a gravitational field). Per Gene's vision, I want to see more mixing of XML with everyday language arts. Learn some tags at the same time you're learning basic punctuation, and the difference between a verb and a noun. No one finishing the Shuttleworth sequence, for example, should feel "afraid of HTML". Kirby

3 7

Another update from the field...
by kirby urner Aug. 20, 2006

Aug. 20, 2006

Per blog, I've been huddling with M language, basis of MUMPS, given its huge installations and emerging dino language status. CS has a new challenge: whereas round one was about a scatter gun of languages "making it" into commercial use, round two is about recruiting critical numbers to a veritable dinotopia of exotic solutions that nevertheless work, and starting over would be more expensive than simply holding on to a good thing. What I think this means in practice is marketing CS to humanities majors as a way to gain an esoteric proficiency in some vintage museum quality hieroglyphics, sufficient to pay quite a few bills, if not all of them. That might be an avocation more than your core identity (you're a scuba diver for a living, not a J programmer -- but yes, you write J and it at least pays for your air). I'm afraid I'm pretty harsh in my criticisms of current CS (this is back office stuff, so I feel free to take the gloves off), for its dry-as-bones mono-lingual culture. We should be advertising the diversity of lifestyles available through this intro-level portal (CS0). As recruiters, we're hear to expand their horizons, not force 'em into cubicles (or visions thereof) before their time. On another front (closer to Plone), I'm busy trying to talk a client *out* of using a content management system (CMS) where plain old static web pages would do. The problem with pouring your content into a ZODB or some such, is your stuff has no legs, no autonomy, without this complicated wrapper. For some, that's the whole point. For others, they see a CMS as a crutch, something to save them from learning any XHTML/CSS. What they don't realize: (a) the learning curve is just as steep, at least up to the basic level in both areas (operating a CMS API as an end user is really no easier than mastering beginner-level DreamWeaver) and (b) you're hitching your star to the arc of a CMS, whereas straight HTML is pretty much guaranteed to run anywhere. In my client's case, the CMS is an open source PHP project, and looks to be capable and high caliber. As soon as you want to add eCommerce features, polling, forums, other bells and whistles, a CMS starts to look pretty good (I love Plone by the way, even if I'm not a master archetypes programmer). But if you're just wanting a "cyber business card" web site, as this client does, static HTML is likely a better solution. I recommend this party line to Zope sales by the way. Don't be afraid to turn people away on the basis of "you just need to learn some webmaster basics or hire a smart HTML jockey for your needs" i.e. don't market a CMS as "the right solution for everybody" any more than "computer programming for everybody" means we all program the same way, or all have to use Python. Kirby

1 0