When I was in high school, you’d have to be a megalomaniac or the most popular kid around to think of yourself as having a fan base.
Yeah, well, either that, or you were in a band.
When I was in high school, you’d have to be a megalomaniac or the most popular kid around to think of yourself as having a fan base.
Yeah, well, either that, or you were in a band.
I'm on my way to PyCon in Dallas. For you non-geeks, PyCon is the annual conference for the Python programming language. I'm not presenting anything, but we are having a SchoolTool/CanDo development sprint during slow times in the three day conference proper and four days afterward. We're bringing over the main SchoolTool developers, Ignas from Vilnius and Jean-Francois Roche from Brussels, and from CanDo we'll have Jeff Elkner and several of his students (and his brother Alan) along with lead developer Paul Carduner. We are focusing on the resource scheduling/reservation functionality, which has been about 80% done for a while, but needs some polish to really close the "sale" with schools.
The registration for this PyCon has been strong, and it feels like a buzz is building up again around Python. I previously attended PyCon two years ago, which was when I met Jeff. That was the last PyCon in the DC area, and the unspoken theme was "What is this Ruby on Rails thing, where did it come from, why is it kicking our ass, and what are we going to do about it?" There was the distinct feeling that Python had fallen behind on web frameworks just when what would come to be known as "Web 2.0" was taking off, and unlike the Ruby community, it was going to be difficult, if not impossible, to get everyone to push behind one framework.
Last year, the response had begun to take shape, but the move to Dallas discouraged me, and apparently quite a few others from attending.
The past month or so I've been working with various new Python web components, like SQLObject, which is used in TurboGears and Pylons, but also makes a great standalone object-relational mapper for whatever database you're trying to manage, and Breve, which is a lightweight s-expressions style XML templating engine that can be plugged into several of the Python web frameworks or again, easily be used standalone. I definitely feel that Python's web options stack up to the Rails juggernaut now, but there is still a lot of diversity of options and approaches in the Python toolkit (it is one case where "there is more than one way to do it" in Python), but now we've got good freedom of choice diversity rather than bad every man for himself diversity. So I'm excited about learning more about the exciting new bits and meeting some of the developers.
Also, the Python in education community, which has had mixed success at essentially grassroots advocacy over the years, has received an incalculable boost by the very prominent use of Python in the OLPC system. PyCon will be my first chance to lay my eyes (and hopefully my hands) on an XO prototype and learn more about the current state of the software. I can't wait for that.
I've neglected to mention that I participated in a podcasted discussion on Teachers Teaching Teachers of the various implications of Google Apps for Education and other Google-stuff for schools. I'd particularly like to thank Paul Allison for inviting me to participate, and particularly for referring to me as "a friend of Writing Projects." One thing I forgot to mention: I wonder if the metaphor for AdSense ads in Google Apps isn't that they're like, say, a billboard on the front of the school (pretty clearly not ok), but that they're more like the advertisements in magazines that the library subscribes to (pretty much universally accepted). Somewhere inbetween the two, perhaps, but those are good reference points.
Also, Pat Delaney's follow-up post on the use of Google Apps at his school is very illuminating, and his overall conclusions are spot on.
I always find commentary about military reform to be more illuminating for the problems of school reform than comparisons with business. Like this piece by William S. Lind on the Swedish conception of "military intelligence" as underrättelser, or "correction from below:"
Higher level commanders are even more victims of the current system than are their juniors. With sufficient guts, junior leaders can ignore the intel. Unless a senior commander is the sort who recognizes that his headquarters is a Black Hole and stays away from it as much as possible, he has no alternative to the virtual reality his G-2 presents to him. He is not only flying blind, he is flying blind while thinking he sees. Out of such double-blindness many great defeats have come.
What is missing here is precisely underrättelser, correction from below. Instead of dumping the errors on the users, the whole intel system should avidly seek correction from below to minimize them. Errors cannot be eliminated, because no matter how good the intel, it will be incomplete and some will be wrong. But correction from below, from the people who are directly encountering the enemy, is the only way to reduce them. By making "correction from below" literally their name for military intelligence, the Swedes have made the intel system's most necessary characteristic definitional. Intellectually, that is a remarkable achievement.
Defining military intelligence as "correction from below" also carries the culture of a Third Generation military over into the intelligence process. Just as another of those words that speak volumes, Auftragstaktik, builds tactics on the understanding that the levels of command nearest to the fight have the clearest tactical picture, so underrättelser builds military intelligence on the same understanding. The two work hand-in-glove: junior leaders act on the basis of what they see, not detailed orders from remote headquarters, and they simultaneously feed what they see into an intelligence process that is eager for their corrections. Neither action eliminates uncertainty in war, because nothing can, but both speed adaptation to it, which is the goal in maneuver warfare.
I think the relevance to high-stakes testing and NCLB is obvious.
...on May 2 of 2004, Principal Fain reported to the Norwich Bulletin that at the Kelly Middle School [800 students] there were 27 police visits and 14 arrests for violent activity.
One of those incidences involved Julie Amero. On April 14, 2004, Julie was a substitute teacher when a fight broke out in the school. The regular staff all closed their doors and pretended nothing was going on. A janitor warned Julie not to get involved. Julie ignored him and arrived too late. A thirteen year old girl was badly beaten up by another girl at the school. Julie arrived in time to comfort the bleeding child. Julie was the last person the girl would see before she slipped into a coma.
A few months later, Julie is again substituting and is accused of cruising the internet all day instead of engaging the students.
Since 9/11 it has seemed pretty clear that Pakistan is the most dangerous country in the "war on terror," in fact, it is such a dangerous country, we're afraid to even say how dangerous it is lest we catastrophically destabilize it with our own rhetoric, and besides, nobody has any bright ideas for fixing it. So we attacked Iraq instead.
How many people actually care about this? I don't know, but I'm happy about it. Like everything OLPC does, this is pretty ballsy... if they can put a UI on a distributed revision control system like Bazaar that will actually make sense to kids, it will be a powerful combination, not only for programming but all kinds of collaborative text editing.
I made the Soupe à l’Oignon Gratinée from last weekend's New York Times Magazine for our Valentine's dinner. It is easy to make, if a little time consuming; perhaps more importantly, it is easy to remember and find all the ingredients (1 baguette, butter, Emmental cheese, 8 onions, tomato puree) while walking around the supermarket.
Goes well with rack of lamb.
I guess if you don't know much about the history and philosophy of free software, or, more to the point, you don't care about the movement, Doug Johnson's piece in Education World, "Free is Good: A Beginner’s Dozen of No-Cost Computing Tools" seems innocuous enough. But it is another small passive-aggressive thumb in the eye of those of us who advocate for free software in education.
The biggest hurdle we face long term is simply educating people about what free and open source software is and how it is different than proprietary software. Articles like Doug's make this job harder. It is exhausting trying to undo the damage caused by this kind of passage:
There are basically three types of “free" software:
- Open source software uses code that the creator has placed in the public domain and that a large body of users then re-writes and adds to. The Linux operating system is probably the most famous open source product available.
- Minimally-featured versions of commercial products are made available by a producer who then hopes that features available only in the purchased version will sell the software. The popular e-mail reader Eudora operates that way.
- Web-based software applications that derive revenue from advertising are growing in popularity. Yahoo mail uses that economic model.
Open source software is one type of free software? Open source software is in the public domain? Commercial software can be is free software? Whether or not these assertions make your hair stand on end largely depends on whether or not you think the free software community, Free Software Foundation, GNU Project and/or Richard Stallman himself are worthy of any respect or consideration whatsoever.
The evidence indicates that the educational technology establishment, in which Doug is a significant voice, does not think so; the exclusion of the history, principles and ethos of the free software movement from the mainstream discourse about educational technology is complete, unwavering and systematic.
This is particularly ironic coming from someone like Doug who has written a book on Learning Right From Wrong in the Digital Age and is interested in what the librarians like to call the "the ethical use of information." The free software movement is based on a fundamentally ethical critique of proprietary software distribution. If anyone should be interested in substantively engaging with this movement, it should be someone like Doug.
I cannot explain this phenomenon, I can only point it out.
My copy of the new edition of Philipp von Weitershausen's Zope 3 book arrived yesterday from Amazon US. It's got many updates, technically and organizationally, and if you're interested in SchoolTool development, a must-have.
A Richer Picture digital portfolio system will allow an administrator to create a full export of the system's MySQL database. In general terms, this is an excellent practice; the school can always dump their data and create their own complete version of the database locally, in MySQL or any other relational database. In practice, this process can be a little bumpy.
What you get is a zip file with a bunch of comma-delimited text files, one for each table, and an SQL file. I was hoping this file of SQL commands would automatically recreate the database. Unfortunately, it doesn't. I'm not entirely sure what it does, but generally it seems designed to export the data rather than import it. Regardless, it does serve as a good ersatz schema, and turning it into a import script mostly just requires changing the
WRITE statements to
LOAD statements. Even then, I ran into a string of problems trying to deal with the commas and line endings contained within the text of individual database entries. I think they need to be more explicit about escaping the line endings in the export, although I've found the relevant MySQL docs somewhat vague. I've currently wrangled it into a position where I think I'm importing all the records, but the comments containing line endings are getting chopped. I've gone back and forth with David about this on email a bit, and hopefully we'll get this completely resolved, and I'll post the full script.
If I was more experienced with relational databases, this would be old hat, I suppose. These problems come from underspecified nature of the CSV data format. Unlike, say, XML, there is no one right way to handle character escaping and the like (not that everyone gets it right in XML, but at least you can definitively say they're wrong, for what that's worth). You can export XML from MySQL, and I'd prefer it, even in a really braindead dialect. In theory, the right way to do this would be to use some of the IMS XML specs, but given that I have no tools that make processing data in that XML format easier than in any other, it doesn't give an significant practical advantage at this point (and SIF is a whole message passing system, which would be great to have in this case, but getting it running on either end is a big job).
Anyhow, I've got most of the raw data in my own MySQL database. Hopefully the next steps will be more interesting and enlightening.
When I left my position Feinstein High School technology coordinator (and that position ceased to exist), the school was ready to try implementing digital portfolios to augment and eventually replace the paper portfolios that are central to the school's work. In Providence, digital portfolios are something that teachers have been hearing about for about a decade, in part because of the work of David Niguidula as part of his work locally over the years at Brown, the Coalition of Essential Schools and his current company, Richer Picture. So when I left, I encouraged them to hire David, which they did, and apparently he now has a contract with the whole district, although I don't really keep up with the politics so that might be wrong.
Anyhow, in the subsequent years, FHS has incrementally incorporated the Richer Picture digital portfolio system. This is a slow process, as digital portfolios require every element of your technical infrastructure to work well before the portfolios will really thrive. In the meantime, I've continued to take responsibility for the school's standards-based evaluations, which have been put together over the years using either a hacked alpha version of SchoolTool or the lowest common denominator of IT, spreadsheets.
The problem with these expediencies is that the schools data remains not so much locked in silos as sitting in randomly scattered piles. This causes me and FHS's principal, KC Perry great consternation at the end of every semester; it is not what we envisioned when we started the school five years ago. However, it is also a difficult trap for me to escape, because I inevitably am sucked off into long internal debates about the "right" way to solve the problem, and why SchoolTool isn't quite ready for the task, but will be real soon now.
However, this is the first year that KC has pushed teachers to start entering their assessments directly into the the portfolio system, and at the end of the fall semester, I needed to be able to extract those assessments from the portfolio database. The Richer Picture digital portfolio now runs on PHP & MySQL, and the web application supports downloading a complete export of the database, so I decided that the most practical way to actually make some progress in integrating and analyzing old and new data going forward was to develop some tools that would work with the portfolio database.
Thus ends part 1...
So after cutting out just about every ed-tech blog I read, I've still got 149 subscriptions, and I'm sharing about the top 5% of those posts, in terms of objectively measured quality (I've got an algorithm, but it is proprietary, sorry), so if you're looking for 10 or so good links a day from outside the echo chamber, you can subscribe to my shared feed.
In somewhat related news, it looks like a way to filter out posts that include the phrase "School 2.0" appeared the day after I asked for it. Pipes should be able to do it, but there is one problem, I need Pipes to be integrated into my RSS reader, so that, for example, all of my incoming feeds could be run through the same filter before I read them. Also, I want to be able to flag things as I'm reading them and have Pipes do different things after that point. So either Yahoo! needs an feed reader (or don't I know about the one they've got?) or Google Reader needs to add similar functionality (and I'm quite confident they will).
I'm generally NOT spending a lot of time trying out random free-as-in-beer web apps by startups I've never heard of, but I'm trying to make an exception for stikkit, a note-taking tool and the first app from Rael Dornfest's new startup. I say "trying" because I simply can't stay focused on any kind of note taking system. But Rael was responsible for many of my favorite books and hacks, including Peerkat, a defunct RSS aggregator which I still pine for, so I consider keeping up with Rael's ideas to be a very wise strategic investment.
Anyhow, stikkit seems very clever and rich, and worthy of a pilot with some curious students. It actually rolls some of the core ideas of Chandler into a web app... I'll be writing more on Chandler as soon as I finish Dreaming in Code.
I was reading "MySchool 2.0" by David Jakes this morning and something finally snapped. I can't do this any more. Writing isn't a problem, but I can't read this ed-tech crap any more. It is insipid. I'm just not getting anything out of it. I've cut everyone out of my ed-tech blogroll except principals, a couple graybeards, and people actively involved in open source.
Those of you who have managed to stop wetting your pants over a Senate bill with one sponsor in the minority party might want to start supporting positive legislation, like The Employee Free Choice Act, introduced into the House Tuesday with 232 co-sponsors. The act would:
.... restore workers’ freedom to choose a union by:
- Establishing stronger penalties for violation of employee rights when workers seek to form a union and during first-contract negotiations.
- Providing mediation and arbitration for first-contract disputes.
- Allowing employees to form unions by signing cards authorizing union representation.
Now, I know many of you have resigned your children to the tender mercies of the invisible hand of Tom Friedman, but restoring the right to organize in the private sector is really the only thing that is going to mitigate the tremendous and growing inequality of income and wealth that the global economy is creating. If you've been involved in trying to organize a union in the past few decades, or know someone who has, you've seen how far the enforcement and implementation of existing laws has tipped in favor of capital. And yes, we need to organize labor globally to bring balance to the global economy, but we can't do that if we don't have strong unions here.
This will probably never become law as long as Bush is in the White House, but it is time to start putting the pressure on, and we'll get this through in 2009.
I'm about ready to write some kind of script to filter out any post containing the phrase "School 2.0" or any variation thereof. Without going into a long screed, let me just point out that "School 2.0" is not even the most interesting possible educational variation on a recent O'Reilly Media propogated concept Isn't "Make School" much more evocative? At least if you've read Make magazine (and if you haven't, you are a l-o-s-e-r). What Make School would look like is pretty obvious: a little '60's auto body shop class plus Papert's constructionism plus The Met plus ZOOM plus The Center for Bits and Atoms plus OLPC plus etc etc... You get the idea. It would be a blast, and it would have to look like that. "School 2.0" could be anything. You can't start with a vague buzzword and build around it. It just won't work.
First, let's pause and reflect on the fact that this is distributed as a PDF, rather than, say, a Web 2.0 technology like a wiki or blog.
Second, the main feature by Karoli Hayes is just wrong. Basically, her son was paid to transcribe an old rock drumming part, which was subsequently published without giving him credit. She misses the key point, that her son was doing "work for hire," which, from the Wikipedia:
...is an exception to the general rule that the person who actually creates a work is the legally-recognized author of that work. According to copyright law in most countries, if a work is "made for hire", the employer—not the employee—is considered the legal author.
This is a doubly vague case because she doesn't specify who the original artist was (was he the same as the drummer who commissioned the transcription?) and regardless, copyright around drum parts is a particularly murky territory. Again, from the Wikipedia:
Arrangers in pop music recordings often add parts for orchestral or band instruments involving new material such that the arrangers may reasonably be considered co-composers, although for copyright and royality purposes usually are not. Rhythm section parts are usually improvised or otherwise invented by the performers themselves using chord symbols or a lead sheet as a guide.
If you scribble a doodle on a napkin, that's copyrighted. Guitar part: copyrighted. Can you publish a transcription of a guitar part without the permission of the copyright holder of the song? No. But a drum part and transcription? I (a drummer with several albums to my credit) am left wondering.
Lastly, the newsletter is published under a Creative Commons by-nc-sa license, but their download page says:
We ask that you DO NOT upload or post the newsletter on any other website. You are welcome to link to this newsletter and print it out (AS IS) for distribution on your immediate campus only.
So... what's the point of using a Creative Commons license at all?
Things got a little hectic around here this week, and I didn't get to finish the series of posts I had lined up to disassemble Dan Meyer's Defense of NCLB. So I'll briefly summarize one of my remaining points.
Dan's piece is framed as some PR advice for his fellow teachers:
...I’m embarrassed of you. The public thinks we’re a bunch of whiney, overentitled babysitters and ever since NCLB debuted back in 2002, you’ve done precious little to improve our public relations.
The problem with Dan's argument here is that public opinion has remained evenly divided on NCLB, but when people are polled on individual strategies of NCLB, many of them are overwhelmingly unpopular, including some of the ones Dan is encouraging teachers to drop. Look at this Phi Delta Kappan/Gallup poll. For example, 82% are concerned that NCLB will cause less emphasis on art, music, history, and other subjects. You can look at the data yourself for more, but I think it is pretty clear that making these arguments is not bad PR for teachers at all, but in fact the opposite.
Pretty much everyone who was awake this morning linked to the "Web 2.0 ... The Machine is Us/ing Us" video, which is nicely done, but what's its argument? The first half is all about the importance of separating form and content on the web by moving from HTML to XML. From there, Prof. Wesch asserts that "with form separated from content, users did not need to know complicated code to upload content to the web." Um... what? Because those XML editors are so much easier to use than the WYSIWIG HTML editors? XML is a good thing, but practically speaking, our lives have been made easier by database backed web applications that can spit out HTML, XML, JSON or whatever you want. Occasionally this is done primarily with specifically XML technologies (XSLT, etc.), but primarily, this tool chain hasn't risen to dominance.
Beyond that point (which I think is important, as it is the pivot between the first and second half of the piece, coming exactly in the middle), the obvious importance of the form of this video itself refutes the importance of a separation of form and content and that separation's role in "Web 2.0." This is a video, it isn't XML, and you can't separate its form from its content, and it should be quite clear that its popularity is due to its form rather than its content. If you re-worked this as a textual post, its incoherence would be obvious, and it would be ignored.