Google for Classroom Collaboration

The OLPC project has proven value to a one-to-one computing approach facilitating collaborative learning. 

Similar projects, from first-world XO deployments, to the open 1-to-1 project in Maine, are following suite. 

The XO's mesh networking and collaboration functionality is buggy at times, but quite useful in a rural setting where infrastructure is unavailable. 

In my life, I have found collaboration to be best done via Google's web apps, with Gdocs and Gwave a staple of my business and scholastic life, providing real-time, multi-media collaboration anywhere with internet. Any document created with Google Docs are automagically stored in the cloud, where they are available for download or remote editing. Taking advantage of these free services makes tons of sense in an educational computing setup.

In stick-based 1-to-1 education, where the kids carry all of their files and programs around on a bootable flash drive, GDocs would provide a full backup in case of loss of stick. Storing editable files in the cloud would also let kids work on school projects, or parents/teachers monitor school projects, from home computers, without needing to boot into the stick OS.

In first-world educational laptop deployments, Google apps act as a "path of least resistance" for collaboration. Rather than having the IT staff install a custom Jabber server, simply configuring Google apps for school use accomplishes very nearly the same goal.

In very small schools or very large schools, Google apps acts as a "path of least resistance" for collaboration. Small schools often lack the resources necessary for internal collaboration servers. Google Apps for Education is free, and easy to set up. For larger institutions, it's a significant strain on internal resources to provide real-time collaboration via traditional methods. Google scales up excellently. 

There are obvious failings with this proposal. Right now, the primary interface for Google Docs is via web browser. In a modern educational setting, the last thing you want your students doing is having a web browser up in a classroom setting. Facebook and Flash games provide a powerful distraction from note-taking and group work.  Also, using a web browser as the primary text processor leads to issues when internet is unavailable.

Fortunately, there are solutions. HTML5, an upcoming web standard, will allow for GDocs to be used offline. Additional coding using the Abiword engine, used in the "Write" activity, with the GDoc API, would produce a platform-and-web-independent word processor with local and cloud file storage.

What do you think? In what other situations would cloud-based collaboration pay off? What sort of scenarios should software like this be designed for?

Open 1-to-1, "Maple Syrup," and customization.

Last Friday saw the conclusion of the FOSSed conference in Bethel, Maine. The conference was a hugely influential experience for the project and provided me with faces to go with three usergroups I've been designing for, Administrators, Students, and Teachers. I have quite a few thoughts I'd like to get on paper.

After my introduction to the Open One-to-One computing project running in Maine, Caroline and I have come to the conclusion that it may serve as an excellent platform to base the USR Bootable Flashdrive off of. With this in mind, I'm helping to evolve the O121 image development system using the LiveCD customization process used by the USR project to create a live O121 flash drive. Nicknamed "Maple Syrup," named after the popular sugar product of New England, this system will serve as a technological tool for K-12 education.

Right now, the Open 1-to-1 image is created manually, by modifying a fresh install of Ubuntu Netbook Remix on the target hardware. This approach produced a disk image implemented in several schools across Maine to great success. While making deployment quite easy, this approach makes virtualization and live-CD / live-USB creation nearly impossible.

To fix this, I found a means of converting the write-to-disk only image provided by O121 into a more useful form. The program Remastersys allows one to create a bootable ISO image from a running Ubuntu system. From a bootable CD ISO, I can use my existing ISO-USB script to create a persistent flash drive install.

As a refinement of the above procedure, I intend to use a few scripts written by David Farning to create a bootable and installable disk image entirely programatically, from scratch. Modifying his shell scripts  for use with the O121 & "Maple Syrup" project will greatly simplify the build process and expand the usability of the resultant image. Alternatively, there exist several alternate procedures to develop a bootable & installable disk image based on an existing LiveCD.

My next step is to flash the most recent O121 image to a spare netbook, install, configure, and run Remastersys, and debug / modify the resultant ISO.

FOSSed - Day .5 Reflections & Update

It's lunch break at FOSSed, a conference in Bethel, Maine, addressing Free-and-Open-Source-Software solutions for education. 

The morning breakout I attended was an in-depth look into the Open 1-to-1(open1to1.org) project. 

Open 1-to-1 is an answer to the Maine Learning Technology Initiative, a program running since 2002 which put Mac laptops into the hands of 7th and 8th graders across the state of Maine. MLTI serves as an excellent example of technology implemented in the classroom.

Open 1-to-1 is a framework for using Linux in an educational setting from grades 7-12. Using a customized version of Ubuntu Netbook-Remix(UNR) and affordable Asus netbooks, Open 1-to-1 has provided an affordable means for many highschools to implement 1-to-1 computing across Maine and across the world.

I recently installed Ubuntu Netbook Remix on my second-hand HP Mini 1000 Netbook. After a little bit of struggling with Broadcomm drivers, it works fine and has proved to be a very convenient tool

My "Favorites" view. Note Chrome, Qalculate, Sugar, and TweetDeck.

My "Home" view on Sugar on Ubuntu Netbook Remix.

Everything here but Turtle-Art-For-Arduino and Read works!

Yes, Write works. I fixed it. 

Write was broken because of an Ubuntu dependency screwup with python-abiword.

Editing control.tar.gz/control and recreating the Debian package solved the issue.

Including Sugar with O121 will expand the target audience down into the K-6 grade level. O121 will serve as a customizable base for education all the way from Kindergarten up through 12th grade. That's cool.

Progress! [A Visual Guide]

Just in time for Wednesday's FOSSed conference, everything is running awesomely.
I wrote a BASH script which turns a customized ISO image and a 2GB flash drive into SoaS, USR-style.

• the good news: most things work
• the bad: some important things don't

Check it out.

This is the "neighborhood" view, as seen on my laptop.

It successfully connected to a WPA2 WiFi network and to a Jabber Server, out-of-the-box.

Yep, the whole thing is actually running on a flash drive.

I've been developing with 2GB drives from MicroCenter. 

Speed is a little lacking, but nothing overt. Also, these guys are CHEAP!($6/ea)

Yep! Activities work too! Here's the ever-popular "Speak" activity.

When I said some important things weren't working, this is what I was talking about.

"Write failed to start"

This bug also affected "Read" - two of the most important activities, hands down.

I'm betting that the issues here are connected to packaging - both Read and Write depend upon external programs and libraries. There're kinks to be worked out as we rebase from the Fedora versions to the Ubuntu versions.

In summary: We're close. It runs stably, but not flawlessly. Working closer with dfarning on customization and with the SEETA crew to tackle bugs will bring quick progress and a polished product in no-time flat.

Week 3 + 4

Weeks three and four were a whirlwind of accomplishments. From coming to a more-or-less functional understanding of the procedure around customization, to successfully creating a bootable USB disk loading into a working Sugar environment with full driver support on my new Latitude E6400 laptop, huge strides have been made.

Unfortunately, I'm on a very crunched timeline here - FOSSed, a conference dedicated towards free-and-open-source-software for the dedicated purpose of education, is on the 7th.

My goal is to have a stable, functional, and distributable version of USR on a bootable CD drive between now and then. I've made quite a bit of progress, but there's tons left to go.

In the interim, here are a few photos of USR running on my laptop:

The "Maze" activity running on my laptop. 

Visible is the 1gb persistent flash drive I installed the Febuary USR build onto.

The Sugar home-page running on my laptop. 

If you look closely, you can see a graphical glitch in the top-left corner. 

Icons aren't properly being displayed - hopefully this will have been fixed in the July release.

That's all for now. I'll be blogging from FOSSed and will likely post tomorrow, after an all-day work session on programmatic flash-drive generation and bug squashing.

Weeks 1 + 2

The first several weeks of working on this project consisted of less work than expected. I met with Manu when he was in town for a conference and discussed the future of the project and the roles we were to play.  I spent extensive amounts of time re-familiarizing myself with the current state of the Sugar project, the Sugar Labs organizational structure, and discussing, briefly, the Debian/Ubuntu packaging system with Luke.

Delayed Introduction

This blog was created to chronicle my work with SugarLabs on an Ubuntu-Sugar Remix.

The posts are being amalgamated as I transcribe my by-the-monitor notes and merge them with the handful of textfiles in my /home/it/USR folder.

Over the past month or so, I've been working with dfarning, lfaraone, and Caroline Meeks, in addition to a team of programmers from the Indian SEETA, to rebase Sugar, the graphical interface and top-level software best known via its use on the OLPC project's XO-1, onto the Ubuntu Linux distro.

The goal of this project is a bootable thumbdrive providing the 'meat' of the One Laptop Per Child education project in a hardware-independent manner. Flash drives are cheaper, more readily distributable, and, in the first world, much more compatible with existent educational paradigms. Most modern schools have computer labs featuring hardware enough for a classroom of students at a time and many children have computers readily accessible at home. The constant presence of computers more powerful than the XO laptop in the lives of 1st world children slightly invalidates the need for a separate 1-to-1 computing paradigm. A 1-to-1 stick paradigm, however, can be implemented successfully for 5% of the cost.

Luke has been working under David, primarily focusing on packaging of Sugar activities for Debian & Ubuntu.
SEETA has been squashing bugs left and right and tackling testing.
David has been working on customization of a bootable Ubuntu image for out-of-box use with Sugar.
I've been working on taking David's customized image and turning it into a persistent bootable USB stick.

Upcoming posts will provide greater detail about my efforts.