Nortel Tech Conference Day 3

At the general session , Al Worden, an astronaut on the Apollo 15 mission, spoke about the engineering challenges during the Apollo program. It made me think about how all of the macro engineering of the engines and the micro engineering of the computers must all integrate perfectly, or the results could be fatal.I think OLPC could benefit from this kind of thinking by pairing some of the large scale problems with the XO laptop with software or hardware solutions.(and vice versa)

One of the other speakers today was Andy Lippman from MIT, who talked about Nortel's partnership with MIT and some of the benefits it has brought. He also showed some of the cool things the Media Lab and other sections of MIT re working on, besides the XO.Also seeing the concept XOXO pic again made me think of some of the cool things in our future.


I did not go to as many sessions today as I did on Monday, but some of the stuff today really inspired me. New technology focused schools, fully realized mobile broadband, and next generation (maybe we could call this Web3.0?) web apps are all areas that Nortel is working on,and the work I've seen looks amazing.

Tomorrow I head out to the Kennedy Space Center, which will be really interesting to see with a bunch of engineers.

Nortel Technical Conference Day 2

Other then the general session,which focused on the changing face of college education, the events on Tuesday focused on very technical discussions which had no direct relevance to me. So Michael and I decided to go to Universal Studios.


Once we got there the first thing we did was ride all of the major coasters in the park.The best was probably "The Hulk", a giant steel launched coaster with many loops and corkscrews.


Of course, it was always fun to try out the "wimpier" rides. Many of them were in the shade and helped us escape the heat of Florida in the summer.



After exhausting ourselves at Universal, I got a rather funny urge to see how much it costs to see a movie in the universal resort area , completely out of curiosity and totally expecting it to be outrageous. Surprisingly, its actually about 30 cents cheaper to see a movie here then in DC.


Both Michael and I had been wanting to see "The Incredible Hulk " when we got back to DC, so we decided to just go now and complete the day. It definitely lived up to the hype and I would encourage those who are fans of the comics to see the movie.


For dinner we went to Boston Lobster Feast , which had a buffet of all you can eat lobster and other seafood.



All in all , a great day of relaxation in preparation for busier days ahead.

Nortel Technical Conference, Day 1

One of the many opportunities that working with Nortel LearniT has given me has been an invitation to the Nortel Technical Conference in Orlando,Florida. I attended discussions ranging from Nortel's business plans to ideas to enhance the Vancouver Olympic Games.
Some of the sessions I attended focused on mesh technology which is of great interest to our readers so I will elaborate a bit on those.



On Monday, I gave a demo of the mesh technology on the XO, which generally worked out successfully. First, 4 XO's were set out sharing the Chat application, all connected using a Schoolserver. Then I had volunteers try out Record and Distance sharing the former of which worked quite well, the latter not so much. I suspect that Distance is rather sensitive to version differences, so I will have to look into that. After that I showed off some of the schoolserver technology such as the DNS configuration that allows for the Schoolserver to push out a domain of the teacher's choice for the students to connect to. Incidentally all XO's connected through the school server correctly utilized web caching , the first time I've gotten it to work flawlessly.

The session before mine also focused on an implementation of mesh networking. Bob Withrow, the Director of Networking Research at Nortel, gave a demonstration of open 802.11s technology (seen at http://www.open80211s.org/) used to route calls locally rather then through a carrier. For his demo, he used 4 XO's hooked up with special antennae combined with several neo1973 phones running the openmoko software.The XO's were supposed to be used to measure mesh activity and to show how the link reconfigured as the phones moved , but the heavy wireless traffic at the conference prevented this demonstration from working correctly.I did learn quite a bit out how mesh could be implemented into consumer products and the challenges involved.I wonder if Google has looked at this for its Android platform.


More coming soon!

Matt Gallagher

Bugs... They're everywhere!

After reciving two new Mass Production XOs from OLPC via the developer program, I've been testing them like crazy. I've reported 8 new bugs, and more are to follow.

In addition, I submitted a simple patch to Sugar to fix an annoying traceback that is issued when "sugar-install-bundle" is called with no arguments. Hopefully it will be accepted soon.

I'm just getting my feet wet so far.

Pycon 2008 Recap

I'm sorry that this is quite a bit late, but I 'd like to fill you in on some of the interesting thing that were accomplished at PyCon.

(Pic of attendees in main expo hall)

One of the first big OLPC related events at Pycon was a keynote by former OLPC employee Ivan Krstić. He focused on how the OLPC facilitates learning and the ways in which Python helps with that.

(Pic of Ivan's keynote)


Some of the other talks that focused on the OLPC project addressed issues that have plagued software development on the platform, such as a lack of proper unit tests in many applications.

(pic of slide from OLPC unit testing talk)

The second part of the conference that I attended was the OLPC Sprint. (for those who don't know, a "sprint" is a period of rapid collaborative development on a specific project, similar to the "hackathons" of OpenBSD fame)
Led by Mike Fletcher, the OLPC group made significant progress in creating several useful applications for the XO despite having no official representation from the OLPC organization( a fact that disappointed some)

(Pic of OLPC sprint room)

I worked quite a bit with James Hancock, the lead developer of GASP, as he started to work on an XO-friendly networking API and further Pippy (and possibly Develop ) integration.

(Pic of James Hancock)


An unexpected but familar face came in the form of Mel Chua , who I had worked with before on other OLPC related projects. She has been a great source of constructive criticism and advice for GASP , as well as helping us out with documentation.
(Pic of Mel Chua)

By the time I left PyCon, I realized that the OLPC project has many long struggles ahead, but there are plenty of people ready to meet the challenge head on.

Hello from PyCon 2008

PyCon 2008 was huge this year. Over 1000 people attended -- a 70% increase from only a year ago. The OLPC presence is everywhere. Participants used them to browse the web and take notes at presentations.

This morning one of the Plenary talks was an OLPC Update by Ivan Krstic. He gave a keynote at PyCon last year, so it was an good opportunity to look back and see how far the project has come in the last year. Most exciting was seeing his pictures from Uruguay and Peru, where the first two XO deployments are taking place.

I attended a morning talk titled "Programming for the One Laptop Per Child laptop" by Charles Merriam. It was a good discussion of what to expect from developing software for the XO and helped prepare us for the coming sprint.

This evening we are starting the sprints. Matt Gallagher, James Hancock and I are at the Twisted with Games BOF (Birds of a Feather). James is eager to add networking support to gasp, and the twisted developers are talking with him about using twisted to do that.

In the "it's a small world" department, Mel Chua showed up at the start of our sprint. She is a student of Dr. Allen Downey, the original author of How to Think Like a Computer Scientist. I've never met Allen face to face, but I've been working with him for several years on the book. Mel knew about gasp and is interested in sprinting with us.

That's all for now. I'll report back on the progress of the sprint in a day or two.

Logging on #gasp!

Jason Straw came through, using the Arlington Career Center webspace to host our bot and logs. You can see what we've been up to here. Also, the user mailing list archive is up here, so for those what want to pitch in, or just keep tabs on the project, check it out.

In other news, expect to see GASP 0.2 packaged for Ubuntu and the XO within a week! Exciting times, folks.

From the logs:
jhancock__: ayanami: I don't think I have meet you before?
ayanami: jhancock__: Error: "I" is not a valid command.