Monday, April 4, 2011

A Manaiakalani Post - geek speak alert

It's been a long time coming.

Things are looking extremely positive for the moment.

The backports package for the network drivers seems to be solving a great deal of my issues.

Making the connection system wide came with an unintended consequence - the connection was made as soon as the computer was booted up rather than after the user had logged on.

This means more learning time rather than dealing with connection issues. Combine this with a monitoring script that looks for certain states in Network-Manager and the system and you've got a really stable connection. One school (300 netbooks) now has things configured in this manner and the issues I was getting have all but completely disappeared.

The remaining issues are things which I know about. The boot up theme for example, hides what's going on. This is an issue if it's doing something like checking the hard drive, as this can take a little bit of time, and not knowing what's going on can leave the user feeling bewildered and frustrated that their netbook just isn't starting up in the usual way.

Schoolzone, one of the two options for web filtering for school's is presenting all sorts of problems. When I realised that the gnome proxy tool doesn't set the authentication details for apt-get, I added them in manually. Only to find that Schoolzone blocks bzip2 files.

My first thought was to talk to Schoolzone about unblocking bzip2 files except that I would guess that this situation arises quite often. Trojans/Viruses which bring down a payload (for Windows environments at least) probably have things compressed and bzip2 seems as good a compression format as any. So I'm wondering if I can write an overlay to "apt-get update" which pulls down uncompressed files and does the right things to them using a server somewhere on the Internet to store the uncompressed files.

However, by the time I get to doing something like that, all of the schools in the cluster will probably no longer be using schoolzone which would make testing difficult.

Last week I had the pleasure of meeting Dr. Helen Barrett and her daughter, Erin.

It was a dynamic situation. It was interesting having someone around the school who I was able to leverage off. If I was thinking about something, I could ask them for their experiences of it. First off they asked me why I hadn't included Openshot on the netbooks.

There are actually 2 reasons.
  1. Openshot hasn't impressed me in terms of stability.
  2. The screen real estate needed for Openshot makes it less than ideal on the netbooks.
But then I started thinking. What would happen if you were able to find software to replace the classroom computers? In which case you could replace all of those computers with monitors to plug the netbooks into. 5 or 6 to a classroom would do. This would lead to savings. The upgrade cycle on monitors is a hell of a lot longer than on computers. If they're unable to work on the monitors, they still have the netbooks to use.

They then said "At the very least, audacity should be on there". So I've started trials. I've shown a few of the more curious students how to install it and have said "go forward and play". One has even offered to make me some instructional videos.

The problem is that this is at odds with the Manaiakalani project. The netbooks are a pencil with which is used to write/draw etc. on the book which is the Internet.

There is a precedent. The netbook image comes with games already installed - chess, tux math, tux typing (though I'd be quite happy to remove the tux series of games as they a. lock up the sound controls b. aren't the most stable of applications out there) as well as the Ubuntu default games.

None of these applications have saved files which matter. But if you were to create a sound file with audacity, that would have to be stored somewhere - and given that the prevailing attitude to this is that everything on the netbooks should be considered disposable (in terms of software and user-generated content) then this does present an issue which I would love to solve.

To this end, from the very beginning, I created a separate home partition with the thought that eventually I want to be able to empower the student technicians to do just about everything.

I've been throwing the idea around for a while but haven't had a chance to develop it just yet. I think I might have mentioned this in this blog before - I'm calling it "The Keys to the Castle". The idea is that they boot up from a usb drive and it then presents the technician with a range of options. Things like:
  • Reinstall system (preserves user files - so long as they're in /home)
    • This would require saving the /etc/passwd file and putting it in place again after doing the re-imaging.
  • Destructively re-image system
  • Reset passwords
  • Hard drive check (fsck)
  • Reset guest (To start initial login script)
In terms of individual applications misbehaving, most errors can be removed by removing the user's configuration file for that application. i.e. errors with wxcam (the webcam software) can normally be solved by typing:
rm ~/.wxcam

Probably the most destructive thing about reimaging a netbook is the loss of bookmarks in Firefox. So I would love to be able to do something like, even when doing a destructive reimage, saving those bookmarks and have a procedure for putting them back in place once the user's account has been created. It would be even better if I could find a way of putting those bookmarks online - this might be fairly trivial given the services out there.

So for my "holidays" I'm looking forward to a week of sleep and then time to get the sound card working as it should (I've had it working once though I have to sort through the procedure - i.e. is a Kernel upgrade needed?) and then starting development on "The Keys to the Castle".

I've been talking about making this a nationwide project. The idea being that it should be an overlay to standard Ubuntu for school children. So removing Facebook integration, blocking applications which are even slightly objectionable (hot-babe springs to mind though it's only in the restricted repositories) and hopefully starting development on a range of educational software with the user as the focus (Childsplay was very much a horror in terms of it working well - keys would stop working etc. The Tux games I've mentioned before).

I think I'm in a pretty good position for educational software. There are more than enough stakeholders within the environment I'm working in now. So this is a real possibility if/when I find the time.

Of course, this needs to be balanced against my Manaiakalani responsibilities. I know this probably sounds silly but when I do start getting paid for my role within the Manaiakalani project, it's very unlikely to be for any development. I'm more likely to be more... a technical coordinator (I rather like the term "facilitator" but that's off the table given that it'd be seen as a clash in titles) with development being a value add rather than my core job.

No comments:

Post a Comment