J5’s Blog

I just read an AP article talking about the design of the XO’s Sugar interface.

Born from Davidz’s work on OLPC we now have a Fedora Core 6 LiveCD and tools for building your own Fedora based targetted LiveCD. Go, go, go DavidZ.

The aKademy videos from Dublin have been posted in OGG Theora format. I you leave for the holidays with my talk, Competition and Cooperation - An honest look at the dynamics between the KDE and GNOME communities. Sound is a bit fuzzy but you can hear most of it.

Happy holidays.

Last weekend was my last cooking class. It was all about sauces. I was a bit out of it since I had just flown back from the OSDL DAM meeting in Portland and got home around 2 in the morning. I opted to make a pasta dish with a beshamel cheese sauce because I wanted to learn how to make my own pasta. The pasta took a bit to make so we set it aside and used premade pasta. Not my best work but it still turned out tasty.
My best dish so far was the swordfish with jalapeno, avocado mayonnaise I made during the grilling class.

All in all, the classes were well worth the money. I am looking into taking some more but I don’t want them to totally interfere with the ski season so I might just take individual classes. So what shall it be? A tour of France or a tour of Italy? I’m not sure yet.

In this day and age when we are all looking for the next killer app sometimes one finds a gem if they just look a little smaller. I just stumbled on a little python app called Gourmet which is a personal recipe database. Some of the nice features include a measure converter, serving multiplier and shopping list generator. It is a bit rough around the edges but mostly all it needs is some HIG love. I would also like to see a weight to measure calculator since most chefs go by weight and not volume when measuring things like flour and butter. I ended up packaging it up for Fedora. If I have time I will start to hack on it. One of the first things I want to do is free its dependency on both python-imaging and python-reportlab. Cairo will most likely suffice without bringing in dependencies like tk, tcl and tkinter.

Rock on Collabora. These guys are really building a great Free Software company and are an asset to GNOME and FOSS. If you have a need in the collaboration space I highly suggest hiring these guys. Apart from rocking on OLPC video conferencing they have Simon McVittie rocking out the next generation of D-Bus python bindings which replaces the Pyrex backend with a pure C implementation that should be easier to work with going forward.

My next task is setting up a repository where developers can contribute and host application bundles which can be downloaded directly to an OLPC machine and installed. This will allow people to try out the different application that will be available and make it easier for countries, teachers and kids to get involved in developing targeted content for the device. As it stands right now they have to go through me to get activities into the builds and believe me, taking myself out of the equation will make things run a lot smoother.

As I get ready to go to the OSDL DAM (desktop architects meeting) in Portland I am suddenly reminded why standards for standards sake can be harmful. This is the case of RelaxNG vs. W3C XML Schemas. I remember looking at W3C XML Schemas wanting to write a Schema to XML parser code generator back when I was in school. I never got around to building it because Schemas just was huge and unintuitive. What I did get out of my exploratory phase was someone pointing me to the OASIS ’s RelaxNG work which was way simpler but not created at the W3C.

Now Tim Bray has thrown his hat in the ring and said RelaxNG is the clear winner. This is an example of why there needs to be an evolution of standards from real world use. The competition between the two standards was good in many respects. It allowed different approaches to the same or similar problems to be evaluated. The shame in all this is they have both become standards before the competition could organically pick a clear winner. By many accounts XML Schemas is a bad standard because of its complexity but, because it is a standard there will always be an argument of why it should be supported. It is a bad argument but there you have it.