africasource
I'm at a free software conference, in Africa. It's wonderful to meet other hackers in the continent, and especially seeing quality home-grown hackers working on home-grown solutions. The participants are about 65% African and 35% international. Some of the internationals are from other developing nations, and some are from Europe and North America.
Now, it's Africa, and it's a funded meeting. So it's not like a self-organized meeting of people sharing the same technical passion, like GUADEC or something; it's a European NGO that, partnering with some local organizations, has secured funding to bring together participants from all over the world.
Naturally it has an NGO flavor. Organizers and a social focus prevail. This isn't the most natural context in which hackers can get together, but neither does contemporary Africa provide a natural context for free software development in general. There are many social problems (lack of computers, expensive bandwidth, etc) that affect the ability for promising programmers to learn, cooperate, and code. Once someone has broken through the barriers to participation, the pressure to make the dollars tends to outweigh the more abstract principles of free software.
But people do care, and hackers somehow come out of this. Guido Sohne amazes me. Organizers like Joris Komen from SchoolNet Namibia do hard, enthusiastic work towards equalizing the social problems that prevent the development of good hackers.
And so, in the end, the social work and the hacking are symbiotic. I guess now that I realize that, it's not such a bad idea -- still, next time, this one needs to be African-organized.
If only plane tickets between Nairobi and Windhoek weren't so damn expensive!
gstreamer
The maintainers (and yes, they do exist -- they're just plural) of GStreamer put together a nice release. It was a lot of work, but we're pointed in the right direction now. Nice work, guys.
gnome and languages
I think it's clear to most people that graphical applications will not continue to be developed in C. My primary concern (as the primary author of guile-gnome) is that Gnome applications can still be developed using any language with a binding.
I'm not convinced that we need a common language runtime, though -- GObject, coupled with bindings generators, already provides a sufficient mechanism for integration. It would be more difficult to make guile (and I'm sure any language) on runtime than to make bindings for C libraries.
So I don't understand, really. Is the argument that GNOME platform libraries should be able to be implemented in a higher-level language? I can see the reason for a runtime there. Maybe others can shed light on this issue?
Comments are closed.