Brazil’s free software forum: to the favelas

By Andy Oram
April 20, 2006

I came to Brazil to deliver a presentation on documentation. But I quickly realized, after the organizers accepted my proposal, that I would come here to learn as much as to teach. I am concerned broadly with the information needs of Brazilian computer users, administrators, and programmers--and I want to understand deeply what the needs and opportunities for documentation here are. In short, I’ve come for information about information about information technology.

As it turns out, education is also the goal of the One Laptop Per Child project, which had a strong presence here. (Project Athena and X Window System pioneer Jim Gettys came to promote it--just one of the many luminaries from North American who made the long journey down to show up at this conference.)

The OLPC systems are trying to pull off a stupendous achievement: to provide interfaces that are richer and easier to use than any interfaces that have come before, while rigorously cutting down on the memory and power requirements that have always accompanied advances in interfaces. In this way the project hopes to create systems that, when manufactured in the hundreds of thousands or millions, are cheap enough to offer in bulk to entire villages.

Thoughts on development

After chatting with Gettys yesterday about One Laptop Per Child, I happened to pass today, on the highway to the conference, the first favela I’ve ever seen. On a flat area, a substrate composed completely of garbage and detritus, houses of tin and cardboard were arrayed. A few people of various ages wandered the open field, along with a couple donkeys hitched to carts, and a horse that bent its head down as if confusedly believing that nutritious ground cover would meet its teeth there, instead of the filth that spread in every direction. While I felt in one sense as if I were gazing at a raised-up vision of the nineteenth century, the raw materials of the depressing scene came entirely from the cast-off junk of modern life. And as far removed as the scene appeared to be from order and progress, the shacks were laid out in perfectly straight lines as if directed by a city planner of the damned--because the poor need efficiency the most.

It may have seemed callous, therefore, for me to go tonight to a famous Porto Alegre restaurant called CTG, which offered not only an open food bar, but waiters coming around to tables constantly with every kind of meat cooked in the most sumptuous styles. Nowhere could “zero hunger” seem so much like not only a program but an enforced requirement. But I have reached the decisive position that it is not inconsistent to strive to improve the lot of those who suffer, while appreciating the good fortune (which can be temporary in any case) to be able to enjoy life. What do we hope to share with the poor, if we feel no hope in our own lives?

The restaurant also featured live performances of a folk music called chorinho, which I’m told is local to the region around Porto Alegre and was very popular in the 1950s. In a somewhat artificial but still impressive display, dancers came out to show off the dances that traditionally accompanied chorinho. Some of the moves reminded me of North American country dancing, while others featured near-martial displays involving spears, long knives, and ropes that I suppose were meant to represent whips. I wondered whether the rural folk who created this art form would have ever come to it if they had possessed One Laptop Per Child.

[Dance performance with costumes]

Costumed dancers reproducing local chorinho folk tradition in performance setting.

Technological optimists tell us that information technology can help people preserve their local traditions and culture, but I worry. And while the favelas may seem to be unrelenting in their degradation, with nothing positive to offer, it must be remembered that the great innovation of samba came from favelas in Rio de Janeiro. (Computer users who enjoy the Samba file and print sharing software might consider making donations to organizations who work with the poor in Brazil.)

One Laptop Per Child has so many natural enemies--governments afraid of the spread of communication skills, companies such as Microsoft afraid of competition, teachers afraid of new educational paradigms, and the eternal adversaries posed by bureaucracy and fixed ways of doing things--that I can’t bring myself to get too enthusiastic. But only for fear of being disappointed. OLPC is a magnificent project, in concept as well as technical potential. Some of points made by Jim Gettys today in a very professional presentation include:

Project organizers are convinced that modern applications can be much leaner in memory requirements, with no loss of functionality. Improvements that projects make in a drive to fulfill the needs of OLPC will end up benefiting all computer users.

Power consumption is also way out of control; rich user-interface applications such as Flash and applets perform scads of unnecessary processing that eat up battery power. Again, changes in this area will help the entire computer industry.

OLPC may be the wedge that finally brings IPv6 into mainstream use. This is partly because the project will require millions of new addresses if it comes even close to meeting its goals of widespread use. It’s partly because the project eschews the IPv4 work-around of hiding systems behind firewalls with network address translation; it wants each system available to the wider world for administration purposes and peer-to-peer applications.

In short, Gettys says, nearly every innovation driven by OLPC will improve things for all computer users. And to underline this, he asked projects to do OLPC-related work in their main project trees, instead of forking them as if to ghettoize OLPC and treat it as an anomaly.

I talked today to one person who works with the poor: Joice Kafer, webmaster for the cooperative Solis that I described in yesterday’s blog. With about ten other young people educated in information technology, Joice volunteers to work with the children and elderly in poor neighborhoods, to teach them skills they can use for personal communications and for employment. The project, Gnurias, concentrates completely on free software, and uses systems donated by businesses or set up in local schools (where the computers often sit unused because no one there is trained to work on them). It’s too new to evaluate its effects, but the participants hope it will increase people’s sense of the world around them, keep kids in school longer, and raise economic levels.

Another social project of a somewhat lighter nature (although exceedingly serious underneath) is the cartoon series Hackerteen. Under the guise of classic good-and-evil stories with cliff-hanging disaster plots, it offers teaching to young people on many levels: technical background on how the Internet works, guidelines for protecting oneself online, and the ethics of participating online.

Information gaps in Brazil, and who’s filling them

I’m very interested in the Brazilian market for computer documentation in Portuguese. In many countries, even though English is a second language, it is so widespread among computer professionals that they can get their information needs in English and provide an unsatisfactory market for translations. Brazil seems different: a country burgeoning in IT people--there’s even an LPI partner there--who need documentation in their native language.

One interviewee gave the opinion that most Brazilian programmers know English. Other people gave me a very different opinion. (There may also be differences between programmers and other computer experts, such as administrators, but I sensed that programmers were not totally ready for English documentation.) The second opinion is that only 15-20% of computer professionals feel comfortable reading a book-length treatment of a computer topic in English. Certainly, professors hear complaints from students if they assign an English text.

To make the situation worse, those who are fluent in English tend to leave for enticing jobs in other countries where they can earn a lot more.

I visited the booth of a relatively new Portuguese publication called Linux Magazine, a subsidiary of the magazine by that name based in Germany, and widely read in many European countries. (It is not related to the U.S. publication Linux Magazine.) Editor-in-chief Rafael Peregrino showed me copies of the main magazine, aimed at system administrators with modest training, and a newer magazine for end users. He said the system administration magazine is widely distributed to bookstores and sells about 7,000 copies a month, including 3,000 subscriptions. He also pointed out that he knows each copy is read by an average of three to four people. His young company is now profitable--and it even has competition.

And then, of course, to fill the information gap, there is this conference. The energy is infectious. The floor is still bustling at eight o’clock at night (yes, presentations go from 10:00 to 21:00 without stop). The scope is completely unbounded, so long as the topic is free software. At one presentation I hear of user design changes in the desktop, while at the next I find out the kernel API functions and new features that make virtualization better.

Some people say the energy is somewhat less than previous years. This can be attributed to the natural effect of aging--a conference is always most exciting in its first few years--and to competition from LinuxWorld Brazil, a recent addition to the scene. But I’m very glad I’m here.

O’Reilly translates a number of its books in Portuguese, but I am looking for ways to develop stronger community bonds that will help us choose books that can sell, and find the most appropriate publishers.

Other blog postings on this conference:

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.