Slashdot Mirror
C-SPAN Interviews Wikipedia Founder
TrentL writes "Wikipedia founder Jimmy Wales (aka Jimbo) was recently interviewed on C-SPAN's primetime program Q&A. Topics included the origins of Wikipedia, governing philosophy, and criticisms from members of the print encyclopedia community." From the article: "I had the idea basically from watching the growth of the free software movement. So all of the software that really runs the Internet, Linux, Apache, the Web serving software, it's all written by volunteers collaboratively working together using free licenses. And it's really good quality stuff."
Wikipedia is amazing. A shining example of what people can do from working together as a community for the spread of information and making the world a better place. It is great to see Stallman's influence reaching to such extents that very awesome sites like Wiki are started and become what they are today.
I saw part of the interview and it's a shame that he didn't give credit to Ward Cunningham, the guy who invented wiki and showed what was possible wrt community building with the Portland Patterns Repository.
Wikipedia has not changed the nature of truth. Wikipedia has only made it easier to access free, democratic information, thus allowing readers to make more informed choices about truth based on a larger range of fact and opinion. It is, in my opinion, very much preferable to watching news or the discovery channel, where fact-checking is such a tedious and after-the-fact process that it barely ever occurs.
If you want to criticize someone for homogenizing the truth, look to secondary school and university educators and textbook publishers, who cannot afford to have a definite perspective on truth, due to lobbyist groups and bureaucrats.
You, and people like you are the reason Wikipedia has the problems it does.
Actually, Wikipedia *doesn't* have this problem, en masse. From a traditional computer security theoretical standpoint, Wikis are appalling. In real life, it seems that they do generally work. Maybe over time, they'll take some tweaking (as the content stabilizes), but the "it's prone to horrible malicious attacks" argument lacks a bit when you consider that it actually works.
As long as IP addresses are expensive IDs (i.e. a user can't just get another at will), the problem is partly solved, anyway. When I catch one instance of vandalism, I list other submissions from that IP, and start ripping out other changes. Vandals very rarely are useful contributors.
Someone that contributes 95% useful information with a few wrong things thrown in could probably cause some damage -- but nobody seems to want to really hurt Wikipedia thus far. [shrug]
Also, most of the vandals seem to be schoolchildren, and the vandalism is pretty amateur, whereas the regular Wiki contributors have worked such that the grammar and writing style of the bulk of Wikipedia is of excellent quality. Against this backdrop, vandalism tends to stand out -- someone who has graduated from high school with a solid English background seems to be less likely to be interested in running around vandalising other people's donations. "Teacher" is a popular article to vandalize, for instance, as are those of pop bands.
Slashdot sees a lot of trolls, but I think that part of the "troll psychology" is that trolling is considered fun -- successful trolling takes some skill, causes little or no damage (at least on the individual level, though Slashdot being flooded with trolls can get annoying), and people see an immediate reaction to what they've written. On Wiki, where vandalizing articles does hurt people, the most common reaction is just to see some inert text followed by the vandalism being backed out. There's no "modding up", and messages don't become part of a timeless archive (as they can be backed out).
I, personally, think that creating/improving vandalism flagging to Wikipedia would be one of the more useful research projects out there (i.e. this is applicable to a lot of things besides Wikipedia, successfully doing this can directly cause a lot of good, and there is interesting data mining research involved), and I'm guessing that if someone hasn't already jumped on this, someone will at some point.
Any program relying on (nontrivial) preemptive multithreading will be buggy.