Lots of people will view this as a sellout, but it isn't. The primarly difference between ESR and RMS is that ESR is pragmatic, and RMS is 100% principled and doesn't allow his principles to be modified by pragmatism.

ESR's stance is that the end (more openness -- and that means more Freedom, too) justifies the means (deal with more proprietary software now).

Lots of people will also claim that there is danger of a slippery slope here, that if we allow *any* proprietary software, then we won't know where to stop. I simply don't think that's true. As Linux increasingly dominates the marketplace and the world wakes up and realizes that Software Freedom just makes more sense, then we'll see a shift away from the new-old way of doing things. (The old-old way was when code was delivered to the customer with the compiled executable(s).)

I believe we'll see the same sort of progress in DRM and music/movie/etc. copyrights and the related P2P battles. As artists wake up and realize that the Internet enables them to survive in a different way than the current studio systems allow, many of these issues will morph and the current battles will go away. Why do you need a studio pushing your single in stores and on radio when the Internet can simply bypass these traditional advertising means and 'Net-based word-of-mouse advertizing can do all of/most of the work?

Of course, these processes will take years, and I think Linux-on-the-desktop will be the first one to see significant progress.

Especially if the community heeds ESR's advice now.

For a Natural Language Processing project, I wrote a perl script that would take all of your Gaim logs, build a database of things you are most likely to say, in the style you are most likely to say it. Then, as you typed, it non-obtrusively suggested the rest of the word or next word that you were currently typing. You could hit a key and it would fill it in for you

If IM bots get sophisticated enough, they could start doing something like this. Not for typing something out, but for generating messages from a person that sound a lot like them. Perhaps copying a message of sending a link they had sent earlier, just changing the link location. Then if the person replied, they could use information from the logs on how to best respond like the person.

Hmm, perhaps I shouldn't be giving anyone more ideas...

I should find some reference material for this, but I'm too lazy.

My understanding is that in the past the parties chose their own candidates without the help of the state governments. This generally meant that a few powerful members of the party -- the elite or rich -- chose their candidate in what was seen by the public as secret, corrupt "smoke-filled rooms". Especially in a two-party system, this bothered the voting public, so reforms were passed to help make sure there was a fair and open system for the parties' to choose their candidates.

Is it me or does the line from the article:

1. Lowering soldiers' core body temperature might keep them from overheating.

sound like they are discussing the issues surrounding overclocked soldiers.

Alternatively, you can build anything using nothing but XORS.

This is false. With only XORS, the only functions that can be computed are parity functions -- those that depend only on whether the number of active inputs is even or odd.

Centrallix - http://www.centrallix.net