Tuesday, July 10, 2007

Learning about PGP and microformats

Because I haven't posted about anything remotely technical in a very, very long time:

Today I finally got a PGP key because I needed one to get an account on the OLPC dev server.. (For my parents: A PGP key is a special number that sort of acts like a "password" for messages. It helps make communications like email and such private.)

Aside from "it's something related to crypto," I didn't actually know what PGP was. (I'm still mildly fuzzy on it, so if anyone's into that kind of stuff and wouldn't mind explaining...) So I did a little reading, and got my key with the help of this tutorial. My favorite part was the way Phil Zimmermann challenged the US regulations on exporting PGP (they considered cryptosystems with 40+ bit keys "munitions"). Via Wikipedia:
Zimmerman challenged these regulations in a curious way. He published the entire source code of PGP in a hardback book, via MIT Press, which was widely sold in bookstores. Anybody wishing to build their own copy of PGP could buy the $60 book, cut off the covers and separate the pages, then run them through a scanner and an OCR (optical character - or text recognition) program to create a set of text files containing the source code. One could then build the application using the GNU C Compiler and GNU Make utility, which are freely available on the Internet. The compiled files could be put on a server anywhere in the world, and anyone could download them. The principle was simple: the export of munitions--guns, bombs, planes, software--is restricted; but the export of books is protected by the First Amendment. Anybody could buy a book and cut it up, so any computer programmer with common UNIX skills could build the software. However, the question was never tested in court.
I also learned about microformats while working on the OLPC sample library, or rather, trying to find out how it ticks; it's all html and css right now. If you're looking for neat free childrens' books to read your kids, siblings, or students, check it out. They're multilingual, too. And we're looking for translators, people to read the books out loud and send us the sound-files (bonus points for , and of course... more books.)

Microformats are "adding simple semantic meaning to human-readable content which is otherwise, from a machine's point of view, just plain text." That was the idea behind Ad Libris (which is on the back-burner for now as we gather content to test it out with).

In other words, we read text (for instance, this blog post) and it means something to us; you can say "ah, the post is about pgp and microformats, and Mel wrote it." But your computer sees a bunch of letters (or ones and zeroes, really) and can't answer those kinds of questions. Microformats tell your computer that hey, Mel is the author! Hey, this post is about PGP! so it can begin to answer intelligent questions like "Computer, show me all the stuff Mel's written about OLPC in the last year" without someone having to build a custom database of time-sucking glory.

Nifty stuff. Happy Mel. Man, the world is full of great ideas.

2 comments:

nikki said...

Would that make microformats analogous to, say, tagging journal entries and being able to search them by tags later on?

Blues said...

@nikki: almost, except microformats let you mark up individual components of the page with information far more detailed than a tag.

For example, I could put up my name and address, and mark up each component as first-name, last-name, nickname, street-address, locality, area, country, etc., and someone with the right software (say a greasemonkey script), could directly import that into their address boo.

Strange that I actually got to this post searching for a microformat about PGP.