Thursday, July 27, 21:20 CDT
I meant to go to sleep at a sort of reasonable hour last night, but instead I read The Last Unicorn, by Peter S. Beagle. Good book. Fairy tale, but aware of it... The kind of book I'd call the best kind of fantasy, except that I like the other kinds at least as much. If that makes any sense.

Less than a month left now before I'll be starting college... Weird.

Read a couple of interesting Napster / music distribution related pieces on Salon today. Why the music industry has nothing to celebrate, which isn't exactly news to a lot of us, but well written anyway, and Courtney Love does the math, which is a surprisingly good (and surprisingly clued in) rant from an artist's perspective about the music industry, 'net distribution, and all that good stuff.

(Hmm... Yet another distributed file sharing system. But this one appears to have a well thought out design and actual security. I think I'd like to try running this, once I have an always-on connection.)

Surfing freshmeat, I stumbled across gentoo, a graphical filemanager for Linux that uses a two-panes with buttons at the bottom design. I suppose the interface is sort of hideously cluttered, but it seems efficient and has some cool features. Maybe more purely useful than GNOME's included version of the GNU Midnight Commander, or at least a good complement to it.

One thing that sort of bothers me is that gentoo has its own huge internal list of filetypes, display characteristics and icons for them, and actions to take on different kinds of files. I'm not knocking this - it's an impressive amount of work, and the filehandling is pretty sophisticated. It can group things not just as .jpg's or .gif's for example, but as subgroups within the category of "images". Nifty, and actually kind of useful.

What bugs me is how much duplicated effort is involved here... It seems like every different free filemanager and desktop project, not to mention countless different applications, go to the effort of coming up with a set of filetype data and ways to deal with it. "Files with an extension of .html should be edited with such and such, and viewed in Netscape... Files with an extension of .jpeg should be edited with the GIMP and viewed with xv..." etc., etc. Why isn't there some kind of open, cooperatively maintained set of file data usable by anyone, with maybe some kind of library that would allow everything from textmode apps to flashy graphical desktops to take advantage of it? Use XML and some other buzzwords, and make it usable on pretty much any platform... I'd think it'd benefit a lot of different projects, especially KDE and GNOME, in a big way...

I'm probably just being totally ignorant here, but I *think* I see something where free software could take a pretty big step forward if someone with far more skill than I and an actual understanding of the situation did something about it. (Heh. Why do none of my ideas ever amount to anything, anyway?)