Slashdot Mirror
Innovation on the Edge?
MCassatt asks: "It's a truism in many fields that breakthroughs come from the edge: the scandalous Impressionists become pretty pictures for posters and umbrellas; the world of science fiction becomes the world of science.
The wonderful, the fantastic, and the mad of today are tomorrow's mainstream. Are there examples of this in computer science? Not extreme programming, but extreme programs?"
Lots of fun little things to place with at Google Labs.
http://www.lmcdms.com/
DMS is the US Government's international secure email implementation. At a glance it looks like a bunch of crappy obsolete code and operating systems trying to do email, but when you stop and think about what is DMS is doing, it is pretty damned impressive.
the thing is if a program is really innovative and radical you really cannot tell. example an example of new radical science in work would be an "ionic engine" it took princeples found from earlier sciences and applied them. we don't have this in CS. Nobody comes up with a theory about how "the computer space" works, and then tries to prove it, because everything is pretty much well documented and everything is understood because we created, so you really can't have really extreme programs. unless that is if someone uses a function really weird and gets it to something else, and i really don't see alot of that.
For The Best Jazz/Hip-hop fusion > COlD DUCK
I'd like to put forward this Turing machine, implemented using the rules of Conway's game of Life. It astounded me when I first saw it, and it astounds me still. Have a look at some of the components using the provided applet. If you've ever played with Life, you'll know how hard it is to create anything non-random at all.
Sweetcode often has interesting pieces of programming too.
Having been involved in the development of Kai's Power Tools, I'd have to say that Kai's user-interface designs had a strong influence on what's out there today.
Our philosophy while writing those programs was based on the observation that existing UI paradigms were created for processors hundreds of times slower than current machines; why not leverage that power to create interfaces beyond the standard buttons, menus, and 16x16 pixelated cursors?
Say what you will, the OSX Dock (for example) is indisputably Kai-like. I think that's a good thing.
Weeks of coding saves hours of planning.
Software that enables one to turn a bunch of ordinary off-the-shelf computers into a distributed cluster to run message passing programs on were pretty radial at the time, but now it seems everybody does it. I run my codes just as often of Linux clusters as on big IBM SP/3 machines, and for a lot of tasks, the Linux clusters cannot be beat.
I think running the Microsoft Word paperclip applet should be considered an extreme sport, at least. I think, on a serious note, that more and more people are going to start using apps that allow them to view data constructs in visual terms, like the network map thingamajig I saw for instant messaging the other day. It allows you to see circles, cliques, newbies, etc., and how they're distributed through the IM world. New ways of looking at data for those visual types.
I would count the early arcade games, and the Apple II games.
These machines and programs jammed an enormous amount of programming functionality into incredibly tight spaces. Many of the old arcade programs ran on 4K, 8K, or 16K 8 bit computers, and ran on machines with clock speeds of under 1 MHz, and effective instruction rates of mere hundreds of thousands per second. Even a fully loaded Apple II gave you under 32K of actual program space to work with, once you subtracted the low RAM, the hires graphics areas, and the BASIC ROM space, and people did a whole lot with that 32K.
The last two games I've purchased (Simcity 4 and C&C Generals) require minimums of 500 MHz and 800 MHz processors respectively and 128M of RAM. Of course, they do a lot more, but they are certainly not 500, or 800, or 8000 times as entertaining as the Cocktail Space Invaders machine that graces my hall entryway and is such a hit when we throw parties.
Early arcade games were heroic, wildly successful efforts. Truly examples of extreme programming.
Without a doubt, video motion detection is going to be huge. Programs like Homewatcher, GOTCHA, and many others (I'm too lazy to set up links) can sense motion very accurately, take timestamped images, upload them to a webserver, send them via faz and email, call your phone, run external programs, etc, etc. If you live in a dangerous neighbourhood like me(and if economic downturns persist, perhaps you soon will) they are hugely useful. Couples with cheap cameras and cheap low power hard drives, systems like this could make crime very dangerous for the potential thief if they were extremely widespread.