From: Chris D. <da...@ya...> - 2004-02-08 05:08:16
|
First off, let me say thanks to everone working on this project. I have played with it a bit and it is quite cool. I did some VERY informal benchmarking with some neural network stuff that I wrote and got some nice numbers. I ran the program under cygwin and then recompiled and ran it under colinux. It looked like there was only a 15% performance hit under colinux. (the program uses a mix of floats, ints, arrays and linked lists with a little bit of file I/O if you were wondering) Anyway... to my question... I guess I am wondering if the TAP driver is ONLY needed for networking. That is, I know that if you try to run colinux without it, it dies. Could colinux be configured to continue "booting" even if the driver is not installed? Why? I was just thinking about how cool it would be if I had colinux and a small disk image on my keychain flash drive. Then I could just plug it into whatever Windows computer I happen to have handy, click an icon and have linux up and running. No drivers required. (nice if it not YOUR computer) I realize that the xml configuration file has the location of the disk image hard coded, but it would be easy enough to address that. And yes, I know that flash has a "limited" number of write cycles... Just wondering if it can be done. Thoughts? Again, thank you for such a great project. --Chris ===== There is a very fine line between "hobby" and "mental illness." --Dave Barry home: http://www.dahlweb.net |