#define AUTHOR "hvozda@netcom.com ("Eric S. Hvozda")"

/*
 * > Anyway, 2 days is still about 47.9 hours too long to change a simple
 * > thing like that. 
 * 
 * Whoa!  Who here is building kernels in 6 mins flat from scratch?
 * 

I think the point was that you shouldn't have to rebuild a kernel
to turn on/off the "autonice" feature (it should be a dynamically
tweakable option).

For which 6 minutes is 5.9 minutes too long... :-)

 * (takes me 28 mins on a DX4-100 i386, 512Mb IDE, 20Mb RAM)

Remember:  DX4 is super-fast CPU speed, but I/O is still bound to its normal
rate which cannot be faster than the undivided clock speed of the CPU.
In any case, the upper limit of I/O on EISA is still 33 MHz, last I looked.

I used to have a kernel in 12 minutes on a 486/66 512MB IDE, 16MB RAM.
I don't think 6 minutes would be out of the question on a PCI '586.

[If we could get it going on the new sun4u series, you'd have a kernel
 from scratch in about three minutes, and the world in four hours.]

 * 
 */

#undef AUTHOR	/* "hvozda@netcom.com ("Eric S. Hvozda")" */




				--*greywolf;
--
America is quite possibly the only country to go from barbarism to
decadence without the requisite intervening period of civilisation.