Excellent references. Thank you. One of the references mentioned a real mode linux kernel. I have a lot to learn. Probably I'm confusing the old real mode OS and the protected mode OS of today. Maybe accessing the ports can't be easy anymore. Still, its worth doing. All those controller lines with RS-232 really supplement the data lines which are fast enough for most uses. And the driver produces some real power. If you run into more references I would appreciate them. Thanks. >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<< On 9/9/01, 4:24:40 PM, Mike Hicks <hick0088 at tc.umn.edu> wrote regarding Re: [TCLUG] Help! Programming the UART.sdm: > Rick Engebretson <eng at pinenet.com> wrote: > > > > I'm trying to rediscover how to control the RS232 UART every PC has two > > of. This powerful interface could be more useful, particularly for > > process control or robotics. I'm not looking for a driver, unless it is > > highly versatile. > > > > In the original PCs one simply input or output to the UART BIOS address > > and the bit field was yours. The DOS driver for the COM ports began the > > process of hiding control by adding a new layer. > http://www.linuxdoc.org/HOWTO/mini/IO-Port-Programming.html > Though even that document recommends using the normal kernel I/O > interfaces. > http://www.linuxdoc.org/HOWTO/Serial-Programming-HOWTO/ > http://www.easysw.com/~mike/serial/serial.html > -- > _ _ _ _ _ ___ _ _ _ ___ _ _ __ #define EDINGDONG /* The > / \/ \(_)| ' // ._\ / - \(_)/ ./| ' /(__ daemon is dead */ > \_||_/|_||_|_\\___/ \_-_/|_|\__\|_|_\ __) > [ Mike Hicks | http://umn.edu/~hick0088/ | mailto:hick0088 at tc.umn.edu ]