🚀 go-pugleaf

louie@sayshell.umd.edu (Louis A. Mamakos) writes:
> Serial ports don't have to interrupt on every character.  If there
> were a moderately intelligent serial controller, plus DMA support for
> the I/O transfers, then things could be quite zippy, thank you.

The validity of this line of thinking depends on what you're doing.
If your idea of serial port traffic consists of unixy transfers,
DMA'd serial ports are great. But if you're exchanging messages in some
higher-level protocol, such as is often encountered in machine-control apps,
this doesn't work very well. A device like the 8530 used by NeXT will
scratch its nose on Sunday if you ask it to, but that's only with respect
to hardware protocol. It still needs some user-programmable software protocol
smarts to back it up. Adding an 8-bit CPU to back the serial ports would make
serial loads just disappear, as far as the main processor is concerned, for a
far wider class of applications.

There's a cultural issue here that's preventing progress toward really-useful
machines. UNIXy capabilities are very important, but they're not the only
kinds of things people do with computers.

The next insanely-great computer needs to do realtime for real. It will be
built a little differently. It will be programmed a little differently. It
will be a *lot* more useful.

Steve abell@netcom.com