In message <Pine.NEB.4.05.10001190001130.7171-100000@audrey.Ivy.NET>
Miles Nordin writes:
>On Tue, 18 Jan 2000, Jonathan Stone wrote:
>
>> And forcing XON/XOFF with DIRECT_CONSOLE is just ... broken, given the
>> existence of BIOS serial consoles which want RTS/CTS hardware handshaking.
>
>I belive this was a deliberate design, so that you can use three-wire
>cables. 

Ick. I bet none of them are Emacs users, then.

>For example if you are using those DB25-to-RJ11 shims that public
>libraries like to use sometimes, 

Are you sure it isn't DECconnect MMJ, with an off-centre latch?

>or a homemade cable, or anything without
>all seven wires, then you want RTS ignored.  There is an OpenPROM option
>for ignoring RTS on real computers, so DIRECT_SERIAL was (i think) a vague
>effort to bring something almost as good to PeeCee toys. [...]

Maybe two years ago.  These days, PeeCee server motherboards are real
computers, with microcontrollers to do EMP (remote front-panel
reset/power-on over a serial wire), and BIOS options to use serial
consoles.  For which the BIOS would really rather use CTS/RTS
handshaking, it wants DCD asserted, the speed is configurable up to
19.2k or 56k or 115200, there's a Hayes dialler in there, even....
and the only way to get our bootblocks to bypass INT13 also forces
XON/XOFF?  `Ick'.  Are you trying to tell me ``Use FreeBSD'', or what?


>You can't toggle it like in OpenPROM, but the ignore-RTS default
>leaves most people happier.

I'm skeptical. Even if true now, it may not be, very soon.  And even
if you're right, there's a real tension here between old i486 users
and new Pentium-3 motherboards aimed at the "real computer" server
market. So, if NVRAM is not an option, why not put a patchable
parameter block into the bootblock code, with speed, flow control,
"default" console device, etc? And pass it onto the kernel via
bootinfo?  That way, we could potentially slurp it out and preserve it
across upgrades, even.  (Getting it from the bios would be nicer yet,
but that's Real Computer territory again.)

The bottom line is: today there's no way I can both set the serial-
console speed to other than 9600 _and_ use hardware handshaking rather
than XON/XOFF, correct?  That _does_ seem broken, given how new
motherboards want to do real-computer serial consoles.
No two ways about it.

Anyway, enough curmudgeoning. Thanks to you and everyone else for the
help. I just hope I can at least get 9600 Xon/Xoff to work, tomorrow.