[GTALUG] Actual ttyS0 MIA

[Stewart C. Russell]

On 2020-05-07 4:09 p.m., Alvin Starr via talk wrote:
>
> The headers and cables tended to be pretty standard.

Thanks, Alvin. TIL I learned that they're actually not. Well, there's 
one standard, and there's the thing that ASUS uses. Guess who bought a 
"standard" cable but has an ASUS motherboard?

The "standard" way assumes an IDC 10-way header on the motherboard and 
an IDC 9-pin RS-232 connector on the other end. Because of the 
difference in pin ordering, a straight-through ribbon cable ends up 
mapping header pins 1-9 to 1, 6, 2, 7, 3, 8, 4, 9.

     1 2 3 4 5 6 7 8 9

     1   2   3   4   5
       6   7   8   9

You can see the pinout for this on a weird little embedded 386SX clone I 
have (ZF MicroSystems OEMmodule): 
https://archive.org/details/zf_systemcard_technical/page/10/mode/2up

ASUS, bless 'em, decided that they should be different and mapping 
header pins 1-9 to RS-232 connector pins 1-9. More logical, maybe, but 
it means you can never use IDC 9-pin connectors with an ASUS board.

I've rewired it and all is well.

> There are a number of programs for managing serial ports but I have had reasonably good luck with minicom. 

minicom's usually my go-to program too. One of my devices insists on 
talking 7M1, and minicom can't set that as a default. I had to compile 
the venerable C-Kermit for out-the-box mark parity support. C-Kermit 
doesn't come as an Ubuntu package, and building it is a bit special. 
CUNY stopped supporting it (and Frank) in 2011, so development has 
slowed way down. A shame, 'cos you can bootstrap Kermit using almost 
anything and end up with a robust 8-bit file transfer path.

cheers,
  Stewart