Minicom - data is corrupted between two PC's ??

[David Timms]

Mikkel L. Ellertson wrote:
> David Timms wrote:
>> Once I checked all the above, one computer can start minicom fine, but
>> the computer {a hp nx6320 notebook} was having:
>> Device /dev/ttyS0 lock failed: Operation not permitted.
>>
>> It seems if I attempt to start minicom from a user account, then the
>> lock file is generated in /var/lock/*tty* , but then has some problem.
>> su to root and running minicom after this shows identical error.
>> rm the lock file and running minicom from root is then successful.
>>
>> However, with a 3-wire crossover between the two PC's at 115200 8n1 just
>> junk chars were coming up on the other computer in one direction and
>> nothing in the opposite. After a time (20s), suddenly junk also appeared
>> on the other computer (not the letters typed but >ascii chars}
>>
>> I reset to 38400 and 9600. result now is one direction characters come
>> out normal, and in the other direction I see weird characters:
>> ?...?p??.?? . . ? ?. ?  ? ? ?   ? . ?? ? ?   .???.??    .???.
>> +++++++++?+#++?++++?+#+?++_++?+??+?_?+_?.......................+++
>>
>> I'm going to find my DB9F plug with a 2-3 short in it later today to
>> test each end separately.
>>
>> It appears that the data rates are mismatched, yet the status text says
>> 9600 8N1 on both machines.
>> Any hints for resolving this one ?
>>
>> Thanks, DaveT
>>
> Did you remember to change to Software Flow Control to yes and
> Hardware Flow Control to no in the port setup on both machines? The
> default is hareware flow control, and that is not going to work with
> a 3-wire cable.
I have both hw and sw hs set to off. In windows, this is how I have 
always used it at 115200, with zero problems. Is linux not so good in 
this regard ?

> You may also want to change to a standard kernel instead of the Zen
> kernel.
I'm using the normal {non zen} kernel.

Thanks, DaveT.