[nos-bbs] attach asy patch (for NinoTNC) but need to test all hardware

[maiko at pcsinternet.ca]

I should clarify that the LOW LATENCY FLAG is only enforced for cases
where you specify the f option, such as is currently done for NinoTNC
use, I am investigating how we can ditch the option, since the python
serial functions don't seem to need it. In other words, it is not set
for 'regular' TNC use. IF you are seeing garbled callsigns or 'corrupt
AX25 headers', please trace your port for me (trace port 0x211 
trace.log)
so that we can see for real what is actualy happening. I doubt we would
need to set the flag for all cases of 'attach asy', but perhaps it might
be needed ?

The PySerial code does it for all ports, but I have read that in some
cases it's best to leave it off. Took this long for me to notice, live
and learn - never expected LATENCY to be an issue on linux ? wrong.

Maiko / VE4KLM

On 2024-09-27 14:33, Maiko Langelaar (Personal) wrote:
> Good day,
> 
> Been a while since I last posted, more changes coming, but I felt we 
> need to
> get this one out to the masses. Even if you are NOT using a NinoTNC, 
> this fix
> to the unixasy.c source file affects all hardware initialized through
> 'attach asy',
> so any kam, pk232, teensy, tinytnc (?), whatever, please give this a 
> read :
> 
> https://www.langelaar.net/radio/ve4klm/jnos2/documents/development.txt
> 
> So for the few remaining folks using 'attach asy', please test and make 
> sure
> your existing hardware still works properly, report any issues direct 
> to me.
> 
> We appreciate your support and time, thank you.
> 
> Maiko / VE4KLM