Mon, Sep 13, 2021 at 03:32:19PM +0200, groepaz_at_gmx.net wrote: >Am Montag, 13. September 2021, 15:15:35 CEST schrieb Marko Mäkelä: >> Hi Michał, >> >> >>Somewhat unrelated topic: Is there is a virtual serial port that would >> >>allow cbmlink to work across the RS-232 interface on the CBM II and >> >>264 series machines? I am thinking how feasible it would be to >> >>implement a virtual C2N232 device. >> > >> >What do you mean by virtual serial ports? >> >> Sorry for being unclear. I was referring to VICE. It would be great if >> an emulated 6551 or CMD SwiftLink or whatnot would be accessible to >> programs running on the host system. On Microsoft Windows, that would be >> called "Virtual COM port". I realize that it might require an operating >> system kernel module. For example, a regular Unix domain socket or named >> pipe may not be able to support all primitives defined in <termios.h>. > >they can be used over a network socket (which may or may not work for whatever >you want to do). i am smelling x/y problem here though - so what is the actual >usecase here? :) The use case would be using cbmlink between the RS-232 port of an emulated 8-bit Commodore computer and a host system. I would not expect the <termios.h> functions to be defined on network sockets, but I must admit that I never tried it. I would think that at least sending or receiving a BREAK signal would require some special handling and is likely not supported over TCP/IP. But, it does not look like the BREAK signal is being used in that protocol. It is used on the C2N232 as a kind of an interrupt to reset the device to a known state. But, sending a NUL byte should be equivalent when the device already is in the idle state. MarkoReceived on 2021-09-13 19:00:03
Archive generated by hypermail 2.3.0.