On 02/06/13 17:31, W. Martin Borgert wrote:
>> As a workaround, set up a udev rule to:
>> > - set the group and mode on the device
>> > - create a non-changing symlink name to the changing device name
>> > - tell ModemManager not to touch the device
>> >
>> > Create a file, say /etc/udev/rules.d/99-local.rules
>> >
>> > SUBSYSTEM=="tty", ATTRS{idVendor}=="067b", ATTRS{idProduct}=="2303",
>> > GROUP="users", MODE="0660", SYMLINK+="gps"
>> > ACTION=="add|change", SUBSYSTEM=="usb", ENV{DEVTYPE}=="usb_device",
>> > ATTRS{idVendor}=="067b", ATTRS{idProduct}=="2303",
>> > ENV{ID_MM_DEVICE_IGNORE}="1"
> Many thanks for this idea! For me, personally, it probably is
> not an option, however:
>
> Unfortunately, a "PL2303 Serial Port" is a relatively common USB
> serial adaptor which might be used both inside an USB modem or an
> external adaptor for any kind of serial device. So there is no
> other way than probing the device (e.g. sending AT commands at
> 115200 bps and hoping for a readable response, or just reading at
> 4800 bps and looking out for readable NMEA datagrams).
>
> Maybe in the very distant future, MM would not probe the serial
> device at all - instead I imagine a "serial detector" software
> component, that does all kind of probing and only sends signals on
> the DBus about the result ("modem found at /dev/ttyUSB0", "GPS
> found at /dev/ttyUSB1", "voltmeter found...").
In MM git master, USB<->RS232 adaptors are never probed by default when
detected, so there shouldn't be any direct conflict. Instead, they will
only be probed to look for modems if Scan() is called explicitly.
--
Aleksander
_______________________________________________
networkmanager-list mailing list
[email protected]
https://mail.gnome.org/mailman/listinfo/networkmanager-list
