开云体育

ctrl + shift + ? for shortcuts
© 2025 开云体育

Re: Help with setup

 

Good morning folks,

I have a Linux box and the problem, that "ADEVICE plughw:1,0" in the config file does not work without complaining that a resource or device is busy. But commenting out this statement works, i.e. I can hear packet sound on the headphones. I am not sure if PTT cm108 works - because it is not yet checked.

Here are some details of my situation:

$ cm108
VID PID Product Sound ADEVICE ADEVICE HID [ptt]
--- --- ------- ----- ------- ------- ---------
** 0d8c 013c USB PnP Sound Device /dev/snd/pcmC1D0c plughw:1,0 plughw:Device,0 /dev/hidraw0
** 0d8c 013c USB PnP Sound Device /dev/snd/pcmC1D0p plughw:1,0 plughw:Device,0 /dev/hidraw0
** 0d8c 013c USB PnP Sound Device /dev/snd/controlC1 /dev/hidraw0

$ lsof /dev/snd/pcmC1D0c
pulseaudi 14545 username mem CHR 116,3 678 /dev/snd/pcmC1D0c
pulseaudi 14545 username 36u CHR 116,3 0t0 678 /dev/snd/pcmC1D0c

$ ps axv | grep 14545
14545 ? S<sl 0:05 186 40 2422663 21372 0.2 /usr/bin/pulseaudio --daemonize=no --log-target=journal

I am currently a bit confused where I probably find the second process of the resource/device for isolating it, so that the plughw statement can be issued. It is up to now unclear why I see two lines after the lsof statement in my case and what actions could follow on.

I would appreciate any helpful ideas on this matter.

Greetings.
Karl, DK7AL


Re: Winlink Error

 

Okay, appreciate it.? Suspected it could be an issue because I did notice a delay and hang time on radio.? Was just throwing me off because I could clearly see messages coming through before winlink disconnected KISS.
?
Got a digirig on the way now.


Re: Winlink Error

 

The direwolf Radio Interface Guide ? ? section 7.4 explains why you don't want to use VOX built in to the radio.?
You need to use a different interface which provides the push-to-talk control to the radio.
?
73,
John WB2OSZ


Winlink Error

 

?
Hey, I've spent a few hours looking for a solution to this,? Sound card is setup correctly, packets are being sent via VOX (uv-5r), stations are replying.? However it seems winlink is disconnecting without acknowledging the RX packets.? ?
?
Any ideas?? Is this most likely a winlink or direwolf setting?? Exactly the same behavior with "soundmodem.exe" program.

Surely I'm missing one simple setting somewhere.
?
Thank you!
?
-Karlan
AI5TS


Re: Multiple Radio Culprit #direwolf #linux

 

Thanks John.? I'm looking into that.


Re: Multiple Radio Culprit #direwolf #linux

 

No the sound cards don't have unique identifiers anywhere I can find.? I'm using a RPI4b.
Thanks for the advice.
?


Re: Multiple Radio Culprit #direwolf #linux

 

tnx
?


Re: Multiple Radio Culprit #direwolf #linux

 

There is an easy solution.
This situation is covered in section 7.6 of the Radio Interface Guide.
?
73,
John WB2OSZ


Re: Multiple Radio Culprit #direwolf #linux

 

开云体育


Yeah, this is a system thing at not a Direwolf thing per se.?? Do these two Digirig units have unique serial numbers?? You can see all the complete details by looking at the output of "sudo lsusb -vvv".? Maybe you're using a Raspberry Pi 5 where one Digirig can be on the USB2 controller and the other digirig is on the USB3 controller?? You ultimately need SOMETHING to distinguish them so UDEV can name them uniquely.

--David
KI6ZHD


On 04/08/2025 02:47 PM, Steven via groups.io wrote:

I've incorporated two radios into my RPI setup.? I have two digirig soundcards configured as channels 0 and 2 in the .conf file? They's both USB audio.? When rebooting Pi, the addresses (device IDs) swap requiring a swap of the USB cables to the digirigs.? Its a PITA.
Is there some way to code each of these USB devices to prevent this?? I realize the problem may be more a Linux issue.
?


Re: Multiple Radio Culprit #direwolf #linux

 

Yes. Run the 'cm108' command. It will identify the USB devices by their ports. It will show you how to make a 'rules' file. Good luck?


Multiple Radio Culprit #direwolf #linux

 

I've incorporated two radios into my RPI setup.? I have two digirig soundcards configured as channels 0 and 2 in the .conf file? They's both USB audio.? When rebooting Pi, the addresses (device IDs) swap requiring a swap of the USB cables to the digirigs.? Its a PITA.
Is there some way to code each of these USB devices to prevent this?? I realize the problem may be more a Linux issue.
?


Re: Direwolf Lives Again... maybe #direwolf

 

April Fool??

On 04/07/2025 5:04 PM PDT Gordon Lyman (glyman2) W4ZD via groups.io <gcl1@...> wrote:
?
?

[Edited Message Follows]

Its funny what you find when you google search direwolf while trying to get your packet station fixed? :)
?
?
?
?
?
?
?
?


Re: Direwolf Lives Again... maybe #direwolf

 

Feels like someone should update the group's avatar picture


Direwolf Lives Again... maybe #direwolf

 
Edited

Its funny what you find when you google search direwolf while trying to get your packet station fixed? :)
?
?
?
?
?
?
?
?


TXTail keeps resetting to 0

 

everything was working fine then the next day I could no longer send messages over kissutil. TXtail keeps changing to 0. I reset to 10 and sttempt to send out a message and it goes back to 0 and gives me an error code. I am still quite green anything will help. Aslo was set in the config file to 10.


Re: Weather reporting in Celsius?

 

This is mentioned in the APRS Protocol Reference, near end of Chapter 5.
?
?
"APRS Data Units
?
For historical reasons there is some lack of consistency between units of data
in APRS packets — some speeds are in knots, others in miles per hour; some
altitudes are in feet, others in meters, and so on. It is emphasized that this
specification describes the units of data as they are transmitted on-air. It is
the responsibility of APRS applications to convert the on-air units to more
suitable units if required."
?
?
For more details, see APRS12c.pdf .
?
73,
John WB2OSZ
?
?


Re: Weather reporting in Celsius?

 

The APRS Protocol Specification states that temperature in weather report packets is always specified in Fahrenheit. Whether your APRS client will convert between those "wire protocol" units and your local units is up to it.

Andrew, KA2DDO
author of YAAC

________________________________________
From: [email protected] <[email protected]> on behalf of Ian Morrison via groups.io <Ianm3920@...>
Sent: Friday, March 28, 2025 10:52 AM
To: [email protected]
Subject: [direwolf] Weather reporting in Celsius?

I have CumulusMX running on a PI3 feeding YAAC with the output sent to Direwolf. All is well. Packets sent with weather and packets received.
The weather starts out reporting Celsius but it is transmitted from Direwolf in Fahrenheit.
I must be missing something, but what?
Help would be appreciated.

Ian - VE3EP


Weather reporting in Celsius?

 

I have CumulusMX running on a PI3 feeding YAAC with the output sent to Direwolf. All is well. Packets sent with weather and packets received.
The weather starts out reporting Celsius but it is transmitted from Direwolf in Fahrenheit.
I must be missing something, but what?
Help would be appreciated.

Ian - VE3EP


Re: Using Direwolf to emulate a host mode TNC

 

Thanks for your input Thomas. I have spent the day getting this
working on a virtual Debian system -- and I specifically picked 32bit
system because I was having issues with 64 bit, and your description
of what is going on there makes perfect sense. Since it is running in
as a guest OS on my mac, I simply point the emulated serial port at
direwolf on a raspberry pi in my garage hooked to a radio and
soundcard. It works flawlessly. I understand it lacks some newer
AX.25 features, but running like this it can open up a world of
possibilities with retro fun. That is all the software that is
looking for a hostmode TNC. I do have real hostmode TNC's, but
direwolf is a little bit better and with a digirig + pi, it is even
more compact.

If I run a windows xp guest in qemu or virtbox, I can have the serial
port point to TFKISS, and I am off to the races with WinFBB, DOSFBB,
or WinPac2.8 (or whatever cool stuff is out there that needs TF2.7b).

Programs looking for WA8DED hostmode proms should work as well as
TF2.7b is just an extension of this hoslmode firmware.

I was looking at it primarily for LinFBB a while back because I wanted
to use direwolf with it, and the AX.25 stack in Linux was broken. Now
that is fixed the importance of it diminishes slightly. It would be
cool for someone to write a hostmode stack that DED compatible from
scratch, but I do not have that much knowledge. I can cludge/hack and
fix things with my own patches a little, but not a wiz bang
programmer.

-73 de Chris KQ6UP

On Mon, Mar 24, 2025 at 6:14?PM Thomas Leibold via groups.io
<leibold@...> wrote:

Hi Chris,

There is a program called TFKISS, but it needs to be ported to a more current version of Linux.
I found two source versions of TFKISS and some more in binary form only. It appears that the authors (from Nord<>Link in Germany) had problems with copyright violations (it is mentioned in the documentation) and possibly stopped publishing source code as a result.

The two versions of code are apparently both from a common ancestor 2.7b dated 10/5/1996.
The first is the port of "The Firmware" from firmware code (to be installed on a TNC-2 compatible hardware TNC) to application software running on Linux.
The second is a version that at some point was released for (or with) OpenBSD and differs from the first mainly through the addition of the configure system with support for Linux, Solaris, NetBSD.

The software compiles and runs on Linux without any modifications (yes, even modern Linux), as long as it is being compiled as a 32-bit application (use the -m32 flag when compiling it on a 64-bit platform). The code is full of pointer to/from integer conversions that can greatly optimize code (when the target is a tiny EPROM) but extremely problematic when those pointers are suddenly 64-bit long!

For use with Direwolf I was unsuccessful in establishing an IP socket connection (tfkiss_extsocket 127.0.0.1:8001). It later turned out that this option doesn't do what I thought it is for (TCP/IP socket connection to a KISS TNC, like Direwolf). However using the pseudo-tty option (-p) in Direwolf, I was able to specify "device /tmp/kisstnc" in the tfkiss.ini file and run tfkiss with Direwolf.
I also played with tfkiss with a hardware KISS TNC over a real serial port and that worked just fine too.

Or at very least provide a UNIX socket like /tmp/tnc0 for TFKISS to connect to.
While a Unix domain socket would provide an additional method for guest applications to connect to Direwolf, I don't think that it would be beneficial in this case:
- the -p (pseudo-ttty/terminal) option is already available and works.
- a UNIX domain socket option would probably have the same problem as the pseudo-tty option (blocking when nobody is reading the data sent to it).
The correct approach (in my opinion) would be to make tfkiss use the TCP/IP socket to connect to Direwolf's KISS network port.

While tfkiss is functional by itself, it appears its main purpose was to be a compatibility layer between a KISS modem and other client software, especially TNT from the same group. When used as a compatibility layer, the tfkiss_extsocket port is what applications like TNT can connect to. Unlike tfkiss (which is usable but not very user-friendly), TNT is/was a full-featured packet communications package (but I haven't tried that software yet).

While it may not take much effort to implement a KISS socket connection in tfkiss (to connect to Direwolf), it may not be the best way to spend your time. The tfkiss software is dated and while it implements AX.25 version 2.0, it does not implement any of the level 2 enhancements such as larger packet sizes (for 9600 baud or faster packet) or improved error handling (selective reject of only bad packets instead of *all* unacknowledged packets). Nor does it support FX.25 forward error correction. I'm fairly certain that it would be less effort to write new software from scratch than trying to bring tfkiss up to current standards.

73,
Thomas
KK6FPP





--
Thanks,
Chris Maness


Re: Using Direwolf to emulate a host mode TNC

 

Hi Chris,

There is a program called TFKISS, but it needs to be ported to a more current version of Linux.
I found two source versions of TFKISS and some more in binary form only. It appears that the authors (from Nord<>Link in Germany) had problems with copyright violations (it is mentioned in the documentation) and possibly stopped publishing source code as a result.

The two versions of code are apparently both from a common ancestor 2.7b dated 10/5/1996.
The first is the port of "The Firmware" from firmware code (to be installed on a TNC-2 compatible hardware TNC) to application software running on Linux.
The second is a version that at some point was released for (or with) OpenBSD and differs from the first mainly through the addition of the configure system with support for Linux, Solaris, NetBSD.

The software compiles and runs on Linux without any modifications (yes, even modern Linux), as long as it is being compiled as a 32-bit application (use the -m32 flag when compiling it on a 64-bit platform). The code is full of pointer to/from integer conversions that can greatly optimize code (when the target is a tiny EPROM) but extremely problematic when those pointers are suddenly 64-bit long!

For use with Direwolf I was unsuccessful in establishing an IP socket connection (tfkiss_extsocket 127.0.0.1:8001). It later turned out that this option doesn't do what I thought it is for (TCP/IP socket connection to a KISS TNC, like Direwolf). However using the pseudo-tty option (-p) in Direwolf, I was able to specify "device /tmp/kisstnc" in the tfkiss.ini file and run tfkiss with Direwolf.
I also played with tfkiss with a hardware KISS TNC over a real serial port and that worked just fine too.

Or at very least provide a UNIX socket like /tmp/tnc0 for TFKISS to connect to.
While a Unix domain socket would provide an additional method for guest applications to connect to Direwolf, I don't think that it would be beneficial in this case:
- the -p (pseudo-ttty/terminal) option is already available and works.
- a UNIX domain socket option would probably have the same problem as the pseudo-tty option (blocking when nobody is reading the data sent to it).
The correct approach (in my opinion) would be to make tfkiss use the TCP/IP socket to connect to Direwolf's KISS network port.

While tfkiss is functional by itself, it appears its main purpose was to be a compatibility layer between a KISS modem and other client software, especially TNT from the same group. When used as a compatibility layer, the tfkiss_extsocket port is what applications like TNT can connect to. Unlike tfkiss (which is usable but not very user-friendly), TNT is/was a full-featured packet communications package (but I haven't tried that software yet).

While it may not take much effort to implement a KISS socket connection in tfkiss (to connect to Direwolf), it may not be the best way to spend your time. The tfkiss software is dated and while it implements AX.25 version 2.0, it does not implement any of the level 2 enhancements such as larger packet sizes (for 9600 baud or faster packet) or improved error handling (selective reject of only bad packets instead of *all* unacknowledged packets). Nor does it support FX.25 forward error correction. I'm fairly certain that it would be less effort to write new software from scratch than trying to bring tfkiss up to current standards.

73,
Thomas
KK6FPP