Just downloaded FLAMP from Source Forge, Issue present.

DE KI4DKS

FLAMP Relay


DE KI4DKS K

go figure.

Turns out the HASH values for my copy of FLAMP 2.2.12 differs from Source Forges version 2.2.12. Something fishy is going on.
Please standby while I'll investigate further.

Robert

Better yet perform a binary compare with my copy to yours, that way you don't have to execute it.
Robert

开云体育

Please modify this file: src/dialogs/fl_digi.cxx

line 1647, was "export, "cabrillo",

change to: "export", "cabrillo", "UDP fields",

recompile and test

David

To clarify, is this issue on Linux and Windows?
Robert

Well FUBAR,

DE KI4DKS

FLAMP Relay

<DATA 72 905F>{8F4F:2}L0c3LsDOqjV4P73UNoaMhZztJmI4HPDaEN3x3FU8fB6juA6roP7nPOVtxI7jbXuW
<DATA 72 8873>{8F4F:4}ZcSu58FeOe14+KjYrwfDfxbcGsdIGmW2gNg+OCKRRiWf/8IHNXWdUumuJoBQSUNL
<DATA 72 C15D>{8F4F:6}T8iHmgUBgC7eqwQfLar30h87qpJJudmqOycIB5WTk1M9gzUH4qt6aATjVJnjZ8ot
<CNTL 10 E8C1>{8F4F:EOF}
<CNTL 10 48CD>{8F4F:EOT}

DE KI4DKS K

If the plot gets any thicker I'm not going to be able to swim in it.

Can you right click on flamp.exe file and from this context menu select CRC SHA -> SHA-256 see if you get the same hash value. You might need to use another option for the SHA-256 hash calculation as this context menu might be from the 7zip install on my machine. I done recall if this is the case.

If you compiled your own version the above would likely fail, recompile with optimization turned off and see if the issue persist.

Name: flamp.exe
Size: 2689536 bytes (2626 KiB)
SHA256: ED1257D69C77FFCBAD6154E11176452D55D03F4C241B4C5012CA6F5FFB454E10


We can perform one more option, going to modify the program to output a bunch of data along the function path see what shows up. I'll post here once I hear back from you.

Robert

My experience running fldigi version 4.2.05.20 on Linux Mint 21.3 (and the terminal error spew) is very similar to yours.

I didn't realize that fldigi was crashing out instead of closing gracefully until I ran it from the command line. After clicking the Close (X) button, it takes 3 seconds for the window to disappear, but that's also true of version 4.2.05, which exits normally. The crash is odd because the program functions as it should as far as I can tell. On exit there should be nothing for the program to do but end its processes and release the resources it was using.

Robert:

Your response reminded me that I have an "ancient" flamp.prefs file as I haven't needed to reconfigure it for a long time.
I deleted it and "forced" a new one.? I see there is at least one new field.
Regardless, the replacement didn't affect anything.

Here is the information requested...

The "screenshot" is what gets Tx'ed.
I have tested any number of individual block relays, singles and multiples (with both spaces and commas as delimiters) with no difference.
Typical modem is MFSK-32, but I have tried others in testing without any change.

I should add that the Windows machines (running an older version of fldigi) is only used for SDR reception and not connected to a radio.? I "tested" them (after updating flamp to v2.2.12) in order to see if it was a Linux issue.

This evening, I recompiled flamp on one of the three Linux boxes (all are connected to a radio) and still get the "aborted" relay.
I seem to recall that previously the "blocks" field used to persist until AFTER the relay was complete before it was cleared (per the configuration setting).? For v2.2.12, it is cleared BEFORE the Tx even starts.

Thanks for your help.

de KI4DKS

I followed the instructions.? Everything looked good.? Dialout was previously setup. My device was ttyusb4.? I pressed the "Test CAT" button.? It was unusually slow, but it eventually allowed the "Test PTT" button.? When it was pressed, I got an error.? ?

--
KE0VCD?
ARES
SATERN
CERT

I see you are using 4.02.05 on Linux and .04 on Windows. Should have see that right from the start. FLAMP required FLDIGI 4.02.04.02 or higher for full functionality, try the latest version of FLDIGI see if that corrects the issue. If not please send the above requested data.

Robert

Good day David.

I can’t understand why in MFSK-xx modes RST “default” is -10. In other modes 599. I noticed this not only in Alpha versions, but also in previous releases. I reviewed all the settings several times, I thought there were default RST values for various mods. Have not found. I don’t consider this some kind of “big” bug)))

TNX

Looks like we need to drill down a little farther. The image is what I'm seeing with version FLAMP 2.2.12. Going to need some very specific information.

From FLAMP select File->Folders, From this file location please post FLAMP.pref, the relay file in question (located in the relay folder). What blocks did you select, version number of FLDIGI and what modem was used to transmit the data.

Good information, Thanks, Robert

开云体育

try these

Dave

Spoke to a friend who runs Win11 and uses the ALPHA versions.
This issue appears AFTER 2.2.11.1 and continues to 2.2.11.14

I don't know if the Linux ALPHA versions have this issue.

de KI4DKS

On Mon, May 6, 2024 at 10:36 AM, Dave, W1HKJ wrote:

<TX>PARIS PARIS PARIS PARIS PARIS PARIS PARIS PARIS PARIS PARIS E<RX>

Dave, you are my friend forever!
If this does not help me I'll switch to FT8 only, forever. :)

Now to have my wife add some unknown text and for me to try it without looking at the screen since it displays it all in the beginning.

I wonder if the ARRL bulletins are available in text format. That might work.

Thanks again for all your efforts.
73, Bo W4GHV since '54

OS Versions:

Win10:? fldigi v4.2.04, flamp v2.2.12, standard Windows precompiled
Win 11: fldigi v4.2.04, flamp v2.2.12, standard Windows precompiled
Linux Mint v21.3 (3 machines):? fldigi v4.2.05, flamp v2.2.12, compiled on machine

Already tested unchecking and checking "Clear missing blocks on non-canceled transmit(s)"

Relay file complete, following initial Rx and also tested after restart & reloading the queue.
Also tested with multiple other complete relay queues.

Tested adjusting the "Hdr Rpt" & "Data Rpt" fields, no joy.

It sends a full relay correctly, ONLY a partial fill is affected.

By "rolling back" to flamp v2.2.11 on all machines, it works correctly.

de KI4DKS

Look like your sound card ID name (or the code that processes it) is the issue, return a bad string which libstdc++ choked on.

Robert

Hey all,

I did a fresh install of Linux Mint 21.3 this morning, making a dual boot Windows 10 / Linux system. I followed the wiki instructions here:



to compile and install flrig-2.0.05.45 and fldigi-4.2.05.20. Everything seemed to go fine, with one small hiccup in the wiki instructions:

$ sudo aptitude build-dep fldigi

did not work, even though aptitude is installed. So I tried this:

$ sudo apt build-dep fldigi

and that seemed to work fine. No warnings or errors where given when running configure and make. Fldigi seems to be functioning just fine.

But, when I close fldigi, it is giving me a "Aborting fldigi due to a fatal error" message and a stack trace in the shell (see the attached file for the full message). This only seems to happen after I have transmitted with fldigi, and then close it. If I open fldigi and then immediately close it without transmitting anything, I do not get the "Aborting" message. I guessing that maybe libstdc++.so.6 and/or libfltk.so.1.3 are the likely culprits?

So next I uninstalled fldigi 4.2.05.20, and then compiled and installed fldigi 4.2.05. I do not get the "Aborting" message when closing fldigi any longer.

I guessing maybe there is a dependency with 4.2.05.20 that I am missing? I'm not sure what to check. Any ideas?

Thanks,
Steve K4RVA


PS. I have used Linux Mint, and other flavors of Linux, on this hardware in the past but I do not recall this happening before. Here are the likely relevant bits:

Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz
16GiB
GeForce GTX 1050 Ti
C-Media Electronics Inc. USB Audio Device (a Sabrent USB sound card)
Prolific Technology, Inc. PL2303 Serial Port / Mobile Action MA-8910P (USB-to-serial interface)

One other thing you can try, increase the relay repeat header and data counter to more then one.

Robert

Looks like I'm going to need more information, on Windows 10 I'm not seeing the issue.

Try reproducing the issue by using the following configuration checkbox enabled and disabled.

[ ] Clear missing blocks on non-canceled transmit(s)

Any differences?

Is this version of FLAMP self compiled or from a repository.

Was the relay file a completed file during the relay operation or fragmented? If fragmented did FLAMP contains the requested fills sections?

Try relay again with a complete relay file to make sure that isn't the issue.

Robert