Re: [wsjt-devel] High CPU load with 2.7.0-rc2

2023-07-15 Thread Björn Ekelund via wsjt-devel
Crossing posts so I am not sure which post responds to what but using the
win64 hamlib dll linked below,
WSJT-X 2.7.0-rc2 still crashes when trying to use the IC-7610. Regardless
if the radio is on or off.

Björn SM7IUN

On Sat, Jul 15, 2023 at 2:49 PM Black Michael  wrote:

> This has been fixed in the latest hamlib...
>
> New hamlib for installation directions
>
> #1 Shut down WSJTX
>
> #2 Download either the 32-bit or 64-bit DLL matching the 32/64-bit version
> of WSJTX -- hopefully your browser doesn't block it but may warn you
> multiple times.
>
> If you can do a "Save As" you can save it directly in the appropriate
> WSJTX directory
> C:\WSJT\WSJTX\bin and replace the libhamlib-4.dll that is there.
>
> http://n0nb.users.sourceforge.net/dll32/libhamlib-4.dll
>
> http://n0nb.users.sourceforge.net/dll64/libhamlib-4.dll
>
> Linux/Unix/Mac users need to compile the latest tar file from
> http://n0nb.users.sourceforge.net
> Note: If compiling on Unix-like systems please uninstall any Hamlib
> package you have before installing the new build
>
> #3 If you don't save directly you need to open a file browser and move the
> file that way.
>
> If you're not familiar with that here's a video on the file browser -
> https://www.youtube.com/watch?v=AyVqCJrs9dk
>
> Once installed on Linux do "ldconfig".
>
> rigctl --version should then show a relatively recent date of the download
>
> Mike W9MDB
>
>
>
>
>
> On Saturday, July 15, 2023 at 12:30:39 AM CDT, Björn Ekelund <
> bj...@ekelund.nu> wrote:
>
>
> 1.30. I cannot believe why anyone would run an outdated firmware on a
> radio.
> In DXLog.ne we simply do not support outdated firmwares.
>
> Björn
>
> On Fri, Jul 14, 2023 at 12:13 AM Black Michael via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
>
> Also what firmware version are you running?
> I made some changes which SHOULD be both forward/backward compatible with
> the new firmware.
>
> Mike W9MDB
>
>
> On Thursday, July 13, 2023 at 02:07:03 PM CDT, Björn Ekelund via
> wsjt-devel  wrote:
>
>
> The currently published version of the hamlib dll still crashes when
> selecting ICOM IC-7610, regardless of async setting.
>
> It does not seem to work for other ICOM models either, but it does not
> crash the program.
>
> Control via DXLab Commander is a good workaround while this is being
> worked on.
>
> Björn SM7IUN
>
> On Wed, Jul 12, 2023 at 3:20 PM Black Michael via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
>
> Save this into a file called "hamlib_settings.json" and place in the same
> folder as WSJT-X.ini
> Should improve things until I get this fixed.
>
> {
>  "config": {
> "async": "0"
>  }
> }
>
>
> Mike W9MDB
>
>
>
>
>
>
>
>
> On Wednesday, July 12, 2023 at 07:44:53 AM CDT, Roger Newey (K7GXB) <
> k7...@narundi.net> wrote:
>
>
>
>
>
> Hi Mike,
> FYI:
> Win10 Task Manager shows:
>   WSJT-X CPU during Rx: 35-38%
>   WSJT-X CPU during Tx: 35-36%
>
> System:
> WSJT-X running WSPR "wsjtx-2.7.0-rc2-win64.exe" as downloaded 07-Jul-2023
> (Installed over WSJT 2.6.0);
> Rig: ICOM 7300; Win10 on dedicated Celeron J4125 with 16GB RAM;
> (J4125  has Cores. 4 ; Threads. 4 ; Burst 2.70 GHz ; Processor Base 2.00
> GHz ; Cache. 4 MB).
>
> Regret I don't have comparable CPU data for 2.6.0 but will install it if
> that helps.
>
> 73,
> Roger (K7GXB)
>
> -Original Message-
> From: Black Michael via wsjt-devel 
> Sent: Tuesday, July 11, 2023 20:47
> What rig do you have?
> Mike W9MDB
> ...
> On my Win10 PC, this new version idles at 32% usage which is essentially
> the same as the earlier "4.6~git Jun 28" version.  The older
> libhamlib-4.dll v4.5.4 idles at less than 2%.
> Drew K9CW
> ...
> ---
>
>
>
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
>
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] High CPU load with 2.7.0-rc2

2023-07-14 Thread Björn Ekelund via wsjt-devel
1.30. I cannot believe why anyone would run an outdated firmware on a radio.
In DXLog.ne we simply do not support outdated firmwares.

Björn

On Fri, Jul 14, 2023 at 12:13 AM Black Michael via wsjt-devel <
wsjt-devel@lists.sourceforge.net> wrote:

> Also what firmware version are you running?
> I made some changes which SHOULD be both forward/backward compatible with
> the new firmware.
>
> Mike W9MDB
>
>
> On Thursday, July 13, 2023 at 02:07:03 PM CDT, Björn Ekelund via
> wsjt-devel  wrote:
>
>
> The currently published version of the hamlib dll still crashes when
> selecting ICOM IC-7610, regardless of async setting.
>
> It does not seem to work for other ICOM models either, but it does not
> crash the program.
>
> Control via DXLab Commander is a good workaround while this is being
> worked on.
>
> Björn SM7IUN
>
> On Wed, Jul 12, 2023 at 3:20 PM Black Michael via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
>
> Save this into a file called "hamlib_settings.json" and place in the same
> folder as WSJT-X.ini
> Should improve things until I get this fixed.
>
> {
>  "config": {
> "async": "0"
>  }
> }
>
>
> Mike W9MDB
>
>
>
>
>
>
>
>
> On Wednesday, July 12, 2023 at 07:44:53 AM CDT, Roger Newey (K7GXB) <
> k7...@narundi.net> wrote:
>
>
>
>
>
> Hi Mike,
> FYI:
> Win10 Task Manager shows:
>   WSJT-X CPU during Rx: 35-38%
>   WSJT-X CPU during Tx: 35-36%
>
> System:
> WSJT-X running WSPR "wsjtx-2.7.0-rc2-win64.exe" as downloaded 07-Jul-2023
> (Installed over WSJT 2.6.0);
> Rig: ICOM 7300; Win10 on dedicated Celeron J4125 with 16GB RAM;
> (J4125  has Cores. 4 ; Threads. 4 ; Burst 2.70 GHz ; Processor Base 2.00
> GHz ; Cache. 4 MB).
>
> Regret I don't have comparable CPU data for 2.6.0 but will install it if
> that helps.
>
> 73,
> Roger (K7GXB)
>
> -Original Message-
> From: Black Michael via wsjt-devel 
> Sent: Tuesday, July 11, 2023 20:47
> What rig do you have?
> Mike W9MDB
> ...
> On my Win10 PC, this new version idles at 32% usage which is essentially
> the same as the earlier "4.6~git Jun 28" version.  The older
> libhamlib-4.dll v4.5.4 idles at less than 2%.
> Drew K9CW
> ...
> ---
>
>
>
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] High CPU load with 2.7.0-rc2

2023-07-13 Thread Björn Ekelund via wsjt-devel
The currently published version of the hamlib dll still crashes when
selecting ICOM IC-7610, regardless of async setting.

It does not seem to work for other ICOM models either, but it does not
crash the program.

Control via DXLab Commander is a good workaround while this is being worked
on.

Björn SM7IUN

On Wed, Jul 12, 2023 at 3:20 PM Black Michael via wsjt-devel <
wsjt-devel@lists.sourceforge.net> wrote:

> Save this into a file called "hamlib_settings.json" and place in the same
> folder as WSJT-X.ini
> Should improve things until I get this fixed.
>
> {
>  "config": {
> "async": "0"
>  }
> }
>
>
> Mike W9MDB
>
>
>
>
>
>
>
>
> On Wednesday, July 12, 2023 at 07:44:53 AM CDT, Roger Newey (K7GXB) <
> k7...@narundi.net> wrote:
>
>
>
>
>
> Hi Mike,
> FYI:
> Win10 Task Manager shows:
>   WSJT-X CPU during Rx: 35-38%
>   WSJT-X CPU during Tx: 35-36%
>
> System:
> WSJT-X running WSPR "wsjtx-2.7.0-rc2-win64.exe" as downloaded 07-Jul-2023
> (Installed over WSJT 2.6.0);
> Rig: ICOM 7300; Win10 on dedicated Celeron J4125 with 16GB RAM;
> (J4125  has Cores. 4 ; Threads. 4 ; Burst 2.70 GHz ; Processor Base 2.00
> GHz ; Cache. 4 MB).
>
> Regret I don't have comparable CPU data for 2.6.0 but will install it if
> that helps.
>
> 73,
> Roger (K7GXB)
>
> -Original Message-
> From: Black Michael via wsjt-devel 
> Sent: Tuesday, July 11, 2023 20:47
> What rig do you have?
> Mike W9MDB
> ...
> On my Win10 PC, this new version idles at 32% usage which is essentially
> the same as the earlier "4.6~git Jun 28" version.  The older
> libhamlib-4.dll v4.5.4 idles at less than 2%.
> Drew K9CW
> ...
> ---
>
>
>
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


[wsjt-devel] WSJT-X 2.7.0-rc2

2023-07-07 Thread Björn Ekelund via wsjt-devel
Hamlib is broken. Again. Rc2 crashes when I select IC-7610. IC-7300 and
IC-7851 however seem to work.

Björn SM7IUN
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


[wsjt-devel] Memory leak in 2.7.0 rc1?

2023-05-25 Thread Björn Ekelund via wsjt-devel
My RAM (8 Gbyte) fills up if I let rc1 run for a few days. It does not
happen with 2.6.1.

Interestingly enough the task manager does not show WSJT-X to be the
culprit but when
I stop it, up to 5Gbyte are freed up.

Also, sometimes rc1 allocates 630Mbyte of RAM when started. Other times
90Mbyte. It seems random.

2.6.1 is always around 90Mbyte.

I am running Swedish Windows.

Does anyone else experience the same anomalies?

Björn SM7IUN
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] Fwd: Enhancement suggestion - 30 second cycles

2023-05-08 Thread Björn Ekelund via wsjt-devel
Fair enough, they come with keeping the transmit audio frequency centered
in the passband.

"Use optimal transmit audio frequency" then?

Given the huge and often uninformed debates (not only here), avoiding the
s-word would have a lot of value.

Björn SM7IUN

On Mon, May 8, 2023 at 4:29 PM Black Michael via wsjt-devel <
wsjt-devel@lists.sourceforge.net> wrote:

> #1 It allows you to transmit outside your passband.
> #2 It avoids roll off of power at band edges.
> #3 It avoids non-linearities in the audio digitization.
>
> Mike W9MDB
>
>
>
> On Monday, May 8, 2023 at 09:13:02 AM CDT, Björn Ekelund via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
>
>
> So what more does it do?
>
> Björn SM7IUN
>
> On Mon, May 8, 2023 at 3:43 PM Black Michael  wrote:
>
> That's not the only thing it does so that's not a good choice.  Nobody has
> every come up with anything better than Split.
> Let's just learn that Split from 50 years ago has changed and is much more
> flexible now than before.
>
> Mike W9MDB
>
>
>
>
> On Monday, May 8, 2023 at 08:37:26 AM CDT, Björn Ekelund via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
>
>
> I propose that the option "Split operation" in the radio settings of
> WSJT-X is renamed "Transmit harmonics prevention"
> to remove some of the confusion to those who struggle with digital radio
> technology.
>
> The three options below should be "I don't care if I QRM the band", "Use
> VFO B", and "Use CAT".
>
> Björn SM7IUN
>
>
> On Mon, May 8, 2023 at 10:24 AM Jim Brown via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
>
> On 5/6/2023 10:41 PM, Reino Talarmo via wsjt-devel wrote:
> > This start to be really funny!
>
> There's a saying in the States that "you can't argue with a drunk." A
> corollary is that it's a waste of time to one who only wants the world
> to tell him he's right.
>
> The story related in your post is perfect!
>
> 73, Jim K9YC
>
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] Fwd: Enhancement suggestion - 30 second cycles

2023-05-08 Thread Björn Ekelund via wsjt-devel
So what more does it do?

Björn SM7IUN

On Mon, May 8, 2023 at 3:43 PM Black Michael  wrote:

> That's not the only thing it does so that's not a good choice.  Nobody has
> every come up with anything better than Split.
> Let's just learn that Split from 50 years ago has changed and is much more
> flexible now than before.
>
> Mike W9MDB
>
>
>
>
> On Monday, May 8, 2023 at 08:37:26 AM CDT, Björn Ekelund via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
>
>
> I propose that the option "Split operation" in the radio settings of
> WSJT-X is renamed "Transmit harmonics prevention"
> to remove some of the confusion to those who struggle with digital radio
> technology.
>
> The three options below should be "I don't care if I QRM the band", "Use
> VFO B", and "Use CAT".
>
> Björn SM7IUN
>
>
> On Mon, May 8, 2023 at 10:24 AM Jim Brown via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
>
> On 5/6/2023 10:41 PM, Reino Talarmo via wsjt-devel wrote:
> > This start to be really funny!
>
> There's a saying in the States that "you can't argue with a drunk." A
> corollary is that it's a waste of time to one who only wants the world
> to tell him he's right.
>
> The story related in your post is perfect!
>
> 73, Jim K9YC
>
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] Fwd: Enhancement suggestion - 30 second cycles

2023-05-08 Thread Björn Ekelund via wsjt-devel
I propose that the option "Split operation" in the radio settings of WSJT-X
is renamed "Transmit harmonics prevention"
to remove some of the confusion to those who struggle with digital radio
technology.

The three options below should be "I don't care if I QRM the band", "Use
VFO B", and "Use CAT".

Björn SM7IUN


On Mon, May 8, 2023 at 10:24 AM Jim Brown via wsjt-devel <
wsjt-devel@lists.sourceforge.net> wrote:

> On 5/6/2023 10:41 PM, Reino Talarmo via wsjt-devel wrote:
> > This start to be really funny!
>
> There's a saying in the States that "you can't argue with a drunk." A
> corollary is that it's a waste of time to one who only wants the world
> to tell him he's right.
>
> The story related in your post is perfect!
>
> 73, Jim K9YC
>
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] Repeated crashes with large number of decodes

2023-05-02 Thread Björn Ekelund via wsjt-devel
Thanks Joe,

much appreciated.

Björn SM7IUN

On Tue, May 2, 2023 at 10:20 PM Joe Taylor via wsjt-devel <
wsjt-devel@lists.sourceforge.net> wrote:

> Hi Björn,
>
> Thanks for your report.  Of course you are correct about the cause of
> this bug.  The issue was reported in January, and supposedly corrected
> in bug release 2.6.1.  Alas, it was corrected in only one of two
> necessary places.
>
> It will be fixed in the next release.
>
> -- 73, Joe, K1JT
>
> On 5/2/2023 1:51 PM, Björn Ekelund via wsjt-devel wrote:
> > Decoding the DX0NE pile-up repeatedly crashes WSJT-X 2.6.1. There are a
> > lot of decodes. From what it seems, over 100.
> >
> > My guess is that the array mentioned in the crash report simply runs out
> > because nobody thought there
> > would ever be more than 100 decodes in a cycle.
> >
> > Has anyone else experienced this?
> >
> > Björn SM7IUN
> >
> > Subprocess Error
> > Subprocess failed with exit code 2
> >
> > Running: C:\WSJT\wsjtx\bin\jt9 -s WSJT-X -w 1 -m 3 -e C:\WSJT\wsjtx\bin
> > -a C:\Users\SM7IUN\AppData\Local\WSJT-X -t
> > C:\Users\SM7IUN\AppData\Local\Temp\WSJT-X
> > At line 222 of file C:\JTSDK64-Tools\WSJTX\BBdevelop\lib\ft8_decode.f90
> > Fortran runtime error: Index '101' of dimension 1 of array 'allsnrs'
> > above upper bound of 100
> >
> > Error termination. Backtrace:
> >
> > Could not print backtrace: libbacktrace could not find executable to open
> > #0  0x
> > #1  0x
> > #2  0x
> > #3  0x
> > #4  0x
> > #5  0x
> > #6  0x
> > #7  0x
> > #8  0x
> > #9  0x
> > #10  0x
> > #11  0x
> > #12  0x
> >
> >
> >
> >
> >
> > ___
> > wsjt-devel mailing list
> > wsjt-devel@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
>
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


[wsjt-devel] Repeated crashes with large number of decodes

2023-05-02 Thread Björn Ekelund via wsjt-devel
Decoding the DX0NE pile-up repeatedly crashes WSJT-X 2.6.1. There are a lot
of decodes. From what it seems, over 100.

My guess is that the array mentioned in the crash report simply runs out
because nobody thought there
would ever be more than 100 decodes in a cycle.

Has anyone else experienced this?

Björn SM7IUN

Subprocess Error
Subprocess failed with exit code 2

Running: C:\WSJT\wsjtx\bin\jt9 -s WSJT-X -w 1 -m 3 -e C:\WSJT\wsjtx\bin -a
C:\Users\SM7IUN\AppData\Local\WSJT-X -t
C:\Users\SM7IUN\AppData\Local\Temp\WSJT-X
At line 222 of file C:\JTSDK64-Tools\WSJTX\BBdevelop\lib\ft8_decode.f90
Fortran runtime error: Index '101' of dimension 1 of array 'allsnrs' above
upper bound of 100

Error termination. Backtrace:

Could not print backtrace: libbacktrace could not find executable to open
#0  0x
#1  0x
#2  0x
#3  0x
#4  0x
#5  0x
#6  0x
#7  0x
#8  0x
#9  0x
#10  0x
#11  0x
#12  0x
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


Re: [wsjt-devel] Simulating logging of QSO

2022-02-07 Thread Björn Ekelund via wsjt-devel
Thanks, I have of course considered this but I was hoping for something
more integrated,
my assumption being that the developers have implemented this functionality
for their own needs.

Björn SM7IUN

Den tis 8 feb. 2022 kl 05:14 skrev Bill Frantz :

> You can use two or more computers with loud speakers and microphones to
> set up fox/hound testing. You don’t need to go through a sound card or all
> the way to RF, just put them reasonably close so they can “hear” each other.
>
> 73 Bill AE6JV
>
> > On Feb 7, 2022, at 16:25, Björn Ekelund via wsjt-devel <
> wsjt-devel@lists.sourceforge.net> wrote:
> >
> > I'm trying to verify code in a contest logger that receives logged QSO
> via UDP from WSJT-X.
> >
> > Logging "regular" QSO is easy but I struggle to correctly simulate both
> contest and Fox/Hound QSO.
> > Is there perhaps a clever way to simulate logging of such QSO, including
> correct exchanges etc.?
> >
>
>
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel


[wsjt-devel] Simulating logging of QSO

2022-02-07 Thread Björn Ekelund via wsjt-devel
I'm trying to verify code in a contest logger that receives logged QSO via
UDP from WSJT-X.

Logging "regular" QSO is easy but I struggle to correctly simulate both
contest and Fox/Hound QSO.
Is there perhaps a clever way to simulate logging of such QSO, including
correct exchanges etc.?

I currently have both the binary ("5") and the ADIF ("12") formats
implemented.
Is there a general recommendation on which one to use?

Björn SM7IUN
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel