Watchdog Plugin

[johnwilde]

Thanks for replying. I thought this would lead to a simple fix of just converting the bat to exe. I did this and was left with an exe that works just like the bat when applied manually, but it still fails to load from the command function of watchdog.

I tried your example of calc - and like you I found it worked as a command response (opening up on alarm). Not sure what the difference is?

Anyway, I found a temporary solution that works. I simply put the text of my BAT directly in to the 'command' box of Watchdog. There does constrain content due to the number of characters, but there is enough to get message out.

I was using Swithmail as the software referenced in the manual Send-It-Quiet, doesn't seem to be available anymore.

Would be nice to get a fix, but at the moment I can basic requirement of an e-mail if the boat drags on its mooring.

Thanks again.

[Hakan]

Very well done!

So what's happening is probably the commend is calling the mail program with a switch to send the mail. So the one "answering" is the mail program's .exe-file responding a proper PID to make wxExecute happy?


Would you mind share the command string you used to get i working? I'm curious and others may benefit.
Thanks
Håkan

[johnwilde]

I think you are correct. Maybe I could have played around with making a better exe that the process ‘accepts’?

The command I used was just copied from the BAT files created by the SwithMail software, with a few tweaks. The command (anonymised) is:

C:\PATH-TO\SwithMail.exe /s /from "SENDEREMAIL@gmail.com" /name "SENDER NAME" /pass "SENDER PASSWORD" /server "smtp.gmail.com" /p "587" /SSL /to "RECIPIENT EMAIL" /sub "Mooring Alarm!" /b "Check mooring" /html /priority "high"

The changes required were to provide the full path to the software, rather than just a relative path set in bat. Also, I had to trim the script, removing “cc” etc and keeping text body short to get to sufficiently limited characters.

Obviously SwithMail needs to be installed and path modified to wherever it is installed.

Also, for this use of gmail you need to use a App Specific Password.

Seems to work though - which I’m happy about.

Thanks again for pointing me in the right direction.

[Hakan]

John...
Fantastic you get this to work!
I understand the hard visit will be less nervous?

Many thanks for sharing.

[Rasbats]

John ...

Thanks. Using "mailsend-go" works with your fix as well.
https://github.com/muquit/mailsend-go

Mike

[Be Free]

In my experience, (in the Windows world) if the script will run in a command prompt but does not work when it's called by another program the problem is usually one of permissions or paths.



Make sure that the process that watchdog is using has permission to access the network and make sure that all programs and files are referenced by their full path.


One small point - .bat was used for scripts on the 16 bit version of Windows (pre-NT). From NT forward the preferred extension is .cmd (ignoring powershell scripts for now). There may be some slight differences in the way the script runs as .bat vs .cmd. It is rare and hard to troubleshoot so just avoid the .bat extension all together unless you really intend to use the 16 bit command interpreter (assuming it's even still there).

[Corsair63]

I have a problem with Watchdog plugin.

O ver 5.8.2 latest version of plugin in master catalogue (windows 11 ARM64).

all data comes from Shipmodul Multiplexer via UDP.

I have two echo-sounders one connected to multiplexer via 0183 input and another one also to multiplexer via N2K connection. (both of them Furuno)

Set-up watchdog alarm for depth, with 0183 sounder O displays the depth but plugin doesn't recognise the depth. Nmea log stream attached and also screen shot 1,2 & 3.
if I switch on the plotter (2nd echo-sounder) the display of the plugin recognise the depth and works properly (screen-shot 4)

what is wrong with the readings of the 0183 sounder?
because data is shown in O.

[AedanC]

One of those two sentence types (the one with no f and M in it) does not comply with the NMEA format published at:

https://gpsd.gitlab.io/gpsd/NMEA.htm...low_transducer

[stevead]

I have no idea which repo is used to build the plugin.

However looking at what I presume to be the original source at https://github.com/seandepagnier/watchdog_pi the bug lies not in the NMEA data stream that is being received, but in the watchdog plugin's NMEA parser.

The problem seems to lie in the Depth (DPT) sentence.

A sentences of the form $--DPT,x.x,x.x,x.x*hh<CR><LF> is a valid version 3 sentence.

A sentence of the form $--DPT,x.x,x.x*hh<CR><LF> is a valid version 2 sentence, it omits the maximum range field.

The watchdog's NMEA parser was written for DPT version 2 sentences, which is why the plugin is working for one transducer and not the other. There is an updated NMEA parser that supports both version 2 and 3 sentences which would be the appropriate fix for whoever maintains the plugin.

My guess is that your NMEA 183 transducer (or the chartplotter it is connected to) generates version 2 sentences, whereas the version 3 sentences are presumably being generated by the Ship Modul multiplexer, when it converts the data from your NMEA 2000 depth transducer.

I don't know why it is not working with the Depth Below Transducer (DBT) sentence, which is of the form $--DBT,x.x,f,x.x,M,x.x,F*hh<CR><LF>

Perhaps try filtering the DPT sentence to see if erroneous DPT sentences are causing the plugin to ignore DBT sentences.

Note also the Depth Below Surface (DBS) sentence. Not sure what is generating it but it is misconfigured as it should not have the same values as Depth Below Transducer (DBT). It's irrelevant anyway as the watchdog plugin does not parse DBS sentences.

[Corsair63]

Confused:

I've been investigating about the received NMEA sentences from both devices also from your link and my NMEA library info.

DBT - Depth below transducer
1 2 3 4 5 6 7
| | | | | | |
$--DBT,x.x,f,x.x,M,x.x,F*hh<CR><LF>
Field Number:

Water depth, feet
f = feet
Water depth, meters
M = meters
Water depth, Fathoms
F = Fathoms
Checksum

In real-world sensors, sometimes not all three conversions are reported. So you might see something like $SDDBT,,f,22.5,M,,F*cs

Example:
$SDDBT,7.8,f,2.4,M,1.3,F*0D

Furuno Echo-Sounder (FCV-587):
$SDDBT,10.7,f,3.3,M,1.8,F*39

Furuno Plotter GP1971F:
Doesn't generate neither sends that sentence

As far as I see the sentence from Echo-sounder completely adhere to NMEA standard.

DPT - Depth of Water
1 2 3 4
| | | |
$--DPT,x.x,x.x,x.x*hh<CR><LF>
Field Number:

Water depth relative to transducer, meters
Offset from transducer, meters positive means distance from transducer to water line negative means distance from transducer to keel
Maximum range scale in use (NMEA 3.0 and above)
Checksum

This sentence was incorrectly titled "Heading - Deviation & Variation" in [BETKE]. It’s documented at http://www.humminbird.com/normal.asp?id=853

Example: $INDPT,2.3,0.0*46

Furuno Echo-Sounder (FCV-587):
$SDDPT,3.3,0.0,9999.9*6C

Furuno Plotter (GP1971F):
$SDDPT,3.25,0.00,*7F

deducting from above the one that really send a conforming DPT sentence is the sounder (0183), just the one that is not read by watchdog but read by OCPN.
and the one from the plotter that is not conforming with the standard (or the third field consider 0.0) but is displayed in watchdog and also read by OCPN.

also I see that the example sentence doesn't conform with the standard missing one field for the max. range as per sentence explanation.

or am I wrong?

The one processed by shipmodul would be the plotter one (N2K)

[Corsair63]

Originally Posted by stevead

...
The problem seems to lie in the Depth (DPT) sentence.

A sentences of the form $--DPT,x.x,x.x,x.x*hh<CR><LF> is a valid version 3 sentence.

A sentence of the form $--DPT,x.x,x.x*hh<CR><LF> is a valid version 2 sentence, it omits the maximum range field.

Hi Steve,

as far as I see in my both DPT sentences are of 3 fields, because the one generated (guess) by shipmodul coming from plotter N2K also has it by means of no data but the coma is in there as third separator.

Furuno Echo-Sounder (FCV-587) from 0183:
$SDDPT,3.3,0.0,9999.9*6C

Furuno Plotter (GP1971F) from N2k:
$SDDPT,3.25,0.00,*7F

and this second one is the one working, why? if it has also 3 data fields.

because if it would be generated with only two fields would be:
$SDDPT,3.25,0.00*7F

[Stadlo2]

Hello, I used flatpak to install OpenCpn 5.8.2-0 on my RPI4 (Debian 64) The Watchdog plugin doesn't appear in the plugin menu and I couldn't find a way to install it...Any assistance will be greatly appreciate.
BR, Yannick

[rgleason]

Originally Posted by Stadlo2

Hello, I used flatpak to install OpenCpn 5.8.2-0 on my RPI4 (Debian 64) The Watchdog plugin doesn't appear in the plugin menu and I couldn't find a way to install it...Any assistance will be greatly appreciate.
BR, Yannick

This plugin needs to be rebuilt fusing a new template, and also needs some work, I hope to be doing this in the next week.

[Stadlo2]

Ok, many thanks for that and well appreciated !
BR
Yannick

[Corsair63]

Hi all,

just thinking about the problem of this plugin that do not display the depth from my 0183 Echo-Sounder (from now ES) but displayed the depth from the N2K ES, I tried something else with the 0183.

I have that ES with NMEA output as ver 3.0 giving this output:
$SDDPT,3.3,0.0,9999.9*6C
absolutely conforming with the NMEA standard as per version info of NMEA sentences.
that ES has the option of transmitting version either 1.5 / 2.0 / 3.0 so I switched to version 2.0 and the output was:
$SDDPT,3.1,0.0*55
and voila the plugin started showing the depth.

as per info above it should be considered as two fields sentence that it isn't the standard in new versions of NMEA.

but is displayed a sentence 0183 converted from ES N2K with three fields:
$SDDPT,3.25,0.00,*7F
with null third field instead the proper one from 0183 ES with version 3.0 but only with the third filed as 9999.9

I don't know what to say?

well at least for the time being with version 2.0 in 0183 ES the plugin is working, but should not.

I haven't tested 1.5, it has been too much success and brain storming for one day. And my only neuron is almost firing sparks